Pravastatin for lowering lipids

Abstract

Background

A detailed summary and meta‐analysis of the dose‐related effect of pravastatin on lipids is not available.

Objectives

Primary objective

To assess the pharmacology of pravastatin by characterizing the dose‐related effect and variability of the effect of pravastatin on the surrogate marker: low‐density lipoprotein (LDL cholesterol). The effect of pravastatin on morbidity and mortality is not the objective of this systematic review.

Secondary objectives

• To assess the dose‐related effect and variability of effect of pravastatin on the following surrogate markers: total cholesterol; high‐density lipoprotein (HDL cholesterol); and triglycerides. 
• To assess the effect of pravastatin on withdrawals due to adverse effects.

Search methods

The Cochrane Hypertension Information Specialist searched the following databases for randomized controlled trials (RCTs) up to September 2021: CENTRAL (2021, Issue 8), Ovid MEDLINE, Ovid Embase, Bireme LILACS, the WHO International Clinical Trials Registry Platform, and ClinicalTrials.gov. We also contacted authors of relevant papers regarding further published and unpublished work. The searches had no language restrictions.

Selection criteria

Randomized placebo‐controlled trials evaluating the dose response of different fixed doses of pravastatin on blood lipids over a duration of three to 12 weeks in participants of any age with and without evidence of cardiovascular disease.

Data collection and analysis

Two review authors independently assessed eligibility criteria for studies to be included, and extracted data. We entered lipid data from placebo‐controlled trials into Review Manager 5 as continuous data and withdrawal due to adverse effects (WDAEs) data as dichotomous data. We searched for WDAEs information from all trials. We assessed all trials using Cochrane's risk of bias tool under the categories of sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other potential biases.

Main results

Sixty‐four RCTs evaluated the dose‐related efficacy of pravastatin in 9771 participants. The participants were of any age, with and without evidence of cardiovascular disease, and pravastatin effects were studied within a treatment period of three to 12 weeks. Log dose‐response data over the doses of 5 mg to 160 mg revealed strong linear dose‐related effects on blood total cholesterol and LDL cholesterol, and a weak linear dose‐related effect on blood triglycerides. There was no dose‐related effect of pravastatin on blood HDL cholesterol. Pravastatin 10 mg/day to 80 mg/day reduced LDL cholesterol by 21.7% to 31.9%, total cholesterol by 16.1% to 23.3%,and triglycerides by 5.8% to 20.0%. The certainty of evidence for these effects was judged to be moderate to high. For every two‐fold dose increase there was a 3.4% (95% confidence interval (CI) 2.2 to 4.6) decrease in blood LDL cholesterol. This represented a dose‐response slope that was less than the other studied statins: atorvastatin, rosuvastatin, fluvastatin, pitavastatin and cerivastatin. From other systematic reviews we conducted on statins for its effect to reduce LDL cholesterol, pravastatin is similar to fluvastatin, but has a decreased effect compared to atorvastatin, rosuvastatin, pitavastatin and cerivastatin. The effect of pravastatin compared to placebo on WADES has a risk ratio (RR) of 0.81 (95% CI 0.63 to 1.03). The certainty of evidence was judged to be very low.

Authors' conclusions

Pravastatin lowers blood total cholesterol, LDL cholesterol and triglyceride in a dose‐dependent linear fashion. This review did not provide a good estimate of the incidence of harms associated with pravastatin because of the lack of reporting of adverse effects in 48.4% of the randomized placebo‐controlled trials.

Plain language summary

Pravastatin for lowering lipids

Key messages

• Pravastatin decreases low‐density lipoprotein cholesterol and the effect is dependent on the dose over the range of 5 mg to 160 mg.

• Pravastatin at 80 mg/day is the maximal licensed dose.

• From other systematic reviews we conducted, pravastatin has a similar effect on cholesterol to fluvastatin and has a lesser effect on cholesterol than the other statins.

What are cholesterol and blood fats?

Cholesterol is required to build and maintain all animal cell membranes and is critical to human life. Main components of cholesterol are low‐density lipoprotein, high‐density lipoprotein, and triglycerides. Low‐density lipoprotein transports fat molecules around the body in the blood and delivers fat molecules to cells. High‐density lipoprotein removes fat molecules from cells and transports it to the liver. Cholesterol and its components low‐density lipoprotein cholesterol, and high‐density lipoprotein cholesterol can be measured in the blood. Triglycerides are found in all lipoproteins and can also be measured in the blood. Blood fats are thought to be related to adverse events affecting the heart and blood vessels.

What is pravastatin?

Pravastatin is one of a class of medication called statins that lower blood cholesterol. What are other statins? Do they have any unwanted effects?

What did we want to find out?

How do different doses of pravastatin affect fats in our blood?

What did we do?

We searched for studies that compared pravastatin with a non‐active form of treatment (placebo) over a three to 12‐week duration and measured cholesterol and other fats in the blood. We assessed unwanted effects by looking at how many people left the studies because of unwanted effects. We considered people of any age, with and without evidence of cardiovascular disease.

What did we find?

We found 64 trials that studied the effects of different doses of pravastatin in 9771 participants.

Main results

People taking 5 mg to 160 mg of pravastatin per day lowered their low‐density lipoprotein cholesterol by 18.3% to 35.3%.

The difference in unwanted effects between pravastatin and placebo was uncertain in these short‐term studies.

What are the main limitations of the evidence?

Many studies did not report unwanted effects nor the number of people who left the studies because of unwanted effects.

How up to date is the evidence?

The evidence is up‐to‐date to September 2021.

Summary of findings

Summary of findings 1. Effect of pravastatin on low‐density lipoprotein (LDL) cholesterol.

Pravastatin versus placebo for lowering LDL cholesterol
Patient or population: participants with normal or abnormal lipid profiles Settings: ambulatory clinics Intervention: pravastatin Comparison: placebo
Pravastatin dose (3‐12 weeks)	Relative effect (Mean Difference) (95% CI)	No of participants (studies)	Certainty of the evidence (GRADE)	Comments
10 mg/day LDL cholesterol	MD −21.59 (−23.58 to −19.60)	765 (11)	⊕⊕⊕⊕ High	I2 = 33% Effect predicted from log dose‐response curve, −21.7%
20 mg/day LDL cholesterol	MD −25.70 (−26.77 to −24.63)	3117 (22)	⊕⊕⊕⊝ Moderatea	I2 = 58% Effect predicted from log dose‐response curve, −25.1%
40 mg/day LDL cholesterol	MD −28.37 (−29.22 to −27.52)	4869 (35)	⊕⊕⊕⊝ Moderatea	I2 = 74% Effect predicted from log dose‐response curve, −28.5%
80 mg/day LDL cholesterol	MD −34.39 (−37.00 to −31.77)	881 (4)	⊕⊕⊕⊕ High	I2 = 0% Effect predicted from log dose‐response curve, −31.9%
CI: confidence interval; LDL: low‐density lipoprotein; MD: mean difference
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded one level due to inconsistency (I² > 50%).

Summary of findings 2. Effect of pravastatin on withdrawals due to adverse effects.

Withdrawals due to adverse effects (WDAEs)
Patient or population: participants with normal or abnormal lipid profiles Settings: ambulatory clinics Intervention: pravastatin all doses Comparison: placebo
Pravastatin 10 to 80 mg/day (6 to 156 weeks)	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	No of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Risk with placebo	Risk with all doses of pravastatin
WDAEs	50 per 1000	40 per 1000 (32 to 52)	RR 0.81 (0.63 to 1.03)	6187 (33)	⊕⊝⊝⊝ Very lowa,b	The lack of reporting of WDAE in 31 of 64 RCTs is of concern.
*The basis for the assumed risk is the measure of absolute effect with the placebo group. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio; WDAEs: withdrawals due to adverse effects
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect

^aDowngraded one level due to imprecision (wide confidence interval and crosses the null point).
^bDowngraded two levels due to high risk of bias due to selective reporting.

Background

Our objective was to determine the log dose response effect of pravastatin on blood lipids in order to compare the lipid‐lowering potency of pravastatin to the other available statins.

Description of the condition

Cardiovascular disease is a condition that affects the heart and blood vessels, and is a major cause of death and disability in the developed world, accounting for more than one‐third of total deaths (Kreatsoulas 2010). In the USA, cardiovascular disease causes one in three reported deaths each year (CDC 2011; Roger 2011). Existing evidence shows an association between adverse cardiovascular events and blood concentrations of the lipid low‐density lipoprotein (LDL) cholesterol in adults (ACC 2019, ESC/EAS 2020). The current recommended treatment for secondary prevention of adverse cardiovascular events consists of diet and lifestyle changes plus drug therapy with the drug class widely known as ‘statins’ (CTT 2005).

Description of the intervention

Pravastatin was patented in 1980 and approved for medical use in 1989. Pravastatin is rapidly absorbed, reaching peak plasma concentration within one hour and has an elimination half‐life of 2.7 hours (Pan 1993). Pravastatin is partially degraded in the stomach and partially metabolized by non‐cytochrome P450 (CYP) enzymes (Neuvonen 2008). Pravastatin and the six other available statins are prescribed to prevent adverse cardiovascular events and to lower total cholesterol and LDL cholesterol. Importantly, statins have been shown in individual randomized controlled trials (RCTs), and in a systematic review and meta‐analysis of RCTs, to reduce mortality and major vascular events in people with occlusive vascular disease (CTT 2005). Pravastatin has a role in secondary prevention of cardiovascular disease (Sacks 1996). Pravastatin is one of the less frequently prescribed statins. In Canada in 2010, pravastatin represented 2.5% of the statin prescriptions (Goodman 2010).

How the intervention might work

Pravastatin acts in the liver by inhibiting an enzyme early in the pathway for cholesterol synthesis, 3‐hydroxy‐3‐methyl‐glutaryl coenzyme A reductase (HMG‐CoA reductase). This enzyme irreversibly converts 3‐hydroxy‐3‐methylglutaryl CoA to mevalonate (Moghadasian 1999). This reaction is the third step in a sequence of reactions resulting in the production of many compounds including cholesterol and its circulating blood derivatives, LDL cholesterol and very low density (VLDL) cholesterol (Gaw 2000). When pravastatin is ingested, the drug is routed primarily to the liver where it binds and inhibits HMG‐CoA reductase, lowering cholesterol production. This decrease in liver cholesterol activates sterol regulatory element binding protein (SREBP) processing, thereby increasing the number of LDL receptors displayed on liver cell membranes. The SREBPs also increase the amount of HMG‐CoA reductase, but this does not increase cholesterol synthesis because the enzyme is inhibited by pravastatin. The newly produced LDL receptors remove LDL from the blood, and deliver it to the interior of the cell where the LDL is digested and its released cholesterol becomes available for metabolic purposes (Goldstein 2009). By a method using single nucleotide polymorphisms (SNPs), LDL‐cholesterol and Mendelian randomization to analyze causal factors for coronary artery disease was developed (Tan 2020). The prevailing hypothesis is that statins reduce mortality and morbidity in patients with occlusive vascular disease by reducing liver production of cholesterol, thus causing a reduction in blood LDL cholesterol and a resultant decrease in atherogenesis (Taylor 2013). However, the HMG CoA reductase enzyme is also responsible for the production of ubiquinone (co‐enzyme Q10), heme A, vitamin D, steroid hormones and many other compounds. It remains possible that the beneficial effects of statins are due to actions other than the reduction of cholesterol. These other actions have been referred to as the pleiotropic (producing more than one effect) effects of statins (Liao 2005).

Why it is important to do this review

Statins are the most widely prescribed class of drugs in the world. Prescribing of statins is increasing, as are average prescribed doses. At the present time, clinicians have only an approximate sense of the different potency of the different statins. Previous systematic reviews have assessed the effect of statins on serum lipids (Bandolier 2004; Edwards 2003; Law 2003; Naci 2013a; Naci 2013b; Ward 2007). They have demonstrated that different statins have different potencies in terms of lipid lowering and that higher doses of statins cause greater lowering of serum lipids than lower doses (Kellick 1997; Schaefer 2004; Schectman 1996). A systematic assessment of the potency (Seeman 1976), dose‐response relationship, and variability of effect has been published for atorvastatin (Adams 2015), rosuvastatin (Adams 2014), fluvastatin (Adams 2018), cerivastatin (Adams 2020a), and pitavastatin (Adams 2020b). From these reviews, using identical methodology, it was possible to calculate indirect comparisons. These reviews show that atorvastatin, rosuvastatin, fluvastatin, cerivastatin and pitavastatin have similar slopes of the dose‐response relationship, but major differences in potency: pitavastatin is 77 times more potent than fluvastatin, 6.2 times more potent than atorvastatin, 1.7 times more potent than rosuvastatin and 3.3 times less potent than cerivastatin. Overall order of potency of the studied statins is cerivastatin > pitavastatin > rosuvastatin > atorvastatin > fluvastatin. It is possible that, in addition to a difference in potency, the slope of the dose‐response or the variability of response may be different for pravastatin. Statin‐induced myopathy is common to all statins, and limits the use of statins in many patients. Knowledge of the effects of statins on blood lipids can help us to use them more effectively. We have used the percentage reduction from baseline of the following surrogate markers to describe the dose‐response relationship of the effect of pravastatin: total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides (Boekholdt 2012). It is important to know the potency of pravastatin as compared to the other studied statins: rosuvastatin, atorvastatin, cerivastatin, fluvastatin and pitavastatin. Subsequent reviews of the other available drugs in the class (lovastatin and simvastatin) will also be conducted, in order to compare the results of all the statins.

Objectives

Primary objective

To assess the pharmacology of pravastatin by characterizing the dose‐related effect and variability of the effect of pravastatin on the surrogate marker: low‐density lipoprotein (LDL cholesterol). The effects of statins on morbidity and mortality is not the objective of this systematic review.

Secondary objectives

• To assess the dose‐related effect and variability of effect of pravastatin on the following surrogate markers: total cholesterol; high‐density lipoprotein (HDL cholesterol); and triglycerides.

• To assess the effect of various doses of pravastatin on withdrawals due to adverse effects.

Methods

Criteria for considering studies for this review

Types of studies

Randomized placebo‐controlled trials (RCTs). We included only the period of the trial prior to the first cross‐over due to possible carry‐over effect. Confounding factors, inappropriate dosing, no appropriate placebo data, combined cross‐over data, inappropriate outcomes (medians) and inadequate dietary baseline period, were reasons we excluded studies.

Types of participants

Participants could be of any age, with and without evidence of cardiovascular disease. They could have normal lipid parameters or any type of hyperlipidemia or dyslipidemia. Many patients with normal lipid parameters receive statins (especially those being treated for secondary prevention). Statins are similarly effective in reducing cardiovascular events and lipids in people with normal and elevated lipids (HPS 2011). We accepted participants with various comorbid conditions, including type 2 diabetes mellitus, hypertension, metabolic syndrome, chronic renal failure or cardiovascular disease. Studies with subsets of eligible participants were included only if data for that subset were provided separately. If not, we contacted the trial authors to request that data. Participants were excluded from the review due to confounding, data expressed as median percent change from baseline, inappropriate dosing and inadequate dietary baseline stabilization period.

Types of interventions

Pravastatin and placebo must have been administered at a constant daily dose for a period of three to 12 weeks. We have chosen this administration time window to allow at least three weeks for a steady‐state effect of pravastatin to occur and to keep it short enough to minimize participants dropping out. We included studies where pravastatin was administered in the morning or evening or where it was not specified. Trials required a washout baseline dietary stabilization period of at least three weeks, where all previous lipid‐altering medication were withdrawn. This baseline phase ensured participants followed a standard lipid‐regulating diet and helped to stabilize baseline lipid values prior to treatment. In trials where participants were not receiving lipid‐altering medications or dietary supplements before receiving the test drug, we did not require washout baseline dietary stabilization periods. We did not include studies where participants received co‐interventions that affect blood lipids. Pravastatin 10 mg/day, 20 mg/day, 40 mg/day and 80 mg/day are the doses predominantly prescribed. Because of this, and because most of the trials studied these doses, we presented these doses in the summary of findings tables.

Types of outcome measures

Lipid parameters: we presented the mean percentage change from baseline for different doses of pravastatin minus the mean percentage change from baseline with placebo reported at 3, 4, 5, 6, 7, 8, 9, 10, 11 and 12 weeks for each of the lipid parameters below. We conducted two regression analyzes comparing absolute change from baseline versus baseline and percentage change from baseline versus baseline. We combined data from all time points as a mean value. The outcomes did not determine the eligibility of inclusion into the review.

Primary outcomes

LDL cholesterol

Secondary outcomes

• Total cholesterol

• HDL cholesterol

• Triglycerides

• End of treatment variability was expressed as a coefficient of variation defined as (standard deviation/mean) of LDL cholesterol, total cholesterol, HDL cholesterol and triglyceride measurements for each dose of pravastatin and placebo. It is important to know whether pravastatin has an effect on the variability of lipid measures.

• Withdrawals due to adverse effects (WDAEs). This is the most consistently reported outcome and captures the important adverse effects that lead to stopping the drug.

Search methods for identification of studies

The Cochrane Hypertension Information Specialist searched the following databases for randomized controlled trials (RCTs) p to September 2021: CENTRAL (2021, Issue 8), Ovid MEDLINE, Ovid Embase, Bireme LILACS, the WHO International Clinical Trials Registry Platform, and ClinicalTrials.gov. We also contacted authors of relevant papers regarding further published and unpublished work. The searches had no language restrictions.

Electronic searches

The Cochrane Hypertension Information Specialist searched the following databases for RCTs from date of inception for published, unpublished, and ongoing studies.

• The Cochrane Central Register of Controlled Trials (CENTRAL) via the Cochrane Register of Studies (CRS Web) (searched 22 September 2021).

• MEDLINE Ovid (from 1946 to 21 September 2021), MEDLINE Ovid Epub Ahead of Print, and MEDLINE Ovid In‐Process & Other Non‐Indexed Citations.

• Embase Ovid (from 1974 to 21 September 2021).

• Latin American and Caribbean Health Sciences Literature (LILACS) (from 1982 to 22 September2021).

• Clarivate Web of Science ‐ Science Citation Index (from 1900 to 21 September 2021).

• Clarivate Web of Science ‐ Conference Proceedings Citation Index (from 1990 to 21 September 2021).

• ClinicalTrials.gov (www.ClinicalTrials.gov), (searched 22 September 2021).

• World Health Organization (WHO) International Clinical Trials Registry Platform (www.who.it.trialsearch), (searched 22 September 2021).

The information specialist modeled the subject strategies for databases on the search strategy designed for MEDLINE. Where appropriate, the Cochrane Information Specialist combined the subject strategy adaptations of the highly sensitive search strategy designed by Cochrane for identifying randomized controlled (as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). We present the search strategies for major databases in Appendix 1.

Searching other resources

• We searched Epistemonikos to retrieve existing systematic reviews relevant to this systematic review, so that we could scan their reference lists for additional trials.

• We checked the bibliographies of included studies and any relevant systematic reviews identified for further references to relevant trials.

• We contacted experts/organizations in the field to obtain additional information on relevant trials.

• We contacted original authors for clarification and further data if trial reports were unclear.

We included gray literature by searching the following additional resources.

• ProQuest Dissertations and Theses (search.proquest.com/pqdtft)

• US Food and Drug Administration (www.fda.gov)

• European Patent Office (worldwide.espacenet.com)

These resources were searched using the following keywords: pravastatin, pravachol, pravacol, CS 514

Data collection and analysis

For data collection and analysis, we used Covidence (Covidence) and Review Manager (RevMan 5) (Review Manager 2020).

Selection of studies

Initial selection of trials involved retrieving and reading the titles and abstracts of each paper found from the electronic search databases or bibliographic citations. We have provided a PRISMA flow diagram (Figure 1) to show the process. There were two screening steps. In the first step, two review authors (SA and NA) independently screened titles and abstracts and selected potential references. The same two review authors (SA and NA) then independently analyzed the full‐text articles from the yes group to decide on the trials to be included using Covidence software. We resolved disagreements by recourse to a third review author (JMW).

1.

Pravastatin flow diagram

Data extraction and management

Two review authors (SA and NA) independently extracted the appropriate data from each of the included trials. If there was a disagreement over a value, we reached consensus by data recalculation to determine the correct value. We resolved disagreements by recourse to the third review author (JMW). We used a data extraction form that had been piloted in a previous similar Cochrane Review (Adams 2020b). We directly extracted the mean percentage change from the data or calculated it from the baseline and endpoint values. To extract numeric data from graphs we used a program called WebPlotDigitizer 2022, which calculates the appropriate data. We added the calculated data to the Data and analyses section of the review. We extracted standard deviations (SDs) and standard errors (SEs) from the report or calculate them when possible. We contacted experts/organizations and original authors for clarification and further data if trial reports were unclear. If there was disagreement over a value, we reached consensus by data recalculation to determine the correct value.

Using Covidence (Covidence), we extracted the data for each study are as follows.

• Identification: sponsorship source, country and setting of where the trial was done, author's contact details

• Methods: trial design, group, additional methods' data

• Population: inclusion criteria, exclusion criteria, group differences, baseline characteristicsInterventions and comparisons: various doses of pravastatin for interventions and placebo for comparison

• Outcomes: percentage change from baseline for each dose in total cholesterol, low‐density lipoprotein (LDL)cholesterol, high‐density lipoprotein (HDL) cholesterol, triglycerides for treatment period of three to 12 weeks, and WDAEs with no time restriction

Two review authors (SPA and NA) entered data from placebo‐controlled trials from Covidence into RevMan 5) as continuous and dichotomous data (Review Manager 2020).

Assessment of risk of bias in included studies

Two review authors (SPA and NA) independently assessed all trials using the Cochrane risk of bias tool under the categories of adequate sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other potential biases. We resolved disagreements by recourse to the third review author (JMW). We produced risk of bias tables as outlined in the Cochrane Handbook for Systematic Reviews of Interventions, Chapter 8 (Higgins 2017).

Measures of treatment effect

We analyzed the treatment effects as mean difference (MD) in the percentage change from baseline from placebo by the weighted mean difference (WMD) fixed‐effect‐model with 95% confidence intervals (CIs) for each dose for blood total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides. We analyzed the dichotomous outcome WDAEs as a risk ratio (RR) fixed‐effect model of pravastatin versus placebo with its 95% CI.

Unit of analysis issues

The unit of analysis was the mean value for the people completing each trial. In the case of trials with multiple treatment arms with different doses of pravastatin, we corrected the N value of the placebo group by dividing it by the number of comparisons. We calculated the mean from the data from all the time points. For a cross‐over RCT, the data before the cross‐over point was analyzed.

Dealing with missing data

We expected follow‐up to be reasonably high for these short‐term trials. The data, however, represent treatment efficacy and not real‐world effectiveness of pravastatin on these lipid parameters. When data were missing, we requested them from the trial authors. The most common type of value that was not reported was the SD of the change.

In the case of a missing SD for the change in lipid parameters, we imputed the SD using the following hierarchy (listed from high to low preference).

• SD calculated either from the t statistics corresponding to the exact P value reported or from the 95% CI of the MD between treatment groups.

• Average weighted SD of the change from other trials in the review were calculated by outcome (Furukawa 2006).

It is common for the SD to be miscalculated and in order not to overweight studies where it was inaccurately calculated and lower than expected, when SD values were less than 40% of the average weighted SDs, we used the imputed value by the method Furukawa 2006 as was done in other reviews (Adams 2018; Adams 2020a; Adams 2020b).

Assessment of heterogeneity

The Chi² test to identify heterogeneity is not appropriate because it has low power when there are few studies, but has excessive power to detect clinically unimportant heterogeneity when there are many studies. The I² is a better statistic. The I² statistic calculates between‐study variance/(between‐study variance + within‐study variance). This measures the proportion of total variation in the estimate of the treatment effect that is due to heterogeneity between studies. This statistic is also independent of the number of studies in the analysis (Higgins 2002). We explored the cause of heterogeneity when an I² statistic was more than 50%.

Assessment of reporting biases

We assessed publication bias using funnel plots, as outlined in the Cochrane Handbook for Systematic Reviews of Interventions, Chapter 13 (Page 2019).

Data synthesis

We entered all study data into Review Manager 5 as MD using a fixed‐effect model to determine the weighted treatment effect and 95% CIs for blood total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides (Review Manager 2020). We recorded trial data of each study and dose in (GraphPad Prism 4), to yield a weighted least squares analysis based on the inverse of the square of the SE for each lipid parameter, to generate weighted log dose‐response curves. We defined strong dose‐response relationships when the 95% CIs of the slopes do not cross the zero point and weak or no dose‐response relationships when the 95% CIs do cross the zero point. We entered the number of participants in placebo‐controlled trials who prematurely withdrew due to at least one adverse effect in Review Manager 5 as dichotomous data for each dose and all combined doses of pravastatin and reported it as risk ratio (RR) versus placebo (Review Manager 2020).

Subgroup analysis and investigation of heterogeneity

We assessed heterogeneity using the I² statistic (Higgins 2002). If an I² statistic value was equal to or greater than 50%, we attempted to identify possible causes for this by carrying out a number of planned subgroup analyses via meta‐analysis in RevMan 5 between two comparisons for each of the subgroup analyses, if there were sufficient numbers of trials (at least one trial for each comparison, see below).

We conducted subgroup analysis for LDL cholesterol, but not for secondary outcomes, for which the heterogeneity was high (HDL, and total cholesterol) to analyze the subgroups based on the following factors.

• Men versus women (statins tend to be more effective in women than in men).

• Morning administration time versus evening administration time (evening administration could increase lipid‐lowering effect).

• Drug‐industry‐funded versus non‐drug‐industry‐funded trials (drug‐industry involvement tends to bias in favor of the sponsor).

• Twice daily versus single dose (evening administration could increase lipid‐lowering effect).

• Familial hypercholesterolemia versus non‐familial hypercholesterolemia (statins tend to be less effective in people with familial hypercholesterolemia).

Sensitivity analysis

We carried out a sensitivity analysis to determine if imputing standard deviations (SDs) had an impact on the estimates of lipid‐lowering effect.

Summary of findings and assessment of the certainty of the evidence

Review authors SPA and JMW used the GRADE to assess the certainty of the supporting evidence behind each estimate of treatment effect (Schünemann 2019a; Schünemann 2019b). The five GRADE considerations that were used to assess the certainty of the body of the evidence were: risk of bias, inconsistency, indirectness, imprecision and publication bias. We presented key findings of the review, including a summary of the amount of data, the magnitude of the effect size, and the overall certainty of the evidence, in two summary of findings tables. We used Review Manager software to create the summary of findings tables (Review Manager 2020). We preselected two outcomes for inclusion in the summary of findings tables: the primary outcome ‒ LDL cholesterol lowering for efficacy of pravastatin (by doses of 10 mg, 20 mg, 40 mg, and 80 mg/day), and the secondary outcome ‒ withdrawals due to adverse effects for combined doses (10 mg, 20 mg, 40 mg, and 80 mg/day) as a measure of harm.

Results

Description of studies

This review included 64 RCTs involving 9968 intention‐to treat participants of whom 9771 (98.0%) participants had at least one lipid parameter measured and of whom 9437 (94.7%) had LDL cholesterol measured or calculated by Friedewald (Friedewald 1972).

Of the 64 RCTs, 57 were double‐blind, six single‐blind, and one where method of blinding was not reported. The number of placebo and pravastatin participants were 4178 and 5593, respectively. The number of male and female participants reported in 51 of the 64 trials were 4844 and 2731, respectively. Participants could be of any age. Seven trials studied only individuals with familial hypercholesterolemia.

Results of the search

Database searching identified 9352 citations and 644 other resource citations giving a total of 9996 records. After the duplicates were removed, 5769 records remained. The number of irrelevant records was 5577. From these 5769 remaining records, 192 were obtained as full‐text articles and assessed for eligibility. There were 71 citations for 66 excluded studies with reasons. The final number of included trials was 64 (Figure 1).

Included studies

One hundred and twenty‐one citations to 64 RCTs met the inclusion criteria and had extractable data to evaluate the dose‐related blood lipid‐lowering effect of pravastatin. Each included study is summarized in the Characteristics of included studies. The publication languages of the 64 trials were 63 (98.4%) English and one (1.6%) Danish. Fifty‐seven studies were double‐blinded, six were single‐blinded, and one where the method of blinding was not reported. Trials evaluating the lipid‐altering efficacy of pravastatin were first published in 1987 and continued to be published until 2014 (Figure 2).

2.

Number of included trials according to publication year

The baseline mean (range) lipid parameters were as follows:

• total cholesterol, 6.69 mmol/L (4.61 mmol/L to 10.60 mmol/L), 258.8 mg/dL (178.3 mg/dL to 409.9 mg/dL);

• LDL cholesterol, 4.67 mmol/L (2.69 mmol/L to 8.70 mmol/L), 180.7 mg/dL (104.0 mg/dL to 336.4 mg/dL);

• HDL cholesterol, 1.17 mmol/L (0.93 mmol/L to 1.47 mmol/L), 45.18 mg/dL (35.96 mg/dL to 56.84 mg/dL); and

• triglycerides 1.76 mmol/L (0.60 mmol/L to 5.50 mmol/L), 155.6 mg/dL (53.1 mg/dL to 487.2 mg/dL).

Excluded studies

We excluded 66 trials because they did not meet the inclusion criteria. Reasons for exclusion included confounding in which participants received drugs that affect blood lipid concentrations: immunosuppressants such as cyclosporine, protease inhibitors such as ritonavir and indinavir, food supplements such as fish oils, fibrates such as gemfibrozil, fenofibrate and clofibrate, bile acid sequestrants such as cholestyramine, colestipol, colesevelam, the cholesterol absorption inhibitor ezetimibe, the vitamin niacin and the anti‐oxidant drug probucol, inappropriate dosing, inappropriate outcomes such as median percent change from baseline, and inadequate dietary baseline stabilization period. The reasons for excluding each trial are listed in the Characteristics of excluded studies.

Risk of bias in included studies

We assessed all included trials using the Cochrane risk of bias tool under the categories of allocation (selection bias), blinding (performance bias and detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other potential sources of bias.

Allocation

We judged random sequence generation bias to be low in seven trials and unclear in 57 trials. We judged allocation concealment to be low in five trials and unclear in 59 trials.

Blinding

We judged the risk of performance and detection bias for lipid parameters to be low for all the trials because lipid parameters measured in remote laboratories are unlikely to be influenced by lack of blinding. WDAEs is a clinical outcome, which is potentially influenced by lack of blinding.

For WDAEs, we considered there is an unclear risk of bias for the blinding of outcome assessment in all the 33 trials that reported this outcome. For the 31 trials that did not report WDAEs, we judged blinding of outcome assessment as not applicable.

Incomplete outcome data

Incomplete outcome reporting leading to attrition bias was not a problem in this review as few participants were lost to follow‐up. Overall, 98.0% of the participants completed the treatment and 49 trials had low risk of attrition bias. However, in nine trials, the rate of loss to follow‐up was different between the treatment arms, and there was premature discontinuation of the treatment in some studies. Accordingly, these nine trials were rated at high risk of bias. Moderate rates of loss to follow‐up was reported in six trials, with some of the studies not reporting them separately for each arm of the trial. We therefore judged this as unclear risk of bias.

Selective reporting

Out of 64 trials, 61 (95.3%) reported the primary lipid outcome low‐density lipoprotein cholesterol (LDL‐C), thus selection bias was not a potential source of bias for this outcome.

Out of 64 RCTs, only 33 (51.6%) reported WDAEs. Therefore, selective reporting bias was judged an important source of bias for this outcome. See risk of bias tables in Characteristics of included studies, and for the overall risk of bias.

Other potential sources of bias

The main other potential source of bias is industry funding. Out of the 64 trials, 30 (46.9%) reported funding by industry, 14 (21.9%) reported partial funding by industry and government or charity, 4 (6.25%) reported no industry funding, and in 16 (25%) trials the source of funding was not reported. Out of 30 industry‐funded trials, 28 (93.3%) were funded by Bristol‐Myers Squibb, marketers of pravastatin, and two (3.1%) were funded by another pharmaceutical company. The Bristrol‐Myers Squibb trials might be biased in favor of pravastatin and thus might overestimate the treatment effect while trials funded by rival pharmaceutical companies might be biased against pravastatin and underestimate the treatment effect. Bristol‐Myers Squibb versus non‐Bristol‐Myers Squibb‐funded LDL cholesterol efficacy data were available for the doses of 20 mg/day and 40 mg/day. A subgroup analysis was done separately in RevMan 5 (Review Manager 2020). The subgroup analysis revealed that the lipid‐lowering efficacy of pravastatin in Bristol‐Myers Squibb versus non‐Bristol‐Myers Squibb‐funded trials were different for the doses analyzed; 20 mg/day (−26.39% versus −21.84%; P = 0.03) and 40 mg/day (−29.9% versus −26.44%; P = 0.03). The lipid‐lowering effect estimate was greater in the Bristol‐Myers Squibb‐funded trials. Another potential source of bias is publication bias. Funnel plots can detect publication bias due to missing results. Funnel plots in our review did not suggest any publication bias as the plots were symmetrical in appearance.

Effects of interventions

See: Table 1; Table 2

Absolute change from baseline versus percentage change from baseline

We conducted two regression analyses and there was a correlation between the baseline value and the pravastatin effect on LDL cholesterol when the effect was expressed as absolute change from baseline (P = 0.0010) and no correlation when expressed as a percentage change from baseline (P = 0.2220).

Overall efficacy of pravastatin

We entered values from all data describing the efficacy of pravastatin to lower the lipid parameters from the Data and analyses section as generic inverse variance data into GraphPad Prism 4 to yield log dose‐response curves for pravastatin on the different lipids. The overall, pravastatin efficacy for lowering lipids ise summarized in Table 3.

1. Pravastatin overall efficacy.

Pravastatin dose mg/day	5	10	20	40	80	160
Mean difference in the percentage change from baseline from placebo of LDL‐C (95% confidence interval)	−18.12 (−23.61 to −12.62)	−21.59 (−23.58 to −19.60)	−25.70 (−26.77 to −24.63)	−28.37 (−29.22 to −27.52)	−34.39 (−37.00 to −31.77)	−41.30 (−51.04 to −31.56)
Mean difference in the percentage change from baseline from placebo of total cholesterol (95% confidence interval)	−14.08 (−18.34 to −9.82)	−15.00 (−16.95 to −13.05)	−19.06 (−19.91 to −18.21)	−20.24 (−20.94 to −19.55)	−24.64 (−26.55 to −22.73)	−30.70 (−38.39 to −23.01)
Mean difference in the percentage change from baseline from placebo of triglycerides (95% confidence interval)	−1.96 (−15.24 to 11.33)	−5.53 (−11.28 to 0.22)	−11.03 (−13.44 to −8.62)	−15.28 (−17.46 to −13.10)	−19.99 (−24.46 to −15.53)	−24.00 (−31.70 to −16.30)

LDL‐C: low‐density lipoprotein cholesterol

Primary outcome

Low‐density lipoprotein (LDL) cholesterol

In total, 61/64 (95.3%) trials and 9437/9968 (94.7%) participants contributed to the LDL cholesterol data analysis. The effects of different doses of pravastatin on LDL cholesterol are shown in the Data and analysis section (Analysis 1.1; Analysis 2.1; Analysis 3.1; Analysis 4.1; Analysis 5.1; Analysis 6.1). The GraphPad Prism 4 analysis for LDL cholesterol yielded the log dose‐response straight‐line equation, y = ‐11.3 log(x) ‐ 10.4. The 95% confidence interval of the slope is (‐15.20 to ‐7.397). This equation provides the best estimate of the mean percentage reductions in blood LDL cholesterol from baseline for pravastatin doses ranging from 5 mg/day to 160 mg/day; it uses all the available data. Using this formula, the calculated reductions in blood LDL cholesterol for doses of 5 mg per day to 160 mg per day ranged from 18.3% to 35.3%. For every two‐fold dose increase, there was a 3.42% (95% CI 2.23 to 4.58) percentage decrease in blood LDL cholesterol (Figure 3). The certainty of evidence was judged to be moderate to high. The 20 mg and 40 mg daily dose outcomes were downgraded one level due to inconsistency (I² > 50%).

1.1. Analysis.

Comparison 1: 5 mg vs placebo, Outcome 1: LDL cholesterol

2.1. Analysis.

Comparison 2: 10 mg vs placebo, Outcome 1: LDL cholesterol

3.1. Analysis.

Comparison 3: 20 mg vs placebo, Outcome 1: LDL‐cholesterol

4.1. Analysis.

Comparison 4: 40 mg vs placebo, Outcome 1: LDL cholesterol

5.1. Analysis.

Comparison 5: 80 mg vs placebo, Outcome 1: LDL cholesterol

6.1. Analysis.

Comparison 6: 160 mg vs placebo, Outcome 1: LDL cholesterol

3.

Log dose pravastatin response curve for LDL cholesterol

Values represent the results of each trial for each dose comparison. The standard error bars cannot be seen because they all lie within the points

Secondary outcomes

Total cholesterol

In total, 59/64 (92.2%) trials and 9267/9968 (93.0%) participants contributed to the total cholesterol data analysis. The effect of different doses of pravastatin on total cholesterol are shown in the Data and analyses (Analysis 1.2; Analysis 2.2; Analysis 3.2; Analysis 4.2; Analysis 5.2; Analysis 6.2). The GraphPad Prism 4 analysis for total cholesterol yielded the log dose‐response straight‐line equation, y = ‐8.006 log(x) −8.05. The 95% CI of the slope is (−11.08 to −4.928). This equation provides the best estimate of the mean reductions in blood total cholesterol from baseline for pravastatin doses ranging from 5 mg/day to 160 mg/day; it uses all the available data. Using this formula, we calculated reductions in blood total cholesterol for doses of 5 mg per day to 160 mg per day ranged from 13.65% to 25.7%. For every two‐fold dose increase, there was a 2.41% (95% CI 1.48 to 3.34) percentage decrease in blood total cholesterol (Figure 4).

1.2. Analysis.

Comparison 1: 5 mg vs placebo, Outcome 2: Total cholesterol

2.2. Analysis.

Comparison 2: 10 mg vs placebo, Outcome 2: Total cholesterol

3.2. Analysis.

Comparison 3: 20 mg vs placebo, Outcome 2: Total cholesterol

4.2. Analysis.

Comparison 4: 40 mg vs placebo, Outcome 2: Total cholesterol

5.2. Analysis.

Comparison 5: 80 mg vs placebo, Outcome 2: Total cholesterol

6.2. Analysis.

Comparison 6: 160 mg vs placebo, Outcome 2: Total cholesterol

4.

Log dose pravastatin response curve for total cholesterol

Values represent the results of each trial for each dose comparison. The standard error bars cannot be seen because they all lie within the points

High‐density lipoprotein (HDL) cholesterol

In total, 56/64 (87.5%) trials and 8078/9968 (81.0%) participants contributed to the HDL cholesterol data analysis. The effect of different doses of pravastatin on HDL cholesterol are shown in the Data and analyses (Analysis 1.3; Analysis 2.3; Analysis 3.3; Analysis 4.3; Analysis 5.3; Analysis 6.3). The GraphPad Prism 4 analysis showed that pravastatin doses ranging from 5 mg/day to 160 mg/day had no dose‐related effect on blood HDL cholesterol. The slope of the log dose‐response curve and its 95% CI is 2.089 (95% CI −2.607 to 6.785). All doses of pravastatin caused a small increase in HDL cholesterol. We pooled all trials and doses and the magnitude of the increased from 4.53% (95% CI 3.87 to 5.18; Analysis 16.1).

1.3. Analysis.

Comparison 1: 5 mg vs placebo, Outcome 3: HDL cholesterol

2.3. Analysis.

Comparison 2: 10 mg vs placebo, Outcome 3: HDL cholesterol

3.3. Analysis.

Comparison 3: 20 mg vs placebo, Outcome 3: HDL cholesterol

4.3. Analysis.

Comparison 4: 40 mg vs placebo, Outcome 3: HDL cholesterol

5.3. Analysis.

Comparison 5: 80 mg vs placebo, Outcome 3: HDL cholesterol

6.3. Analysis.

Comparison 6: 160 mg vs placebo, Outcome 3: HDL cholesterol

16.1. Analysis.

Comparison 16: all doses of pravastatin vs placebo for HDL cholesterol, Outcome 1: HDL cholesterol

Triglycerides

In total, 51/64 (79.7%) trials and 7533/9968 (75.6%) participants contributed to the triglyceride data analysis. The effect of different doses of pravastatin on triglycerides are shown in the Data and analyses (Analysis 1.4; Analysis 2.4; Analysis 3.4; Analysis 4.4; Analysis 5.4; Analysis 6.4). The GraphPad Prism 4 analysis for triglycerides yielded the log dose‐response straight‐line equation, y = −15.71 log(x) + 9.934. The 95% CI of the slope is (−26.49 to −4.930) This equation provides the best estimate of the mean reductions in blood triglycerides from baseline for pravastatin doses ranging from 5 mg/day to 160 mg/day; it uses all the available data. Using this formula, the calculated reductions in blood triglycerides for doses of 5 mg per day to 160 mg per day ranged from 1.05% to 24.69%. For every two‐fold dose increase, there was a 4.73% (95% CI 1.48 to 7.97) percentage decrease in blood triglycerides (Figure 5).

1.4. Analysis.

Comparison 1: 5 mg vs placebo, Outcome 4: Triglycerides

2.4. Analysis.

Comparison 2: 10 mg vs placebo, Outcome 4: Triglycerides

3.4. Analysis.

Comparison 3: 20 mg vs placebo, Outcome 4: Triglycerides

4.4. Analysis.

Comparison 4: 40 mg vs placebo, Outcome 4: Triglycerides

5.4. Analysis.

Comparison 5: 80 mg vs placebo, Outcome 4: Triglycerides

6.4. Analysis.

Comparison 6: 160 mg vs placebo, Outcome 4: Triglycerides

5.

Log dose pravastatin response curve for triglycerides

Values represent the results of each trial for each dose comparison. The standard error bars cannot be seen because they all lie within the points

End‐of‐treatment variability

We used a one‐way ANOVA to compare coefficients of variation of the end of treatment lipid values between placebo and the different doses of pravastatin. This revealed no significant differences in the end‐of‐treatment variabilities of LDL cholesterol, total cholesterol, HDL cholesterol and triglycerides for pravastatin as compared to placebo.

Withdrawal data

Thirty‐three (51.6%) of the 64 RCTs reported WDAEs. In six trials, no participant discontinued treatment due to adverse effects or died during the study, therefore, risk reduction was not estimable. There was no dose response with respect to WDAEs therefore a pooled estimate for all doses compared to placebo was conducted. This showed a risk ratio (RR) of 0.81 (95% CI 0.63 to 1.03) for WDAEs (Analysis 7.1). For the placebo group, 128 out of 2563 participants were reported to withdraw due to an adverse effect and, for all doses of pravastatin, 124 out of 3491 participants were reported to withdraw due to an adverse effect. We judged the certainty of evidence to be very low. IWe downgraded two levels due to a high risk of selective outcome reporting and one level due to imprecision (including a null effect).

7.1. Analysis.

Comparison 7: All doses of pravastatin vs placebo for withdrawals due to adverse effects, Outcome 1: WDAE

Subgroup analyses

Male versus female participant data were available for the 40 mg/day pravastatin dose. We analyzed these data separately for total cholesterol and LDL cholesterol‐lowering efficacy in RevMan 5. The subgroup analysis revealed that the efficacy of pravastatin was greater in male than in female participants. The efficacy for the 40 mg/day dose (male versus female) was: (−31.75 versus −22.00; P = 0.03) for LDL cholesterol (Analysis 8.1).

8.1. Analysis.

Comparison 8: 40 mg vs placebo male versus female subgroup analysis for LDL cholesterol, Outcome 1: LDL cholesterol

Single‐dose administration versus twice daily dose administration data were available for the 20 mg/day, and the 40 mg/day doses. We analyzed these data for LDL cholesterol‐lowering efficacy in RevMan 5. The subgroup analysis revealed no difference in the efficacy of pravastatin with respect to twice‐daily administration versus single‐dose administration (Analysis 9.1; Analysis 10.1). Comparison of morning versus evening administration time was not possible because no trial reported morning data.

9.1. Analysis.

Comparison 9: 20 mg vs placebo with respect to single versus twice daily dosing subgroup analysis, Outcome 1: LDL cholesterol

10.1. Analysis.

Comparison 10: 40 mg vs placebo with respect to single versus twice daily dosing subgroup analysis, Outcome 1: LDL cholesterol

Bristol‐Myers Squibb‐funded versus non‐Bristol‐Myers Squibb‐funded trial data were available for the 20 mg/day and 40 mg/day doses. We analyzed these data for LDL cholesterol‐lowering efficacy in RevMan 5. The subgroup analysis revealed that the efficacy of pravastatin was greater in the Bristol‐Myers Squibb‐funded trials than in non‐Bristol‐Myers Squibb‐funded trials for the doses analyzed; 20 mg/day (−26.37% versus −21.84%; P = 0.03) and 40 mg/day (−29.9% versus −26.44%; P = 0.03; Analysis 14.1; Analysis 15.1).

14.1. Analysis.

Comparison 14: 20 mg vs placebo with respect to Bristol‐Myers Squibb versus Non‐Bristol‐Myers Squibb subgroup analysis, Outcome 1: LDL cholesterol

15.1. Analysis.

Comparison 15: 40 mg vs placebo with respect to Bristol‐Myers Squibb versus Non‐Bristol‐Myers Squibb subgroup analysis, Outcome 1: LDL cholesterol

Familial versus non‐familial hypercholesterolemia participant data were available for the doses 10 mg/day, 20 mg/day, and 40 mg/day. We analyzed these data for LDL cholesterol‐lowering efficacy in RevMan 5. There was no difference in the efficacy of pravastatin between familial versus non‐familial hypercholesterolemia participants. The efficacy for the 10 mg/day dose (familial versus non‐familial) was −23.9 versus −20.6 (P = 0.50) (Analysis 11.1; for the 20 mg/day dose (familial versus non‐familial) was −23.00 versus −24.07 (P = 0.77 (Analysis 12.1; and for the 40 mg/day dose (familial versus non‐familial) was −29.18 versus −31.50 (P = 0.74 (Analysis 13.1).

11.1. Analysis.

Comparison 11: 10 mg vs placebo with respect to familial versus non‐familial hypercholesterolaemia subgroup analysis, Outcome 1: LDL cholesterol

12.1. Analysis.

Comparison 12: 20 mg vs placebo with respect to familial versus non‐familial hypercholesterolaemia subgroup analysis, Outcome 1: LDL cholesterol

13.1. Analysis.

Comparison 13: 40 mg vs placebo with respect to familial versus non‐familial hypercholesterolaemia subgroup analysis, Outcome 1: LDL cholesterol

Sensitivity analyses

We carried out a sensitivity analysis to determine if imputing standard deviations has an important impact on the effect of pravastatin for total cholesterol and LDL cholesterol.

For total cholesterol the efficacy of pravastatin we compared before and after deselecting all imputed trial data for the doses 5 mg/day, 10 mg/day, 20 mg/day, 40 mg/day and 80 mg/day. This revealed that the effect was insensitive to using imputed SD: 5 mg/day −14.08 (95% CI −18.34 to −9.82) versus −11.4 (95% CI −18.10 to −4.70), 10 mg/day −15.00 (95% CI −16.95 to −13.05) versus −16.26 (95% CI −19.50 to −13.03).(20 mg/day −19.06 (95% CI −19.91 to −18.21) versus −20.82 (95% CI −24.14 to −17.50) 40 mg/day −20.24 (95% CI −20.94 to −19.55) versus −20.07 (95% CI −21.64 to −18.50) and 80 mg/day −24.65 (95% CI −26.55 to −22.73) versus −24.46 (95% CI −26.40 to −22.52).

For LDL cholesterol the efficacy of pravastatin we compared before and after deselecting all imputed trial data for the doses 5 mg/day, 10 mg/day, 20 mg/day, 40 mg/day and 80 mg/day. This revealed the effect was insensitive to using imputed SD: 5 mg/day −18.12 (95% CI −23.61 to −12.62) versus −15.2 (95% CI −23.83 to −6.57), 10 mg/day −21.59 (95% CI −23.58 to −19.60) versus −21.01 (95% CI −25.17 to −16.84), 20 mg/day −25.70 (95% CI −26.77 to −24.63) versus −28.98 (95% CI −33.02 to −24.94), 40 mg/day −28.37 (95% CI −29.22 to −27.52) versus −31.35 (95% CI −33.30 to −29.38) and 80 mg/day −34.39 (95% CI −37.00 to −31.77) versus −34.17 (95% CI −36.83 to −31.52).

Discussion

This review summarizes the short‐term effects of RCTs comparing pravastatin with placebo on blood LDL cholesterol, total cholesterol, HDL cholesterol, triglycerides and WDAEs.

Summary of main results

Daily pravastatin intake is effective in lowering LDLcholesterol, total cholesterol and triglyceride concentrations and does so in a predictable, dose‐related manner. The Table 1 documents the effect of pravastatin on LDL cholesterol over the dose range of 10 mg to 80 mg/day, the range for which pravastatin is commonly prescribed, and for which this systematic review has the most data. Over this range, LDL cholesterol is decreased by 22% to 34%.

A regression line was calculated for LDL cholesterol. This regression line provided the best estimate of the treatment effect, because it is based on a regression line calculated from all the data for all the doses. The estimates of the average treatment effect from the regression line are similar to the mean value for all the data for each dose (see Table 1).

Furthermore, we have established, using regression analysis, that there is a correlation between the baseline value and the pravastatin effect on LDL cholesterol when the effect is expressed as absolute change from baseline (P = 0.0010). There was no correlation between the baseline value and the pravastatin effect when the effect is expressed as percent reduction from baseline (P = 0.2220). This finding provides support for the approach we have used in this review and explains why systematic reviews reporting the effect of statins on absolute changes in lipid parameters are problematic and potentially misleading (Law 2003; Naci 2013b).

End‐of‐treatment variabilities expressed as a coefficient of variation of pravastatin and placebo were compared to determine whether pravastatin has an effect on variability of blood lipid measurements. Pravastatin did not significantly affect the end‐of‐treatment variabilities of LDL cholesterol, total cholesterol, HDL cholesterol and triglycerides. This is in contrast to atorvastatin (Adams 2015), and fluvastatin (Adams 2018), where those two statins were found to increase variability of total cholesterol and LDL cholesterol.

The greatest value in doing this type of review is the ability to compare pravastatin to other statins in systematic reviews we have conducted using an identical approach. At present, we can compare it to atorvastatin (Adams 2015), rosuvastatin (Adams 2014), fluvastatin (Adams 2018), cerivastatin (Adams 2020a, and pitavastatin (Adams 2020b). The most important finding in this review is that the slope of the dose‐response effect for pravastatin was less than the other five statins studied, which were all similar. For every two‐fold dose increase of pravastatin, there was a 3.4% (95% CI 2.23 to 4.6) percentage decrease in blood LDL cholesterol. In contrast, for every two‐fold dose increase in atorvastatin, rosuvastatin, fluvastatin, cerivastatin and pitavastatin, there was a 4.9% (95% CI 4.5 to 5.4), 4.42% (95% CI 3.85 to 4.99), 6.0% (95% CI 5.4 to 6.6), 6.01% (95% CI 5.61 to 6.40), and 5.35% (95% CI 3.32 to 7.38) decrease in LDL cholesterol respectively. We do not have an explanation as to why the dose response slope is less for pravastatin. We do not think it likely that it means pravastatin is working by a different mechanism of action. There are possible pharmacokinetic explanations such as that the bioavailability of pravastatin decreases with higher doses. The reduced slope could be advantageous as it may mean that pravastatin is less susceptible to dose‐related toxicity.

It is clear that pravastatin is one of the least potent statins. It is of similar potency to fluvastatin and less potent than atorvastatin, rosuvastatin, cerivastatin and pitavastatin for lowering LDL cholesterol.

Does pravastatin increase withdrawals due to adverse effects?

Of 64 placebo‐controlled trials, 33 (51.6%) reported withdrawals due to adverse effects (WDAEs). This analysis represented only 6054 participants, 3491 of whom received pravastatin and 2563 of whom received placebo. The pooled estimate for all doses provided a risk ratio (RR) of 0.81 (95% CI 0.63 to 1.03) (very low certainty of evidence). It was downgraded predominantly because 31 (48.4%) of the 64 placebo‐controlled trials did not report WDAEs. We judged that the risk of selective reporting bias for this outcome was high. WDAE is an outcome that should be reported in all trials and may have deliberately been not reported in trials where they were increased. Risk of participant selection bias is also possible in these trials, as the participants studied could have been selected because they were known to tolerate statins at baseline. Thus, we do not think that pravastatin reduces WDAE as the data suggest.

Furthermore, in this review the slope of the dose‐response effect for pravastatin was less than the other five statins studied. The reduced slope could be advantageous as it may mean that pravastatin is less susceptible to dose‐related toxicity. However, WDAE in these short‐term trials is a very weak and limited way of assessing the harms of pravastatin.

Overall completeness and applicability of evidence

This review included 64 trials with 9771 participants. As such, it provided robust evidence of the dose‐related lipid‐lowering effects of pravastatin. The population studied in this review varied widely but is a good reflection of the patients being treated with statins at present. The average baseline LDL of 4.7 mmol/L is on the high side reflecting that some of the trials were limited to patients with familial hypercholesterolemia. However, the findings of the review can be generalized to the patient populations where pravastatin is being prescribed.

It was unknown when we did the review whether the time or method of dosing of pravastatin is important with respect to lipid lowering. Unfortunately, it was not possible to do a subgroup analysis comparing morning and evening administration. A subgroup analysis comparing twice‐daily versus single daily regimen data was possible for the doses 10 mg/day, 20 mg/day and 40 mg/day. The percentage reductions with twice‐daily dosing were not different from single daily dosing suggesting that evening administration does not increase the effect. This also supported the fact that we combined data from both dosing regimens. A Cochrane Review has attempted to answer this question and concluded that statin lipid‐lowering effect is the same for morning and evening administration (Izquiero‐Palomares 2016).

Practitioners can use this evidence to calculate the expected effect of doses of pravastatin commonly utilized in society. It is unlikely that further research will change these estimates appreciably. However, there was a fair amount of heterogeneity in many of the estimates, and we have downgraded outcomes for inconsistency based on I² statistic less than 50%. It is possible that this was due to differences in the populations being studied (e.g. gender or genetic differences; Thompson 2005). To explore this, where it was possible, we compared the effect of pravastatin in men and women plus in patients with familial and non‐familial hypercholesterolemia. A subgroup analysis comparing male versus female participant data was only available for the 40 mg/day dose. The results showed that the efficacy of pravastatin 40 mg in lowering LDL cholesterol was greater in men than in women −32.75 (95% CI −33.69 to −29.81) versus −22.00 (95% CI −30.82 to −13.18; P = 0.03). However, we judged the amount of data available was insufficient to answer whether the lipid‐lowering effect of pravastatin differed in men and women. If anything, it would be anticipated that the effect would be greater in women because on average they weigh less than men. It is important for trial authors to report data separately by sex, and if this had been done in all these trials, we likely would have been able to answer this important question. This subgroup analysis in the atorvastatin, rosuvastatin, cerivastatin and pitavastatin reviews did show a larger effect in women than men (Adams 2014; Adams 2015; Adams 2020a; Adams 2020b). In the fluvastatin review, there was no statistically significant difference of the effect in men and women (Adams 2018).

The subgroup analysis that showed that the effect estimate was greater with Bristol‐Myers Squibb‐funded trials than in non‐Bristol‐Myers Squibb‐funded trials (Analysis 14.1; Analysis 15.1) suggests that our overall estimates of pravastatin's effects could be greater than the true effect. Future independently conducted trials would be required to test this.

Familial versus non‐familial hypercholesterolemia participant data were available for the pravastatin doses 10 mg/day, 20 mg/day and 40 mg/day. These data were analyzed separately for LDL cholesterol‐lowering efficacy in RevMan 5. The percentage reduction in familial hypercholesterolemia patients and non‐familial patients was not different.

A sensitivity analysis revealed that imputing standard deviations does not have an impact on the effect of pravastatin on lipids.

The risk of performance and detection bias for lipid parameters was judged to be low in this review because lipid parameters were measured in remote laboratories and thus unlikely to be influenced by lack of blinding.

We have used data from the Cholesterol Treatment Trialists’ (CTT) publications to estimate the effect of pravastatin to reduce myocardial infarction and compare that with its effect to reduce LDL cholesterol. For the four RCTs where pravastatin was compared to placebo (LIPID 1998; Sacks 1996; Shepherd 1995; Shepherd 2002), a mean pravastatin dose of 40 mg/day reduced LDL cholesterol by 28.5% and reduced myocardial infarction by 23% (95% CI 16 to 29%). This suggests that a relatively small percentage reduction LDL cholesterol can reduce myocardial infarction similarly to more potent statins. These observations are supported by the network analysis done by (Naci 2013a). It certainly suggests that statins could be acting by some other mechanism to reduce myocardial infarction and calls for more head‐to‐head RCTs comparing different statins.

From the main results in the review the lesser ability of pravastatin to reduce LDL‐cholesterol is probably the reason why pravastatin is not as commonly prescribed today. Many guidelines suggest that patients at high or very high risk require at least a 50% reduction in LDL‐C irrespective of their baseline and this is not achievable with pravastatin. However, as described above these guidelines may be proven to be wrong.

Quality of the evidence

GRADE assessments of certainty were determined through consideration of five domains: risk of bias, inconsistency, indirectness, imprecision and publication bias. The summary of all risk of bias domains for the lipid measurements is unclear to low risk of bias (Figure 6). The unclear judgments for randomization and allocation concealment reflected that the method was not reported. We did not feel that this would bias the lipid parameter measurements which were performed in independent laboratories. Therefore, we did not downgrade LDL cholesterol for risk of bias. For the effect of pravastatin on LDL we downgraded the certainty of evidence to moderate due to inconsistency for the 20 mg/day and 40 mg/day doses, because the I² statistic was greater than 50%. There were no concerns regarding indirectness or imprecision and review of funnel plots did not suggest any publication bias as the plots were symmetrical in appearance (Figure 7).

6.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

7.

40 mg pravastatin/day

Low risk of bias is not true for the harm outcome, WDAE. This was reported in 33 (51.6%) of the 64 placebo‐controlled trials. There is therefore a high risk of selective reporting bias for this outcome and this combined with the high risk of other biases means that we cannot be confident that finding no effect on WDAEs is correct (Table 2). We downgraded the certainty of evidence for WDAE to very low due to imprecision (wide confidence interval and crosses the null point) and high risk of bias due to selective reporting.

Potential biases in the review process

A limitation of this review is that many trials did not report standard deviations (SDs) for the lipid‐lowering effects. Where possible, these values were determined from t‐statistics corresponding to the exact P values reported or from the 95% CI of the mean difference (MD) between treatment groups. In trials where the SD was not reported and could not be calculated, the SDs were imputed by the method of (Furukawa 2006), as the average of this parameter from trials that reported it. Such imputation might weight some studies more or less; however, this has been shown by the sensitivity analysis in this review and in other reviews to not have much effect on the estimate of the effect size (Heran 2008; Musini 2014). Another limitation is that few studies were available to demonstrate the lipid‐lowering effect of pravastatin at very low and very high doses.

Agreements and disagreements with other studies or reviews

The best estimate of the mean per cent reduction in blood LDL‐cholesterol for any dose of pravastatin can be calculated from the log dose‐response equation. Using this equation y = −11.3 log(x) −10.4, a pravastatin dose of 20 mg/day reduces cholesterol by an average of 25.1%. This is close to the range of 22.0% to 26.0% reduction in LDL cholesterol from the six comparative trials from the Drug Effectiveness Review Project (DERP) (Smith 2009), and a range of 22.6% to 25.2% reduction in LDL cholesterol in 70 placebo‐controlled trials from Law 2003.

Comparison of the effect of pravastatin with other statins

The equivalent doses from log dose‐response curves were calculated by the method of Seeman 1976. Pravastatin 40 mg is equivalent to about 40 mg of fluvastatin, 2.9 mg atorvastatin and about 0.7 mg of rosuvastatin in LDL cholesterol lowering.

When pravastatin's effect on LDL cholesterol in the recommended dose range is compared to the other statins, pravastatin is similar to fluvastatin and less than atorvastatin, rosuvastatin, cerivastatin and pitavastatin (see below).

• Pravastatin 10 mg to 80 mg (22% to 32%) decrease in LDL cholesterol

• Fluvastatin 20 mg to 80 mg (21 to 33%) decrease in LDL cholesterol

• Cerivastatin 0.1 mg to 0.8 mg (23% to 41%) decrease in LDL cholesterol

• Pitavastatin 1 mg to 4 mg (33% to 44%) decrease in LDL cholesterol

• Atorvastatin 10 mg to 80 mg (37% to 52%) decrease in LDL cholesterol

• Rosuvastatin 5 mg to 40 mg (41% to 55%) decrease in LDL cholesterol

The average increase in HDL for all doses of pravastatin at 4.5% was similar to the other statins: atorvastatin 4.0%, cerivastatin 5.0%, fluvastatin 3.7%, pitavastatin 4.1% and rosuvastatin 7.3%.

Authors' conclusions

Implications for practice.

• Pravastatin causes a linear dose‐response reduction in the per cent change from control of blood low‐density lipoprotein (LDL) cholesterol, total cholesterol and triglycerides. There is no dose‐response relationship for high‐density lipoprotein (HDL) cholesterol which is increased by 4.53% on average for all doses. This effect on HDL is not different from the other studied statins. Pravastatin doses of 5 mg/day to 80 mg/day resulted in a decrease of LDL cholesterol ranging from 18.3% to 31.9%. From the slope of the line for every two‐fold dose increase, there was a 3.4% decrease in blood LDL cholesterol. Pravastatin doses of 5 mg/day to 80 mg/day resulted in a decrease of total cholesterol ranging from 13.65% to 23.3%. From the slope of the line for every two‐fold dose increase, there was a 2.41% decrease in total cholesterol. Pravastatin doses of 5 mg/day to 80 mg/day resulted in a decrease in blood triglycerides ranging from 1.05% to 20.0%. From the slope of the line for every two‐fold dose increase, there was a 4.73% decrease of triglycerides. We judged lipid parameter outcome bias to be low because of the objectivity of measuring them. We judged withdrawal due to adverse effects (WDAE) outcome bias to be high due to high risk of selective outcome reporting.

• By comparison to our other systematic reviews on statins we were able to determine that the slope of pravastatin dose response is less for LDL cholesterol than atorvastatin, rosuvastatin, fluvastatin, cerivastatin and pitavastatin.

• For the recommended doses, pravastatin reduces LDL cholesterol similar to fluvastatin but less than the other statins.

• Harms of pravastatin as compared to placebo assessed by WDAEs were very uncertain.

Implications for research.

• All placebo‐controlled randomized controlled trials (RCTs) should report WDAEs.

• All trials should report the effects separately in men and women, so it is possible to determine if there are any clinically‐significant dose‐related sex differences.

• There is a need for a network meta‐analysis (NMA) looking at all statins in comparison to each other.

History

Protocol first published: Issue 7, 2020

Appendices

Appendix 1. Search strategies

Database: Ovid MEDLINE(R) ALL <1946 to September 21, 2021>
Search Date: 22 September 2021
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
1 pravastatin/ 
2 pravastatin.tw,kf. 
3 (apo‐pravastatin or aplactin or astin or bristacol or cholespar or "cs 514" or "cs‐514" or cs514).tw,kf. 
4 (elisor or epatostantin or eptastatin or eptastatine or kenstatin or lin‐pravastatin or lipemol or liplat or lipidal or liplat or lipostat or liprevil).tw,kf. 
5 (maxudin or mevalotin or minuscol or novales or nu‐pravastatin or prareduct or prascolend or prastan or prasterol).tw,kf. 
6 (prava or pravachol or pravacol or pravalam or pravalich or pravaselect or pravasin or pravasine or pravator or pravyl or "rms 431" or "rms‐431" or rms431).tw,kf. 
7 (sanaprav or selektine or selipran or "sq 31000" or "sq‐31000" or sq31000 or stanidine or vasopran or vasten or xipral).tw,kf. 
8 or/1‐7 
9 randomized controlled trial.pt. 
10 controlled clinical trial.pt.
11 randomi?ed.ab. 
12 placebo.ab.
13 drug therapy.fs. 
14 randomly.ab.
15 trial.ab. 
16 groups.ab. 
17 or/9‐16
18 animals/ not (humans/ and animals/)
19 17 not 18 
20 8 and 19

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

Database: Cochrane Central Register of Controlled Trials (Issue 8, 2021) via Cochrane Register of Studies
Search Date: 22 September 2021
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
#1 MESH DESCRIPTOR Pravastatin AND CENTRAL:TARGET
#2 pravastatin AND CENTRAL:TARGET
#3 (apo‐pravastatin OR aplactin OR astin OR bristacol OR cholespar OR "cs 514" OR "cs‐514" OR cs514) AND CENTRAL:TARGET
#4 (elisor OR epatostantin OR eptastatin OR eptastatine OR kenstatin OR lin‐pravastatin OR lipemol OR liplat OR lipidal OR liplat OR lipostat OR liprevil) AND CENTRAL:TARGET
#5 (maxudin OR mevalotin OR minuscol OR novales OR nu‐pravastatin OR prareduct OR prascolend OR prastan OR prasterol) AND CENTRAL:TARGET
#6 (prava OR pravachol OR pravacol OR pravalam OR pravalich OR pravaselect OR pravasin OR pravasine OR pravator OR pravyl OR "rms 431" OR "rms‐431" OR rms431) AND CENTRAL:TARGET
#7 (sanaprav OR selektine OR selipran OR "sq 31000" OR "sq‐31000" OR sq31000 OR stanidine OR vasopran OR vasten OR xipral) AND CENTRAL:TARGET
#8 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7)

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

Database: Embase <1974 to 2021 September 21>
Search Date: 22 September 2021
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
1 pravastatin/
2 pravastatin.tw. 
3 (apo‐pravastatin or aplactin or astin or bristacol or cholespar or "cs 514" or "cs‐514" or cs514).tw. 
4 (elisor or epatostantin or eptastatin or eptastatine or kenstatin or lin‐pravastatin or lipemol or liplat or lipidal or liplat or lipostat or liprevil).tw. 
5 (maxudin or mevalotin or minuscol or novales or nu‐pravastatin or prareduct or prascolend or prastan or prasterol).tw. 
6 (prava or pravachol or pravacol or pravalam or pravalich or pravaselect or pravasin or pravasine or pravator or pravyl or "rms 431" or "rms‐431" or rms431).tw. 
7 (sanaprav or selektine or selipran or "sq 31000" or "sq‐31000" or sq31000 or stanidine or vasopran or vasten or xipral).tw.
8 or/1‐7 
9 hyperlipidemia/dt
10 hypercholesterolemia/dt 
11 cholesterol*.mp. 
12 (HDL or LDL).mp. 
13 lipoprotein?.mp.
14 lipid*.mp. 
15 triglyceride*.mp.
16 triacylglycerol*.mp.
17 or/9‐16 
18 randomized controlled trial/ 
19 crossover procedure/ 
20 double‐blind procedure/ 
21 (randomi?ed or randomly).tw. 
22 (crossover* or cross‐over*).tw. 
23 placebo.ab.
24 (doubl* adj blind*).tw.
25 assign*.ab. 
26 allocat*.ab. 
27 or/18‐26 
28 (exp animal/ or animal.hw. or nonhuman/) not (exp human/ or human cell/ or (human or humans).ti.) 
29 27 not 28 
30 8 and 17 and 29

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

Database: Web of Science: Indexes=SCI‐EXPANDED, CPCI‐S, ESCI Timespan=All years
Search Date: 22 September 2021
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ #24 1,412 #20 NOT #23
#23 2,206,199 #21 NOT #22
#22 4,058,029 TS=(human*)
#21 2,453,252 TI=(cats OR dog OR dogs OR horses OR lamb OR lambs OR mice OR murine OR mouse OR rabbit* OR rat OR rats OR rodent* OR sheep) 
#20 1,437 #7 AND #19
#19 690,239 (#18 OR #17 OR #16 OR #15 OR #14 OR #13 OR #12 OR #11 OR #10 OR #9 OR #8)
#18 73,545 TS=(placebo* NEAR/2 trial*) 
#17 33,239 TS=(placebo* NEAR/2 stud*) 
#16 31,025 TS=“controlled clinical trial*"
#15 2,896 TS="cross‐over trial*" 
#14 8,406 TS="crossover trial*" 
#13 6,655 TS="cross‐over stud*" 
#12 19,362 TS="crossover stud*" 
#11. 104,825 TS="random* assign*" 
#10. 31,361 TS="random* allocat*" 
#9 176,622 TS="randomi* trial*" 
#8 329,603 TS="randomi* control*" 
#7 9,431 (#6 OR #5 OR #4 OR #3 OR #2 OR #1)
#6 1 TS=(sanaprav OR selektine OR selipran OR "sq 31000" OR "sq‐31000" OR sq31000 OR stanidine OR vasopran OR vasten OR xipral) 
#5 66 TS=(prava OR pravachol OR pravacol OR pravalam OR pravalich OR pravaselect OR pravasin OR pravasine OR pravator OR pravyl OR "rms 431" OR "rms‐431" OR rms431) 
#4 17 TS=(maxudin OR mevalotin OR minuscol OR novales OR nu‐pravastatin OR prareduct OR prascolend OR prastan OR prasterol) 
#3 38 TS=(elisor OR epatostantin OR eptastatin OR eptastatine OR kenstatin OR lin‐pravastatin OR lipemol OR liplat OR lipidal OR liplat OR lipostat OR liprevil) 
#2 265 TS=(apo‐pravastatin OR aplactin OR astin OR bristacol OR cholespar OR "cs 514" OR "cs‐514" OR cs514) 
#1 9,146 TS=pravastatin*

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

Database: Bireme LILACS
Search Date: 22 September 2021
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
((pravastatin OR pravastatina)) OR ((apo‐pravastatin OR aplactin OR astin OR bristacol OR cholespar OR epatostantin OR eptastatin OR eptastatine OR kenstatin OR lin‐pravastatin OR lipemol OR liplat OR lipidal OR liplat OR lipostat OR liprevil OR maxudin OR mevalotin OR minuscol)) OR ((novales OR nu‐pravastatin OR prareduct OR prascolend OR prastan OR prasterol OR prava OR pravachol OR pravacol OR pravalam OR pravalich OR pravaselect OR pravasin OR pravasine OR pravator OR pravyl OR sanaprav OR selektine OR selipran OR stanidine OR vasopran OR vasten OR xipral)) AND ( db:("LILACS") AND type_of_study:("clinical_trials"))

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

Database: ClinicalTrials.gov

Search Date: 22 September 2021

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
Intervention/treatment: Pravastatin
Other terms: randomized
Study type: Interventional Studies

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

Database: WHO International Clinical Trials Registry Platform (ICTRP)

Search Date: 22 September 2021

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
pravastatin AND randomized

OR

Title: lipid* 
AND 
Intervention: pravastatin

Data and analyses

Comparison 1. 5 mg vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 LDL cholesterol	2	88	Mean Difference (IV, Fixed, 95% CI)	‐18.12 [‐23.61, ‐12.62]
1.2 Total cholesterol	2	88	Mean Difference (IV, Fixed, 95% CI)	‐14.08 [‐18.34, ‐9.82]
1.3 HDL cholesterol	2	88	Mean Difference (IV, Fixed, 95% CI)	2.54 [‐2.98, 8.06]
1.4 Triglycerides	2	88	Mean Difference (IV, Fixed, 95% CI)	‐1.96 [‐15.24, 11.33]
1.5 WDAE	1	0	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

1.5. Analysis.

Comparison 1: 5 mg vs placebo, Outcome 5: WDAE

Comparison 2. 10 mg vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 LDL cholesterol	11	765	Mean Difference (IV, Fixed, 95% CI)	‐21.59 [‐23.58, ‐19.60]
2.2 Total cholesterol	9	494	Mean Difference (IV, Fixed, 95% CI)	‐15.00 [‐16.95, ‐13.05]
2.3 HDL cholesterol	10	614	Mean Difference (IV, Fixed, 95% CI)	4.84 [2.52, 7.17]
2.4 Triglycerides	8	474	Mean Difference (IV, Fixed, 95% CI)	‐5.53 [‐11.28, 0.22]
2.5 WDAE	6	419	Risk Ratio (IV, Fixed, 95% CI)	0.41 [0.09, 1.81]

2.5. Analysis.

Comparison 2: 10 mg vs placebo, Outcome 5: WDAE

Comparison 3. 20 mg vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 LDL‐cholesterol	22	3117	Mean Difference (IV, Fixed, 95% CI)	‐25.70 [‐26.77, ‐24.63]
3.2 Total cholesterol	22	3235	Mean Difference (IV, Fixed, 95% CI)	‐19.06 [‐19.91, ‐18.21]
3.3 HDL cholesterol	21	3105	Mean Difference (IV, Fixed, 95% CI)	4.19 [3.05, 5.33]
3.4 Triglycerides	19	2999	Mean Difference (IV, Fixed, 95% CI)	‐11.03 [‐13.44, ‐8.62]
3.5 WDAE	17	3200	Risk Ratio (IV, Fixed, 95% CI)	0.85 [0.60, 1.20]

3.5. Analysis.

Comparison 3: 20 mg vs placebo, Outcome 5: WDAE

Comparison 4. 40 mg vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 LDL cholesterol	35	4869	Mean Difference (IV, Fixed, 95% CI)	‐28.37 [‐29.22, ‐27.52]
4.2 Total cholesterol	33	4717	Mean Difference (IV, Fixed, 95% CI)	‐20.24 [‐20.94, ‐19.55]
4.3 HDL cholesterol	33	3970	Mean Difference (IV, Fixed, 95% CI)	5.13 [4.14, 6.12]
4.4 Triglycerides	28	3238	Mean Difference (IV, Fixed, 95% CI)	‐15.28 [‐17.46, ‐13.10]
4.5 WDAE	12	1875	Risk Ratio (IV, Fixed, 95% CI)	1.01 [0.64, 1.60]

4.5. Analysis.

Comparison 4: 40 mg vs placebo, Outcome 5: WDAE

Comparison 5. 80 mg vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 LDL cholesterol	4	881	Mean Difference (IV, Fixed, 95% CI)	‐34.39 [‐37.00, ‐31.77]
5.2 Total cholesterol	4	883	Mean Difference (IV, Fixed, 95% CI)	‐24.64 [‐26.55, ‐22.73]
5.3 HDL cholesterol	3	573	Mean Difference (IV, Fixed, 95% CI)	2.27 [‐0.15, 4.69]
5.4 Triglycerides	4	883	Mean Difference (IV, Fixed, 95% CI)	‐19.99 [‐24.46, ‐15.53]
5.5 WDAE	2	449	Risk Ratio (IV, Fixed, 95% CI)	0.74 [0.36, 1.51]

5.5. Analysis.

Comparison 5: 80 mg vs placebo, Outcome 5: WDAE

Comparison 6. 160 mg vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 LDL cholesterol	1	434	Mean Difference (IV, Fixed, 95% CI)	‐41.30 [‐51.04, ‐31.56]
6.2 Total cholesterol	1	434	Mean Difference (IV, Fixed, 95% CI)	‐30.70 [‐38.39, ‐23.01]
6.3 HDL cholesterol	1	434	Mean Difference (IV, Fixed, 95% CI)	5.60 [2.43, 8.77]
6.4 Triglycerides	1	434	Mean Difference (IV, Fixed, 95% CI)	‐24.00 [‐31.70, ‐16.30]

Comparison 7. All doses of pravastatin vs placebo for withdrawals due to adverse effects.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 WDAE	33	6054	Risk Ratio (IV, Fixed, 95% CI)	0.81 [0.63, 1.03]

Comparison 8. 40 mg vs placebo male versus female subgroup analysis for LDL cholesterol.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 LDL cholesterol	3	951	Mean Difference (IV, Fixed, 95% CI)	‐31.30 [‐33.20, ‐29.40]
8.1.1 Male	2	906	Mean Difference (IV, Fixed, 95% CI)	‐31.75 [‐33.69, ‐29.81]
8.1.2 Female	1	45	Mean Difference (IV, Fixed, 95% CI)	‐22.00 [‐30.82, ‐13.18]

Comparison 9. 20 mg vs placebo with respect to single versus twice daily dosing subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 LDL cholesterol	21	3037	Mean Difference (IV, Fixed, 95% CI)	‐25.83 [‐26.91, ‐24.74]
9.1.1 Single dose	18	2873	Mean Difference (IV, Fixed, 95% CI)	‐25.70 [‐26.81, ‐24.58]
9.1.2 Twice daily	3	164	Mean Difference (IV, Fixed, 95% CI)	‐27.89 [‐32.35, ‐23.43]

Comparison 10. 40 mg vs placebo with respect to single versus twice daily dosing subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 LDL cholesterol	32	4515	Mean Difference (IV, Fixed, 95% CI)	‐27.79 [‐28.66, ‐26.92]
10.1.1 Single dose	28	4273	Mean Difference (IV, Fixed, 95% CI)	‐27.75 [‐28.64, ‐26.85]
10.1.2 Twice daily	4	242	Mean Difference (IV, Fixed, 95% CI)	‐28.71 [‐32.77, ‐24.66]

Comparison 11. 10 mg vs placebo with respect to familial versus non‐familial hypercholesterolaemia subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
11.1 LDL cholesterol	2	121	Mean Difference (IV, Fixed, 95% CI)	‐22.45 [‐27.19, ‐17.70]
11.1.1 non‐familial	1	86	Mean Difference (IV, Fixed, 95% CI)	‐23.90 [‐30.24, ‐17.56]
11.1.2 familial	1	35	Mean Difference (IV, Fixed, 95% CI)	‐20.60 [‐27.76, ‐13.44]

Comparison 12. 20 mg vs placebo with respect to familial versus non‐familial hypercholesterolaemia subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
12.1 LDL cholesterol	7	245	Mean Difference (IV, Fixed, 95% CI)	‐23.48 [‐27.01, ‐19.95]
12.1.1 non‐familial	5	129	Mean Difference (IV, Fixed, 95% CI)	‐23.00 [‐27.74, ‐18.26]
12.1.2 familial	2	116	Mean Difference (IV, Fixed, 95% CI)	‐24.07 [‐29.36, ‐18.78]

Comparison 13. 40 mg vs placebo with respect to familial versus non‐familial hypercholesterolaemia subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
13.1 LDL cholesterol	5	204	Mean Difference (IV, Fixed, 95% CI)	‐29.41 [‐33.50, ‐25.33]
13.1.1 non‐familial	1	21	Mean Difference (IV, Fixed, 95% CI)	‐31.50 [‐44.35, ‐18.65]
13.1.2 familial	4	183	Mean Difference (IV, Fixed, 95% CI)	‐29.18 [‐33.49, ‐24.87]

Comparison 14. 20 mg vs placebo with respect to Bristol‐Myers Squibb versus Non‐Bristol‐Myers Squibb subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
14.1 LDL cholesterol	13	2358	Mean Difference (IV, Fixed, 95% CI)	‐25.92 [‐27.16, ‐24.68]
14.1.1 Bristol‐Myers Squibb	10	2169	Mean Difference (IV, Fixed, 95% CI)	‐26.37 [‐27.68, ‐25.06]
14.1.2 Non‐Bristol‐Myers Squibb	3	189	Mean Difference (IV, Fixed, 95% CI)	‐21.84 [‐25.78, ‐17.90]

Comparison 15. 40 mg vs placebo with respect to Bristol‐Myers Squibb versus Non‐Bristol‐Myers Squibb subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
15.1 LDL cholesterol	12	2395	Mean Difference (IV, Fixed, 95% CI)	‐29.28 [‐30.48, ‐28.08]
15.1.1 Bristol‐Myers Squibb	7	1987	Mean Difference (IV, Fixed, 95% CI)	‐29.90 [‐31.22, ‐28.57]
15.1.2 Non‐Bristol‐Myers Squibb	5	408	Mean Difference (IV, Fixed, 95% CI)	‐26.44 [‐29.28, ‐23.60]

Comparison 16. all doses of pravastatin vs placebo for HDL cholesterol.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
16.1 HDL cholesterol	56	8784	Mean Difference (IV, Fixed, 95% CI)	4.53 [3.87, 5.18]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Asselbergs 2004.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 433 Age (years): 52.1 Males: 293 Females: 140 LDL cholesterol: 4.1 mmol/L (158.5 mg/dL) Placebo n: 431 Age (years): 50.6 Males: 268 Females: 163 LDL cholesterol: 4.0 mmol/L (154.7 mg/dL) Inclusion criteria: persistent microalbuminuria (a urinary albumin concentration > 10 mg/L in one early morning spot urine sample and a concentration of 15 mg to 300 mg/24 hours in two 24‐hour urine samples at least once), a blood pressure < 160/100 mmHg and no use of antihypertensive medication, and a total cholesterol level < 8.0 mmol/L, or < 5.0 mmol/L in case of previous myocardial infarction, and no use of lipid‐lowering medication Exclusion criteria: creatinine clearance < 60% of the normal age‐adjusted value and use of ACE inhibitors or angiotensin II receptor antagonists Pretreatment: none
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 4 years) Placebo (duration 4 years)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	Sponsorship source: grant E.013 of the Dutch Kidney Foundation, grant NHS 99.103 and NHS 2002‐B202 of the Netherlands Heart Foundation, and grant from Bristol Myers Squibb Country: the Netherlands Setting: hospital
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization was performed in blocks of 20 based on a computer generated randomization list by the pharmacy of Academic Hospital Groningen, Groningen, the Netherlands.
Allocation concealment (selection bias)	Low risk	Allocation concealment was done by the pharmacy of Academic Hospital Groningen, Groningen, the Netherlands.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAEs were not reported.
Other bias	Unclear risk	This study was financially supported by grant E.013 of the Dutch Kidney Foundation, grant NHS 99.103 and NHS 2002‐B202 of the Netherlands Heart Foundation, and an unrestricted grant of Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (432−412)/(432) * 100 = 4.6% were not included in the efficacy analysis. Placebo: (430−405)/(430) * 100 = 5.8% were not included in the efficacy analysis. The rate of to follow‐up is similar in the placebo and pravastatin arms.

Avellone 1994.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 10 Total cholesterol: 284.5 mg/dL (7.36 mmol/L) LDL cholesterol: 187.5 mg/dL (4.84 mmol/L) HDL cholesterol: 40.0 mg/dL (1.03 mmol/L) Triglycerides: 286.5 mg/dL (3.23 mmol/L) Placebo n: 10 Total cholesterol: 297.8 mg/dL (7.7 mmol/L) LDL cholesterol: 203.1 mg/dL (5.25 mmol/L) HDL cholesterol: 41.3 mg/dL (1.07 mmol/L) Triglycerides: 266.8 mg/dL (3.01 mmol/L) Included criteria: patients with type IIB primary hyperlipoproteinemia Excluded criteria: none Pretreatment: none
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 24 weeks) Placebo (duration 24 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous Triglycerides percentage change from baseline at week 4, 8, 12 Outcome type: continuous
Identification
Notes	Sponsorship source: grant (No. 93.00581.PF41) from the National Research Council (CNR, Rome, Italy) Country: Italy Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Low risk	This research (No. 93.00581.PF41) was supported by the National Research Council (CNR, Rome, Italy)‐Targeted Project "Prevention and Control of Disease Factors"; Subproject SP8 "Control of Cardiovascular Pathology."
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Bak 1998.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day + Step 1 diet n: 53 Age (years): 55.3 Total cholesterol: 7.37 mmol/L (285 mg/dL) Pravastatin 20 mg/day + Step 2 diet n: 53 Age (years): 55.6 Total cholesterol: 7.37 mmol/L (285 mg/dL) Placebo + Step 1 diet n: 54 Age (years): 54.6 Total cholesterol: 7.28 mmol/L (281.5 mg/dL) Placebo + Step 2 diet n: 55 Age (years): 54.8 Total cholesterol: 7.2 mmol/L (278.4 mg/dL) Included criteria: men aged 40 years to 70 years. Entry criteria for fasting serum lipids were total cholesterol concentrations of ≥ 6.5 mmol/L and ≤ 8.0 mmol/L and a triacylglycerols value of ≤ 4.0 mmol/L. Excluded criteria: concurrent use of hormones; every use of lipid‐lowering drugs, fish oil preparations, immunosuppressive drugs, antihypertensive agents and diuretics, antianginal drugs and vasodilators, inhibitors of blood clotting; patients who had impaired hepatic or renal function, an unstable medical condition, a history of alcohol or drug abuse, or excessive obesity (Quetelet index > 30) Pretreatment: The data indicate that the groups were well balanced in all characteristics.
Interventions	Intervention characteristics Pravastatin 20 mg/day + Step 1 diet (duration 6 months) Pravastatin 20 mg/day + Step 2 diet (duration 6 months) Placebo + Step 1 diet (duration 6 months) Placebo + Step 2 diet (duration 6 months)
Outcomes	Total cholesterol percentage change from baseline at month 2 Outcome type: continuous WDAEs up to month 6 Outcome type: dichotomous
Identification
Notes	The lipid data and the WDAE data was combined for both step 1 diet and step 2 diet. Sponsorship source: Bristol Myers Squibb Country: the Netherlands Setting: Participants were recruited from a population‐based cholesterol screening program.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computer‐generated random number to each participant
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	This study was double‐blinded. Lipid parameter measurements were unlikely to be influenced by methods of blinding.
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	High risk	LDL cholesterol was measured at 6 months only, not at 2 months.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	Financial support was obtained by Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	High risk	Placebo Step 1: (54−51) * 100/54 = 5.6% were not included in the efficacy analysis. Placebo Step 2: (55−51) * 100/55 = 7.2% were not included in the efficacy analysis. Pravastatin Step 1: (53−49) * 100/53 = 7.5% were not included in the efficacy analysis. Pravastatin Step 2: (53−46) * 100/53 = 13.2% were not included in the efficacy analysis. Rate of loss to follow‐up is different between the pravastatin step 2 arm and all the other arms of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding with respect to WDAE was not reported.

Barbi 1992.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 8 Total cholesterol: 302 mg/dL (7.81 mmol/L) LDL cholesterol: 231 mg/dL (5.97 mmol/L) HDL cholesterol: 40 mg/dL (1.03 mmol/L) Triglycerides: 156 mg/dL (1.76 mmol/L) Pravastatin 80 mg/day n: 9 Total cholesterol: 297 mg/dL (7.68 mmol/L) LDL cholesterol: 224 mg/dL (5.79 mmol/L) HDL cholesterol: 44 mg/dL (1.14 mmol/L) Triglycerides: 150 mg/dL (1.69 mmol/L) Placebo n: 7 Total cholesterol: 304 mg/dL (7.86 mmol/L) LDL cholesterol: 223 mg/dL (5.77 mmol/L) HDL cholesterol: 48 mg/dL (1.24 mmol/L) Triglycerides: 166 mg/dL (1.87 mmol/L) Included criteria: Patients, age 21 years to 70 years, who had an LDL cholesterol value of greater than the 90th percentile (Report of the National Cholesterol Education Program, 1988) and ≥ 160 mg/dL were considered for this trial. Additionally, triglyceride levels had to be < 250 mg/dL. Excluded criteria: type I, III, IV, or V hyperlipoproteinemia; untreated thyroid or other endocrine diseases; renal or hepatobiliary disease; chronic pancreatitis; collagen‐vascular diseases; myocardial infarction within the past 6 months or clinically significant heart failure; uncontrolled hypertension; history of cerebral vascular accident; serious gastrointestinal disease; obesity greater than 40% above ideal body weight; excessive alcohol consumption; treatment with cortical steroids, estrogens, androgens, lipid‐lowering agents, coumarin anticoagulants, theophylline, barbiturates, or aluminum‐containing antacids; and any other condition judged to impair the patient’s ability to complete the trial Pretreatment: There were no difference at baseline within the parallel treatment groups.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 24 weeks) Pravastatin 80 mg/day (duration 24 weeks) Placebo (duration 24 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: protocol No. 27201‐10 from Bristol Myers Squibb Co., Princeton, New Jersey, and in US Public Health Service grant, HL‐46967 Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Unclear risk	Total cholesterol, triglycerides, and HDL and LDL cholesterol were analyzed by standardized techniques at a centralized laboratory (Scicor, 8200 Scicor Drive, Indianapolis, Indiana 46234) or in a laboratory belonging to one of the authors (Corder).
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome was reported.
Other bias	Unclear risk	This work was supported in part by funds in protocol No. 27201‐10 from Bristol Myers Squibb Co., Princeton, New Jersey, and in US Public Health Service grant, HL‐46967.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Barrow 1991.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 8 Males: 5 Females: 3 LDL cholesterol: 4.41 mmol/L (170.5 mg/dL) Placebo n: 8 Males: 5 Females: 3 LDL cholesterol: 4.24 mmol/L (163.5 mg/dL) Included criteria: 16 participants (10 male and 6 female aged 45 years to 67 years). Each participant had been diagnosed as having mild hypercholesterolemia on the basis of two or more measurements of fasting plasma cholesterol in the range 5.2 mmol/L to 7.8 mmol/L. In addition, each participant had two or more of the following coronary risk factors – male gender, cigarette smoking (≥ 10 cigarettes per day), treated hypertension, personal history of CHD diagnosed angiographically or after myocardial infarction, or a family history of such confirmed coronary disease in a first‐degree relative aged < 55 years. Excluded criteria: any acute condition requiring hospitalization within the previous 3 months, a history of drug or alcohol abuse, unstable angina pectoris, poorly controlled hypertension, heart failure, secondary or homozygous familial hypercholesterolemia Pretreatment: Mean plasma concentrations of triglycerides were similar in both groups at the start of the study.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 52 weeks) Placebo (duration 52 weeks)
Outcomes	LDL cholesterol percentage change from baseline at week 13 Outcome type: continuous
Identification
Notes	13‐week lipid data Sponsorship source: Bristol Myers Squibb Pharmaceuticals Country: UK Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Balanced block design coded by the biometrics group at Bristol Myers Squibb
Allocation concealment (selection bias)	Low risk	Randomization was done by Bristol Myers Squibb.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAEs reported
Other bias	High risk	This work was supported by a grant from Bristol Myers Squibb Pharmaceuticals.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Bayerle Eder 2002.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 8 Total cholesterol: 256.8 mg/dL (6.64 mmol/L) Placebo n: 8 Total cholesterol: 267.2 mg/dL (6.91 mmol/L) Included criteria: Twelve postmenopausal female and four male non‐smoking participants (age ± SD: 54.9 years ± 6.5 years) with moderate hypercholesterolemia, defined as fasting total plasma cholesterol levels between 240 mg/dL and 350 mg/dL and triglycerides 250 mg/dL, were enrolled, who had already been put on a low cholesterol diet normal ophthalmic findings and a refractive error of < 3 diopters. Participants did not take any non‐trial medication, including over‐the‐counter drugs, throughout the study and were asked to abstain from alcohol and beverages containing xanthine derivatives for 12 hours before drug administration. Excluded criteria: regular use of any medication including fish oil preparations or vitamin supplements, age < 25 years or > 65 years, body mass index < 15 or > 85 percentiles, known hypersensitivity to HMG‐CoA reductase inhibitors, impaired hepatic or renal function, diabetes, systemic hypertension, CHD, congestive heart failure, overt vascular complications and homozygous hypercholesterolemia or hyperlipidemia I/III Pretreatment: No difference was detectable between groups before drug therapy.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 1 month) Placebo (duration 1 month)
Outcomes	Total cholesterol percentage change from baseline at month 1 Outcome type: continuous WDAEs up to month 1 Outcome type: dichotomous
Identification
Notes	Sponsorship source: a medical school grant from Bristol Myers Squibb, Austria Country: Austria Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	High risk	LDL cholesterol outcome was not reported for the placebo group.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	The study was kindly supported by a medical school grant from Bristol Myers Squibb, Austria.
Incomplete outcome data (attrition bias) All outcomes	High risk	Placebo group: all participants were included in the efficacy analysis. Pravastatin group: (8−7) * 100/8 = 12.5% were not included in the efficacy analysis. Rate of loss to follow‐up are different in the placebo and pravastatin arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding with respect to WDAE was not reported.

Behounek 1993.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 530 Age (years): 55 Males: 410 Females: 119 Gender data not available: 1 Total cholesterol: 6.74 mmol/L (260.6 mg/dL) Placebo n: 532 Age (years): 55 Males: 403 Females: 128 Gender data not available: 1 Total cholesterol: 6.77 mmol/L (261.8 mg/dL) Overall N: 1062 Age (years): 55 Males: 813 Females: 247 Gender data not available: 2 Total cholesterol: 6.755 mmol/L (261.2 mg/dL) Included criteria: At entry, patients were to have a serum total cholesterol concentration between 5.2 mmol/L and 7.8 mmol/L (200 mg/dL and 300 mg/dL) despite dietary intervention as well as ≥ 2 additional risk factors for CAD as defined in the protocol (male gender, systemic hypertension, cigarette smoking, family history of CAD, prior myocardial infarction, or history of angina pectoris). Excluded criteria: a history of homozygous familial hypercholesterolemia; types I, III, IV or V hyperlipoproteinemia; significant metabolic, renal, hepatic or endocrine disease: history of drug abuse; or obesity (> 40% above ideal body weight). Patients were also excluded if, within 8 weeks of randomization, they were taking corticosteroids, immunosuppressive drugs, investigational drugs, androgens, or estrogens and progestins for indications other than replacement therapy. Pretreatment: Baseline characteristics and the number of CAD risk factors were generally similar in the two groups. Most of the participants had either two or three additional CAD risk factors – 436 (82%) in the pravastatin group and 441 (83%) in the placebo group. Four or five additional risk factors were present in 86 participants (16%) in each group. Contrary to protocol specifications, six participants were taking pravastatin (1.1%) and four taking placebo (0.8%) with only one additional CAD risk factor were included in the study. The number of risk factors was unknown for two participants (0.4%) in the pravastatin group and one participant (0.2%) in the placebo group. The treatment groups were comparable in the prevalence of CAD risk factors except for history of angina, which was more prevalent in the placebo group (P < 0.05).
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 26 weeks) Bedtime dosing Placebo (duration 26 weeks) Bedtime dosing
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous WDAEs up to week 26 Outcome type: dichotomous
Identification
Notes	Sponsorship source: This study was supported by a grant from Bristol Myers Squibb Company, Princeton, New Jersey. Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Unclear risk	Lipid parameters were measured in local laboratories and a central laboratory Bruce D Behounek is from Bristol Myers Squibb Company (conflict of interest).
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	Bristol Myers Squibb Pharmaceutical Research Institute funded the trial.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Bertrand 1997.

Study characteristics
Methods	Study design: randomized, multicenter, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 341 Total cholesterol: 228 mg/dL (5.9 mmol/L) LDL cholesterol: 155 mg/dL (4.0 mmol/L) HDL cholesterol: 47 mg/dL (1.22 mmol/L) Triglycerides: 140 mg/dL (1.58 mmol/L) Placebo n: 347 Total cholesterol: 231 mg/dL (5.97 mmol/L) LDL cholesterol: 157 mg/dL (4.06 mmol/L) HDL cholesterol: 47 mg/dL (1.22 mmol/L) Triglycerides: 139 mg/dL (1.57 mmol/L) Included criteria: Patients 25 years to 75 years old with a left ventricular ejection fraction, assessed by angiography, that exceeded 40% were eligible for inclusion. In addition, all participants were required to have total cholesterol levels between 200 mg/dL and 310 mg/dL and triglyceride levels < 500 mg/dL. Excluded criteria: Patients with a recent myocardial infarction (within 15 days) and patients who had previously undergone PTCA or CABG of the target vessel received drugs not allowed by protocol (fish oil or other lipid‐lowering agents within 1 month of the procedure), corticosteroids or immuno‐suppressive drugs. Pretreatment: The two groups were well matched in terms of baseline clinical characteristics.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 6 months) Placebo (duration 6 months)
Outcomes	Total cholesterol percentage change from baseline at month 2 Outcome type: continuous LDL cholesterol percentage change from baseline at month 2 Outcome type: continuous HDL cholesterol percentage change from baseline at month 2 Outcome type: continuous Triglycerides percentage change from baseline at month 2 Outcome type: continuous
Identification
Notes	Sponsorship source: This study was supported by a grant from Bristol Myers Squibb Laboratories, Paris. Country: France Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAEs were reported for both pravastatin and placebo combined cannot compare WDAEs in pravastatin versus placebo group. During subsequent follow‐up, five participants had adverse events and stopped the treatment.
Other bias	High risk	This study was supported by a grant from Bristol Myers Squibb Laboratories, Paris.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Pravastatin: (347−312) * 100/347 = 10.1% were not included in the efficacy analysis from Table 1. Placebo: (348−311) * 100/348 = 10.6% were not included in the efficacy analysis from Table 1. Absolute rates of loss to follow‐up were moderate in the placebo and pravastatin arms.

Betteridge 1992.

Study characteristics
Methods	Study design: randomized double‐blind, double‐dummy, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day twice‐daily = 40 mg/day n: 43 Males: 31 Females: 12 Total cholesterol: 9.87 mmol/L (382 mg/dL) LDL cholesterol: 7.82 mmol/L (302 mg/dL) HDL cholesterol: 1.17 mmol/L (45.2 mg/dL) Triglycerides: 1.52 mmol/L (135 mg/dL) Placebo n: 43 Males: 31 Females: 12 Total cholesterol: 9.48 mmol/L (367 mg/dL) LDL cholesterol: 7.47 mmol/L (289 mg/dL) HDL cholesterol: 1.13 mmol/L (43.7 mg/dL) Triglycerides: 1.55 mmol/L (137 mg/dL) Included criteria: 128 participants with heterozygous familial hypercholesterolemia. Familial hypercholesterolemia was defined as a total plasma cholesterol concentration ≥ 7.5 mmol/L, LDL cholesterol concentration ≥ 5.0 mmol/L, and tendon xanthomas present in the patient or a first‐degree relative. Patients were also entered into the study if, in addition to raised total plasma and LDL cholesterol concentrations, they had a parent who had died prematurely of myocardial infarction – that is, a father who had died below the age of 50 years or a mother who had died below the age of 60 years. Participants were men and postmenopausal or surgically sterile women aged 18 years to 70 years. Excluded criteria: primary or secondary causes of hyperlipidemia and no significant renal, hepatic, or endocrine (except stable thyroid replacement) disease; patients with hypertension (SBP ≥ 160 mmHg or DBP ≥ 100 mmHg), poorly controlled cardiac failure, obesity (≥ 130% of ideal body weight), recent myocardial infarction (within 3 months), unstable angina, and excess alcohol intake (> 5 units/day); patients with medical conditions (other than CHD) likely to limit life span to < 5 years Pretreatment: Lipid and lipoprotein concentrations were similar.
Interventions	Intervention characteristics Pravastatin 20 mg/day, twice‐daily = 40 mg/day (duration 36 weeks) Placebo (duration 36 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous WDAEs up to week 12 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: UK Setting: lipid clinic
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported separately for each arm.
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment was not reported for WDAEs.

Bhatnagar 1995.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 16 Total cholesterol: 7.5 mmol/L (290 mg/dL) LDL cholesterol: 4.2 mmol/L (162 mg/dL) HDL cholesterol: 1.16 mmol/L (44.9 mg/dL) Age (years): 56.9 Placebo n: 15 Total cholesterol: 7.5 mmol/L (290 mg/dL) LDL cholesterol: 4.4 mmol/L (170 mg/dL) HDL cholesterol: 1.27 mmol/L (49.1 mg/dL) Age (years): 57.6 Included criteria: Patients with non‐insulin‐­dependent diabetes mellitus receiving treatment with diet and sulfonylureas (glibenclamide or gliclazide) for at least 6 months and aged between 18 years and 60 years were recruited from outpatient clinics in the Diabetes Centre and the Lipid Clinic at Manchester Royal Infirmary. The study, which had the approval of the hospital's ethical committee, was restricted to patients with a serum total cholesterol > 6.5 mmol/L on two separate occasions and who had adequate glycemic control (HbA1c (glycated hemoglobin) < 11.5%) within 4 weeks of entering the study normal serum creatine kinase activity and a stable body mass index. Excluded criteria: proteinuria and levels of serum creatinine > 120 μmol/L and serum triglyceride > 10 mmol/L, patients taking lipid‐lowering medication, those with a history of myocardial infarction, or those who had cardiac surgery within the last 3 months Pretreatment: The two pre‐randomization baseline serum lipid values (before or after entry to the 4‐week placebo run‐in phase) were not significantly different in either treatment group.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	SDs can be calculated from P values. Sponsorship source: a grant from Bristol Myers Squibb, Hounslow, U.K. M.I.M. is supported by a Regional Infrastructure Grant to P.N.D. Country: UK Setting: outpatient clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	This study was partially supported by a grant from Bristol Myers Squibb, Hounslow, U.K. M.I.M. is supported by a Regional Infrastructure Grant to P.N.D.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	(35−31) * 100/35 = 11.4% were not included in the efficacy analysis. A total of 35 participants entered the study, but four participants were withdrawn from the study.

Blake 2003.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 57.1 Males: 151 Females: 105 N: 256 Total cholesterol: 231.8 mg/dL (5.99 mmol/L) LDL cholesterol: 143.7 mg/dL (3.72 mmol/L) HDL cholesterol: 40.2 mg/dL (1.04 mmol/L) Placebo Age: 57.3 Males: 147 Females: 97 N: 244 Total cholesterol: 230.6 mg/dL (5.96 mmol/L) LDL cholesterol: 142.5 mg/dL (3.69 mmol/L) HDL cholesterol: 39.5 mg/dL (1.02 mmol/L) Included criteria: men and women aged 21 years or older, who were free of statin use during the 6‐month period prior to enrollment and had no contraindications to statin therapy. Participants had known baseline LDL cholesterol levels of ≥ 130 mg/dL and no prior history of cardiovascular disease. Excluded criteria: prior history of cardiovascular disease Pretreatment: The proportion of current smokers was higher among those randomized to placebo (22.6%) than among those randomized to pravastatin (11.9%).
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 24 weeks) Placebo (duration 24 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Country: USA Setting: multicenter, community‐based trial
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Judgment comment: Components of this work were supported by grants from the National Heart, Lung, and Blood Institute, the Doris Duke Charitable Foundation (New York, New York), the Fondation Leducq (Paris, France), the Donald W. Reynolds Foundation (Las Vegas, Nevada), and Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Blann 2001.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age (years): 65 n: 17 Males: 10 Females: 7 Total cholesterol: 6.5 mmol/L (251 mg/dL) LDL cholesterol: 4.3 mmol/L (166 mg/dL) HDL cholesterol: 1.2 mmol/L (46.4 mg/dL) Placebo Age (years): 61 n: 15 Males: 9 Females: 6 Total cholesterol: 6.1 mmol/L (236 mg/dL) LDL cholesterol: 4.0 mmol/L (155 mg/dL) HDL cholesterol: 1.1 mmol/L (42.5 mg/dL) Included criteria: peripheral artery disease proved by Doppler/ultrasound stenosis of the iliac or femoral arteries or both or stenosis of a carotid artery and total cholesterol between 5.5 mmol/L and 7.5 mmol/L (210 mg/mL to 290 mg/mL) Excluded criteria: diabetes, impaired hepatic or renal function, connective tissue disease, current use of lipid‐lowering or anticoagulant drugs, hormone replacement therapy, venous ulceration, neoplastic disease, or inability to give written informed consent Pretreatment: not reported
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 4 months) Placebo (duration 4 months)
Outcomes	Total cholesterol percentage change from baseline at month 2 Outcome type: continuous LDL cholesterol percentage change from baseline at month 2 Outcome type: continuous HDL cholesterol percentage change from baseline at month 2 Outcome type: continuous
Identification
Notes	Sponsorship source: Bristol Myers Squibb, London, UK Country: UK Setting: outpatients
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	High risk	This study was supported by a grant from Bristol Myers Squibb, London, UK.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Celis 1994.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day Age: 58 n: 25 Males: 11 Females: 14 Total cholesterol: 277 mg/dL (7.16 mmol/L) LDL cholesterol: 197 mg/dL (5.09 mmol/L) HDL cholesterol: 48 mg/dL (1.24 mmol/L) Triglycerides: 166 mg/dL (1.87 mmol/L) Placebo Age: 57 n: 25 Males: 11 Females: 14 Total cholesterol: 285 mg/dL (7.37 mmol/L) LDL cholesterol: 200 mg/dL (5.17 mmol/L) HDL cholesterol: 51 mg/dL (1.32 mmol/L) Triglycerides: 169 mg/dL (1.91 mmol/L) Included criteria: male or postmenopausal female hypertensive patients, aged 18 years to 70 years, with hypercholesterolemia if they were otherwise in good health and if their blood pressure was controlled (DBP < 100 mmHg) total plasma cholesterol from 250 mg/dL through 400 mg/dL after having been on a standard lipid‐lowering diet and without lipid‐lowering agents for at least 1 month Excluded criteria: none reported Pretreatment: At randomization, the two treatment groups were similar in their demographic, clinical and biochemical characteristics.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 6 months) Placebo (duration 6 months)
Outcomes	Total cholesterol percentage change from baseline at month 1 Outcome type: continuous LDL cholesterol percentage change from baseline at month 1 Outcome type: continuous HDL cholesterol percentage change from baseline at month 1 Outcome type: continuous Triglycerides percentage change from baseline at month 1 Outcome type: continuous WDAEs up to month 1 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: Belgium Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported for each arm.
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (25−24) * 100/25 = 4% participants were not included in the efficacy analysis. Placebo: (25−24) * 100/25 = 4% participants were not included in the efficacy analysis. Absolute rates of loss to follow‐up are the same and low in the placebo and pravastatin arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment was not reported for WDAEs.

Chan 1995.

Study characteristics
Methods	Study design: randomized double‐blind placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day Age: 74 n: 30 Males: 14 Females: 16 Total cholesterol: 279 mg/dL (7.21 mmol/L) LDL cholesterol: 200 mg/dL (5.17 mmol/L) HDL cholesterol: 47 mg/dL (1.22 mmol/L) Triglycerides: 127 mg/dL (1.43 mmol/L) Placebo Age: 76 n: 30 Males: 12 Females: 18 Total cholesterol: 285 mg/dL (7.37 mmol/L) LDL cholesterol: 210 mg/dL (5.43 mmol/L) HDL cholesterol: 46 mg/dL (1.19 mmol/L) Triglycerides: 129 mg/dL (1.46 mmol/L) Included criteria: Elderly hypertensive participants over age 65 and with hypercholesterolemia were enrolled from outpatient hypertensive clinics if they were apparently healthy, they showed no cardiovascular risk factors other than hypertension and hypercholesterolemia, and their blood pressure was well controlled (SBP < 160 mmHg, DBP < 95 mmHg). Patients were selected if their plasma total cholesterol level ranged between 250 mg/dL and 400 mg/dL after having been on a standard lipid‐lowering diet (American Heart Association Step 1 Diet), and without hypolipemic agents for at least 1 month. Excluded criteria: not reported Pretreatment: At the beginning of randomization, both placebo and pravastatin groups were similar in demographic, clinical, and biochemical characteristics.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 6 months) Placebo (duration 6 months)
Outcomes	Total cholesterol percentage change from baseline at month 1, 2, 3 Outcome type: continuous LDL cholesterol percentage change from baseline at month 1, 2, 3 Outcome type: continuous HDL cholesterol percentage change from baseline at month 1, 2, 3 Outcome type: continuous Triglycerides percentage change from baseline at month 1, 2, 3 Outcome type: continuous WDAEs up to month 6 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: Taiwan Setting: outpatient hypertensive clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported for each trial arm separately.
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Cheung 1993.

Study characteristics
Methods	Study design: randomized, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day Age: 53 n: 16 Males: 11 Females: 5 Total cholesterol: 265 mg/dL (6.85 mmol/L) LDL cholesterol: 185 mg/dL (4.78 mmol/L) HDL cholesterol: 50 mg/dL (1.29 mmol/L) Triglycerides: 132 mg/dL (1.49 mmol/L) Placebo Age: 60 n: 8 Males: 4 Females: 4 Total cholesterol: 279 mg/dL (7.21 mmol/L) LDL cholesterol: 195 mg/dL (5.04 mmol/L) HDL cholesterol: 50 mg/dL (1.29 mmol/L) Triglycerides: 163 mg/dL (1.84 mmol/L) Included criteria: Participants were between ages 18 years and 70 years and had a mean of two qualifying LDL cholesterol levels within 10% of each other in excess of 160 mg/dL and plasma triglyceride concentrations < 350 mg/dL. Participants were required to have discontinued lipid‐lowering medication 8 weeks to 14 weeks prior to randomization depending on the type of medication. Excluded criteria: any condition that might influence lipoprotein lipid levels such as thiazide or sex hormone use or hypothyroidism Pretreatment: No significant differences are seen in mean age, body mass index, and baseline randomization qualifying lipoprotein lipid measurements. With one exception, all participants had an LDL cholesterol value at randomization above the qualifying level of 160 mg/dL, and only two participants had triglyceride levels above the age‐ and sex‐specific 90th percentile of the Lipid Research Clinics Prevalence Study population.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: This project was supported by NIH grants HL‐30086 (M.C.C., A.C.W.), DK‐35816 (R.H.K.), and an NIH First Independent Research Support and Transition Award HL‐38760 (M.A.A.) and by a grant from the Bristol Myers Squibb US Pharmaceutical Company. Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	This project was supported by NIH grants HL‐30086 (M.C.C., A.C.W.), DK‐35816 (R.H.K.), and an NIH First Independent Research Support and Transition Award HL‐38760 (M.A.A.) and by a grant from the Bristol Myers Squibb US Pharmaceutical Company.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Cipollone 2002.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 8 Total cholesterol: 272 mg/dL (7.03 mmol/L) LDL cholesterol: 189 mg/dL (4.89 mmol/L) HDL cholesterol: 54 mg/dL (1.40 mmol/L) Triglycerides: 112 mg/dL (1.26 mmol/L) Placebo n: 8 Total cholesterol: 256 mg/dL (6.62 mmol/L) LDL cholesterol: 171 mg/dL (4.42 mmol/L) HDL cholesterol: 57 mg/dL (1.47 mmol/L) Triglycerides: 165 mg/dL (1.86 mmol/L) Included criteria: evidence of hypercholesterolemia Excluded criteria: none reported Pretreatment: not reported
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: not reported Country: Italy Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Contacos 1993.

Study characteristics
Methods	Study design: randomized single‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age (years): 52 n: 10 Males: 8 Females: 2 Total cholesterol: 7.8 mmol/L (302 mg/dL) LDL cholesterol: 4.4 mmol/L (170 mg/dL) HDL cholesterol: 1.05 mmol/L (40.6 mg/dL) Triglycerides: 5.5 mmol/L (487 mg/dL) Placebo Age (years): 64 n: 11 Males: 4 Females: 7 Total cholesterol: 7.6 mmol/L (294 mg/dL) LDL cholesterol: 4.9 mmol/L (189 mg/dL) HDL cholesterol: 1.13 mmol/L (43.7 mg/dL) Triglycerides: 4.8 mmol/L (425 mg/dL) Included criteria: Participants with fasting plasma lipid levels of TC (total cholesterol) and TGs > 5.5 mmol/L and > 2.8 mmol/L aged 32 years to 70 years all had a history of primary mixed (type IIB) hyperlipidemia. Excluded criteria: participants with apoE2/E2 phenotype, familial hypercholesterolemia, diabetes mellitus, and secondary forms of hyperlipidemia due to liver, renal, or thyroid dysfunctions and women capable of conceiving Pretreatment: The placebo group was older and had more female participants than the pravastatin group.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 18 weeks) Placebo (duration 18 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous
Identification
Notes	Sponsorship source: the National Heart Foundation of Australia and Bristol Myers Squibb Pharmaceuticals Pty Ltd Country: Australia Setting: outpatients
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE was reported in the paper.
Other bias	Unclear risk	Supported in part by the National Heart Foundation of Australia and Bristol Myers Squibb Pharmaceuticals Pty Ltd
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Davidson 1991.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age (years): 52.3 n: 26 Male: 14 Female: 12 Total cholesterol: 7.52 mmol/L (291 mg/dL) LDL cholesterol: 5.43 mmol/L (210 mg/dL) HDL cholesterol: 1.10 mmol/L (42.5 mg/dL) Triglycerides: 1.88 mmol/L (167 mg/dL) Placebo Age (years): 54 n: 26 Male: 15 Female: 11 Total cholesterol: 7.28 mmol/L (282 mg/dL) LDL cholesterol: 5.13 mmol/L (198 mg/dL) HDL cholesterol: 1.26 mmol/L (48.7 mg/dL) Triglycerides: 1.65 mmol/L (146 mg/dL) Included criteria: men and postmenopausal or surgically sterile women 21 years through 70 years of age. The participants had established diagnoses of primary hypercholesterolemia (LDL cholesterol concentrations ≥ 3.88 mM or in ≥ 75th percentile for age and sex in the US population despite a stable diet and a mean triglyceride value ≤ 3.95 mM). Excluded criteria: homozygous familial hypercholesterolemia; type I, III, IV, V hyperlipoproteinemias; or significant metabolic renal, hepatic, endocrine, or cardiovascular disease; patients taking corticosteroids, estrogens for indications other than replacement therapy, androgens, or investigational drugs Pretreatment: The treatment groups' demographics and plasma lipid levels were similar.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 16 weeks) Bedtime dosing Placebo (duration 16 weeks) Bedtime dosing
Outcomes	Total cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous Triglycerides percentage change from baseline at week 4, 8, 12 Outcome type: continuous WDAE up to week 16 Outcome type: dichotomous
Identification
Notes	Sponsorship source: This study was supported by a grant from the Bristol Myers Squibb Co. and N.I.H. Grant #RR00046. Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported for pravastatin versus placebo separately.
Other bias	Unclear risk	This study was funded by a grant from Bristol Myers Squibb Co. and an NIH grant.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: all participants were included in the efficacy analysis. Placebo: (27−26) * 100/27 = 3.7% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low overall.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding with respect to WDAE was not reported.

Davignon 1994.

Study characteristics
Methods	Study design: randomized, single‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 47 n: 39 Male: 29 Female: 10 Total cholesterol: 307.1 mg/dL (7.94 mmol/L) LDL cholesterol: 227.1 mg/dL (5.87 mmol/L) HDL cholesterol: 47.9 mg/dL (1.24 mmol/L) Triglycerides: 160.9 mg/dL (1.82 mmol/L) Placebo Age: 51.7 n: 40 Male: 27 Female: 13 Total cholesterol: 314.8 mg/dL (8.14 mmol/L) LDL cholesterol: 232.3 mg/dL (6.01 mmol/L) HDL cholesterol: 45.9 mg/dL (1.19 mmol/L) Triglycerides: 182.7 mg/dL (2.06 mmol/L) Included criteria: 168 men and postmenopausal or surgically sterile women (aged 21 years to 70 years) were enrolled in the study with an established diagnosis of type IIa or IIb primary hypercholesterolemiaLDL cholesterol ≥ 150 mg/dL and triglycerides ≤ 350 mg/dL. Excluded criteria: homozygous familial hypercholesterolemia (total cholesterol > 600 mg/dL, presence of xanthomas, and homozygous familial hypercholesterolemia in all first‐degree relatives); type I, III, IV and V hyperlipoproteinemias; or significant metabolic, renal, hepatic, endocrine, or cardiovascular disease treatment with other lipid‐lowering agents, corticosteroids, estrogens, androgens, quinidine, coumarin derivatives, theophylline, barbiturates, antacids or fish oil preparations Pretreatment: There was no statistically significant difference among treatment groups in demographic variables.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 96 weeks) Placebo (duration 96 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs at week 8 Outcome type: dichotomous
Identification
Notes	Sponsorship source: grant from Bristol Myers Squibb Canada Country: Canada Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAEs were not reported separately for each arm of the trial.
Other bias	High risk	This work was supported by a grant from Bristol Myers Squibb Canada, Saint‐Laurent, Quebec, Canada.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

DenHartog 2001.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 64.2 n: 50 Male: 35 Female: 15 Total cholesterol: 6.58 mmol/L (254.4 mg/dL) LDL cholesterol: 4.525 mmol/L (175 mg/dL) Placebo Age: 62.7 n: 49 Male: 37 Female: 12 Total cholesterol: 6.565 mmol/L (253.75 mg/dL) LDL cholesterol: 4.525 mmol/L (175 mg/dL) Included criteria: 99 participants with myocardial infarction or UA (unstable angina) who were hospitalized for less than 48 hours; men 18 years to 80 years or postmenopausal women with serum total cholesterol levels of > 5.2 mmol/L and DL cholesterol levels > 3.5 mmol/L Excluded criteria: history of hypersensitivity to statins or formulation components, severe congestive heart failure or cardiomyopathy, significant liver disease, significant gastrointestinal disease or abdominal surgery that might adversely influence drug absorption, substance or alcohol abuse, history or present use of any other lipid‐lowering or investigational agent, uncontrolled diabetes, thyroid disease, severe renal impairment; dysproteinemia, and primary muscle disease Pretreatment: Both arms of the study were well balanced with respect to most baseline characteristics.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 3 months) Placebo (duration 3 months)
Outcomes	Total cholesterol percentage change from baseline at month 1, 3 Outcome type: continuous LDL cholesterol percentage change from baseline at month 1, 3 Outcome type: continuous WDAEs up to month 3 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: the Netherlands Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Pre‐generated randomization scheme
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding matching placebo
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	Some of the authors are from Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	High risk	22% of participants discontinued prematurely
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment of WDAEs was not reported.

Dupuis 1999.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 55 n: 28 Male: 26 Female: 2 Total cholesterol: 6.39 mmol/L (247 mg/dL) LDL cholesterol: 4.15 mmol/L (160 mg/dL) HDL cholesterol: 1.03 mmol/L (39.8 mg/dL) Triglycerides: 2.19 mmol/L (194 mg/dL) Placebo Age: 56 n: 27 Male: 22 Female: 5 Total cholesterol: 6.41 mmol/L (248 mg/dL) LDL cholesterol: 4.34 mmol/L (168 mg/dL) HDL cholesterol: 1.12 mmol/L (43.3 mg/dL) Triglycerides: 2.19 mmol/L (194 mg/dL) Included criteria: Patients admitted to the coronary care unit of the Montreal Heart Institute with a diagnosis of acute myocardial infarction or unstable angina were eligible if they had admission total serum cholesterol ≥ 5.2 mmol/L or LDL cholesterol ≥ 3.4 mmol/L and serum triglycerides ≥ 4.5 mmol/L. Excluded criteria: presence of heart failure with an ejection fraction of < 40%, administration of lipid‐lowering agents in the preceding 8 weeks, renal failure with serum creatinine level > 200 mmol/L, patients requiring coronary artery bypass surgery, and premenopausal women as well as postmenopausal women on hormone replacement therapy Pretreatment: A few more women, smokers, and participants with previous hypertension were present in the placebo group, but the differences were not statistically significant.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 6 weeks) Placebo (duration 6 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous WDAEs up to week 6 Outcome type: dichotomous
Identification
Notes	Sponsorship source: This work was supported in part by an unrestricted grant from Bristol Myers Squibb, Canada. Country: Canada Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	Unclear risk	This work was supported in part by an unrestricted grant from Bristol Myers Squibb, Canada.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding for WDAE outcome assessment was not reported.

Frederiksen 1993.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 189 Total cholesterol: 7.03 mmol/L (272 mg/dL) LDL cholesterol: 4.94 mmol/L (191 mg/dL) HDL cholesterol: 1.32 mmol/L (51.0mg/dL) Triglycerides: 1.75 mmol/L (155 mg/dL) Placebo n: 96 Total cholesterol: 7.04 mmol/L (272 mg/dL) LDL cholesterol: 4.95 mmol/L (191 mg/dL) HDL cholesterol: 1.34 mmol/L (51.8 mg/dL) Triglycerides: 1.68 mmol/L (149 mg/dL) Included criteria: total cholesterol 6 mmol/L to 8 mmol/L (after diet regimen); ages 20 years to 69 years; men or postmenopausal woman; occurrence of at least one of the following risk factors: ischemic heart disease (IHS), smoking, arterial hypertension, familial disposition to IHS, symptoms of IHS (formerly acute myocardial infarction, angina pectoris) Excluded criteria: women premenopausal (unless sterilized); medications – corticosteroids, antacids, estrogen or progestogens, androgens, lipid‐lowering drugs, immunosuppressants, beta‐blockers, thiazides, alcohol, or drugs; disease – acute myocardial infarction or cardiac surgery or general surgery in the past 3 months, unstable angina pectoris, uncontrolled hypertension; hyperlipidemia – secondary hyperlipidemia, solid triglyceride ≥ 5 mmol/L; other diseases – liver disease, kidney disease, pancreatitis, gastrointestinal disease affecting absorption conditions, metabolic (e.g. diabetes mellitus or gout), or use of other HMG‐CoA reductase inhibitors; overweight – > 40% of ideal body weight; LDL cholesterol < 4 mmol/L Pretreatment: Blood pressure, both systolic and diastolic, was generally slightly higher among placebo‐treated than pravastatin‐treated participants, both among participants followed in hospital and general practice, but the differences were small, insignificant, and without clinical relevance.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 26 weeks) Placebo (duration 26 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6, 10 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6, 10 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6, 10 Outcome type: continuous Triglycerides percentage change from baseline at week 6, 10 Outcome type: continuous WDAE up to week 26 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: Denmark Setting: general practices and hospitals
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	Unclear risk	Source of funding was not reported.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (189−181) * 100/189 = 4.2% were not included in the efficacy analysis. Placebo: (96−91) * 100/96 = 5.2% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in both trial arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Blinding of outcome assessment for WDAEs was not reported.

Furberg 1994.

Study characteristics
Methods	Study design: randomized, single‐center, double‐masked, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day Age (years): 62.7 n: 75 Males: 63 Females: 12 LDL cholesterol: 167.2 mg/dL (4.32 mmol/L) HDL cholesterol: 42.2 mg/dL (1.09 mmol/L) Placebo Age (years): 62.4 n: 76 Males: 66 Females: 10 LDL cholesterol: 164.2 mg/dL (4.25 mmol/L) HDL cholesterol: 40.4 mg/dL (1.04 mmol/L) Included criteria: history of myocardial infarction or angiographic evidence of 250% stenosis in any coronary artery, moderately elevated plasma LDL cholesterol (between the 60th and 90th percentiles for age and gender), and a carotid artery atherosclerotic plaque with an intimal medial thickness of 21.3 mm. In total, 151 male and female prticipants with prevalent CAD between the ages of 50 years and 75 years were randomized to either pravastatin (n = 75) or placebo (n = 76). Excluded criteria: plasma triglyceride concentration ≥ 350 mg/dL, untreated hypo‐ or hyperthyroidism or other endocrine disease, secondary hyperlipidemia, recent myocardial infarction (≤ 6 months), severe or unstable angina pectoris, uncontrolled congestive heart failure or hypertension, significant gastrointestinal disease or surgery that might interfere with drug absorption, excessive ethanol consumption, likelihood of moving from the community within 3 years, limited expected life span, and treatment with certain drugs including corticosteroids, androgens, other lipid‐lowering agents, or antacids containing aluminum salts (from Crouse 1992) Pretreatment: None
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 3 years) Placebo (duration 3 years)
Outcomes	LDL cholesterol percentage change from baseline at month 1 Outcome type: continuous
Identification
Notes	Sponsorship source: a grant from Bristol Myers Squibb, Princeton, New Jersey Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	High risk	This report was supported by a grant from Bristol Myers Squibb, Princeton, New Jersey.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Galvan 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 45 n: 11 Male: 7 Female: 4 Total cholesterol: 7.9 mmol/L (305 mg/dL) LDL cholesterol: 6.1 mmol/L (236 mg/dL) HDL cholesterol: 1.2 mmol/L (46.4 mg/dL) Triglycerides: 1.4 mmol/L (124 mg/dL) Placebo Age: 46 n: 9 Male: 6 Female: 3 Total cholesterol: 7.7 mmol/L (298 mg/dL) LDL cholesterol: 5.8 mmol/L (224 mg/dL) HDL cholesterol: 1.2 mmol/L (46.4 mg/dL) Triglycerides: 1.5 mmol/L (133 mg/dL) Included criteria: 20 participants with primary familial hypercholesterolemia (of the heterozygous type) age 30 years to 60 years; body mass index < 26; fasting plasma glucose levels < 5.7 mM and normal glucose tolerance by National Diabetes Data Group criteria; arterial blood pressure < 160/90 mmHg; plasma triglyceride levels < 2.90 mM Excluded criteria: not reported Pretreatment: Clinical characteristics and the baseline serum levels of total and LDL cholesterol were similar in the placebo and pravastatin groups.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: a grant from Bristol Myers Squibb Country: Italy Setting: lipid clinic
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	Nno WDAE outcome reported
Other bias	Unclear risk	This work was supported in part by a grant from Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Gannagé‐Yared 2005.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 51.6 n: 19 Male: 10 Female: 9 Total cholesterol: 234 mg/dL (6.05 mmol/L) LDL cholesterol: 159.6 mg/dL (4.13 mmol/L) HDL cholesterol: 44.1 mg/dL (1.14 mmol/L) Triglycerides: 146.6 mg/dL (1.66 mmol/L) Placebo Age: 46.3 n: 21 Male: 12 Female: 9 Total cholesterol: 229.8 mg/dL (5.94 mmol/L) LDL cholesterol: 149.7 mg/dL (3.87 mmol/L) HDL cholesterol: 47.8 mg/dL (1.24 mmol/L) Triglycerides: 168.5 mg/dL (1.90 mmol/L) Included criteria: healthy nondiabetic volunteers Excluded criteria: chronic or acute inflammatory diseases, known cardiovascular diseases, diabetes defined as fasting blood glucose > 126 mg/dL according to the last American Diabetes Association classification, elevation of alanine aminotransferase (serum glutamic‐pyruvic transaminase), or creatine phosphokinase levels > 1.5 the upper limit of normal, volunteers who were on any kind of lipid‐lowering therapy Pretreatment: Baseline demographics and risk factors were similar between the two groups.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	SDs can be calculated from P values. Sponsorship source: Dr Azar was supported by grant FM 56 from the council of research of the St. Joseph University and by a grant from Bristol Myers Squibb, Beirut, Lebanon. Dr Gannagé‐Yared was supported by grant FM 35 from the council of research of the St. Joseph University. Country: Lebanon Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Dr Azar was supported by grant FM 56 from the council of research of the St. Joseph University and by a grant from Bristol Myers Squibb, Beirut, Lebanon. Dr Gannagé‐Yared was supported by grant FM 35 from the council of research of the St. Joseph University.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Gibellato 2001.

Study characteristics
Methods	Study design: randomized, double‐blind, double‐dummy, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 34.1 n: 27 Placebo Age: 33.5 n: 26 Included criteria: The study was conducted in active or retired military officer aircrew (pilots, navigators, weapons system operators, flight surgeons) with primary hypercholesterolemia. The study participants were chosen on the basis of having hypercholesterolemia despite dietary intervention. Hypercholesterolemia was defined as a mean LDL cholesterol ≥ 75th percentile or ≥ 130 mg/dL with one of the following CAD risk factors: a) smoking; b) treated hypertension; c) personal or family history of CAD; d) HDL cholesterol < 40 mg/dL; e) total cholesterol/HDL ratio > 5.0; or f) a history of definite cerebrovascular or occlusive peripheral vascular disease. The qualifying LDL cholesterol was based on a mean of two determinations performed at least 1 week apart, after a minimum of 4 weeks of dietary run‐in. Agree­ment between the two LDL cholesterol values was within 15% of the mean value. Mean plasma concentration of tri­glyceride in qualifying specimens was < 350 mg/dL. Mean total cholesterol was ≥ 200 mg/dL despite dietary intervention. Excluded criteria: diagnosed with or had a history of homozygous familial hypercholesterolemia, secondary hyperlipidemias, type I, III, or V hyperlipoproteinemia, sensitivity to HMG‐CoA reductase inhibitors, hypothyroidism or other uncontrolled endocrine disease, renal disease, hepatic disease, dysproteinemia, porphyria, lupus erythematosus, obesity, creatine kinase levels greater than twice the upper limit of normal for the laboratory, pancreatitis, hypertension presence of gastrointestinal disease, cardiac pathologies, or substance abuse; participants treated with corticosteroids, estrogens, androgens, lipid‐lowering agents, fish oil preparations, psyllium, antacids containing aluminum salts, investigational drugs, immunosuppressive drugs, beta‐adrenergic blocking agents, phenytoin, or central nervous system (CNS) active drugs Pretreatment: At baseline, there were no differences between the groups in the measure of total cholesterol (P = 0.695), HDL (P = 0.929), or LDL (P = 0.555).
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 4 weeks) Placebo (duration 4 weeks)
Outcomes	HDL cholesterol percentage change from baseline at week 4 Outcome type: continuous WDAEs at week 4 Outcome type: dichotomous
Identification
Notes	Sponsorship source: This study was supported entirely through a research grant from Bristol Myers Squibb, Inc. Country: USA Setting: naval hospital
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	High risk	LDL cholesterol outcome was not reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAEs were not reported.
Other bias	High risk	This study was supported entirely through a research grant from Bristol Myers Squibb, Inc.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Glasser 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 258 Male: 135 Female: 123 Total cholesterol: 275 mg/dL (7.11 mmol/L) LDL cholesterol: 198 mg/dL (5.12 mmol/L) HDL cholesterol: 42 mg/dL (1.09 mmol/L) Triglycerides: 169 mg/dL (1.91 mmol/L) Placebo n: 89 Male: 41 Female: 48 Total cholesterol: 278 mg/dL (7.19 mmol/L) LDL cholesterol: 201 mg/dL (5.20 mmol/L) HDL cholesterol: 44 mg/dL (1.14 mmol/L) Triglycerides: 168 mg/dL (1.90 mmol/L) Included criteria: patients with ≥ 2 standard CAD risk factors, and with an LDL cholesterol level (calculated from the Friedewald equation) between 160 mg/dL and 190 mg/dL patients with no standard risk factors and an LDL cholesterol > 190 mg/dL; plasma triglyceride concentration required to be < 400 mg/dL Excluded criteria: use of estrogen therapy (beyond that needed for replacement therapy) and any significant comorbidities, whether endocrinologic (e.g. diabetes mellitus, thyroid disease, and so forth), renal (nephrotic syndrome or serum creatinine > 1.5 times the upper limit of normal, urinary protein ≥ 2+, serum albumin < 3.0 g/dL), hepatic (liver enzyme values > 1 time the upper limit of normal), or abnormal creatine phosphokinase levels; acute coronary or cerebrovascular presentations within the preceding 6 months; severe obesity (> 40% above ideal body weight); uncontrolled systemic hypertension; and treatment with concomitant medications known to significantly affect lipid levels or to interfere with pravastatin pharmacokinetics Pretreatment: The only significant differences between the two groups in terms of demographics and baseline clinical variables were that the pravastatin group had more participants with a history of renal disease (8% vs 1%), although none of these participants had evidence of renal disease at the time of screening, and arthritis (34% vs 21%).
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 12 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Company Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	This study was supported by the Bristol Myers Squibb company.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (258−253) * 100/258 = 1.9% participants were not included in the efficacy analysis. Placebo: (89−87) * 100/89 = 2.2 % participants were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in the placebo and pravastatin arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding with respect to withdrawal due to adverse effects was not reported.

Guillen 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day Age: 46.8 n: 51 Male: 15 Female: 36 Total cholesterol: 232.6 mg/dL (6.02 mmol/L) LDL cholesterol: 156.8 mg/dL (4.05 mmol/L) HDL cholesterol: 41.9 mg/dL (1.08 mmol/L) Triglycerides: 170.6 mg/dL (1.93 mmol/L) Placebo Age: 47.8 n: 54 Male: 18 Female: 36 Total cholesterol: 233.6 mg/dL (6.04 mmol/L) LDL cholesterol: 155.8 mg/dL (4.03 mmol/L) HDL cholesterol: 43.0 mg/dL (1.11 mmol/L) Triglycerides: 161.7 mg/dL (1.83 mmol/L) Included criteria: a total of 150 men and women, aged 21 years to 69 years (mean, 47.77 years), with borderline or moderate hypercholesterolemia and two or more CHD risk factors such as smoking, controlled arterial hypertension, or family history of CHD; total cholesterol levels between 200 mg/dL and 260 mg/dL; LDL cholesterol ≥ 130 mg/dL; and triglyceride levels < 350 mg/dL Excluded criteria: younger than 18 years of age; history of drug or alcohol abuse; obstructive hepatobiliary disease or evidence of active liver disease; unstable angina pectoris; poorly controlled congestive heart failure; uncontrolled hypertension despite treatment; serum creatinine > 1.5 times the upper limit of normal; concurrent use of lipid‐­lowering agents, immunosuppressive drugs, or steroids, except in postmenopausal women taking stable doses of estrogen alone or with pro­gestin as replacement therapy; premenopausal women (unless surgically sterilized or provided with an intrauterine device for at least 6 months); patients with homozygous familial hypercholesterolemia; type I, III, IV, or V hyperlipoproteinemia; secondary hyperlipidemia of any cause; nephrotic syndrome; chronic pancreatitis; untreated hypothyroid­ism; extreme obesity; history of an acute medical condition or surgery within the previous 3 months; or any other condition that might interfere with interpretation of the results Pretreatment: No significant differences in sex, age, height, body weight, blood pressure, or lipid values were noted at randomization between the pravastatin and placebo groups.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 26 weeks) Bedtime dosing Placebo (duration 26 weeks) Bedtime dosing
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous WDAE up to week 26 Outcome type: dichotomous
Identification
Notes	Sponsorship source: grants from Bristol Myers Squibb, Princeton, New Jersey Country: Mexico Setting: clinical
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported for both arms of the trial.
Other bias	High risk	This study was supported by grants from Bristol Myers Squibb, Princeton, New Jersey.
Incomplete outcome data (attrition bias) All outcomes	High risk	Pravastatin: (76−51) * 100/76 = 32.9% were not included in the efficacy analysis. Placebo: (74−54) * 100/74 = 27.0% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is high in the placebo and pravastatin arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of WDAE assessment was not reported.

Hoogerbrugge 1990.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 45 n: 40 Male: 22 Female: 18 Total cholesterol: 10.6 mmol/L (410 mg/dL) LDL cholesterol: 8.7 mmol/L (336 mg/dL) HDL cholesterol: 1.13 mmol/L (43.7 mg/dL) Triglycerides: 1.57 mmol/L (139 mg/dL) Placebo Age: 44 n: 22 Male: 16 Female: 6 Total cholesterol: 10.2 mmol/L (394 mg/dL) LDL cholesterol: 8.3 mmol/L (321 mg/dL) HDL cholesterol: 1.14 mmol/L (44.1 mg/dL) Triglycerides: 1.59 mmol/L (141 mg/dL) Included criteria: Sixty‐two participants (38 men and 24 women) with heterozygous familial hypercholesterolemia LDL cholesterol and triglyceride concentrations, obtained at least 1 week apart, had to be > 5.0 mmol/L and < 4.0 mmol/L, respectively. Excluded criteria: none Pretreatment: The two study groups were generally similar with regard to their baseline characteristics and lipoprotein levels.
Interventions	Intervention characteristics Pravastatin 40 mg/day bedtime dosing (duration 24 weeks) Placebo bedtime dosing (duration 24 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 24 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: the Netherlands Setting: clinical
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported. None of the participants had to discontinue medication because of side effects.
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding for WDAE was not reported.

Hoogerbrugge‐vd Linden 1990.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 6 Total cholesterol: 10.3 mmol/L (398 mg/dL) LDL cholesterol: 7.4 mmol/L (286 mg/dL) HDL cholesterol: 1.09 mmol/L (42.2 mg/dL) Triglycerides: 1.59 mmol/L (141 mg/dL) Placebo n: 3 Total cholesterol: 9.8 mmol/L (379 mg/dL) LDL cholesterol: 7.2 mmol/L (278 mmol/L) HDL cholesterol: 1.11 mmol/L (42.9 mg/dL) Triglycerides: 1.99 mmol/L (176 mg/dL) Included criteria: Nine participants (two postmenopausal women and seven men), aged 22 years to 54 years, with heterozygous familial hypercholesterolemia LDL cholesterol concentrations obtained ≥ 1 week apart was more than 5.0 mmol/L in all participants. The mean plasma triglyceride concentration in the same specimens was < 4.0 mmol/L. Excluded criteria: diabetes mellitus, hypothyroidism, or any other cause of secondary hyperlipidemia Pretreatment: not reported
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: not reported Country: the Netherlands Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Jacobson 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 57 n: 182 Male: 115 Female: 67 Total cholesterol: 7.27 mmol/L (281 mg/dL) LDL cholesterol: 5.44 mmol/L (210 mg/dL) HDL cholesterol: 1.15 mmol/L (44 mg/dL) Triglycerides: 1.49 mmol/L (132 mg/dL) Placebo Age: 55.9 n: 63 Male: 38 Female: 25 Total cholesterol: 7.32 mmol/L (283 mg/dL) LDL cholesterol: 5.46 mmol/L (211 mg/dL) HDL cholesterol: 1.18 mmol/L (46 mg/dL) Triglycerides: 1.48 mmol/L (131 mg/dL) Included criteria: Two hundred and forty‐five male and female patients of African‐American descent, ranging in age from 18 years to 75 years, were randomized by 39 centers in a 3:1 ratio to pravastatin treatment (n = 182) or placebo (n = 63); LDL cholesterol level reached ≥ 4.9 mmol/L ( ≥ 190 mg/dL) or > 4.1 mmol/L (> 160 mg/dL) in the presence of either a personal history of CAD or at least two CAD risk factors as set forth by the National Cholesterol Education Program. Maximum allowable plasma triglyceride concentration in qualifying specimens was 4.5 mmol/L (400 mg/dL) with a 4‐week dietary lead‐in period. Participants were instructed to take tablets once daily at bedtime. Excluded criteria: type I, III, IV, or V hyperlipoproteinemia; homozygous familial hypercholesterolemia or secondary hyperlipidemia; poorly controlled diabetes (fasting blood glucose level > 7.8 mmol/L [> 140 mg/dL]); blood pressure of > 160/100 mmHg; those who were in congestive heart failure, had suffered a myocardial infarction or unstable or severe angina within 6 months, or had a stroke or transient ischemic attack within 3 months; nephrotic syndrome or other major renal disease; obstructive hepatobiliary or other major liver dysfunction; gastrointestinal tract surgery or disease that might interfere with drug absorption; excessive obesity (> 40% above ideal body weight); a weekly consumption of > 14 alcoholic beverages; pregnant or lactating women Pretreatment: The two treatment groups were exceedingly well matched with regard to baseline demographic characteristics and serum lipid and lipoprotein profiles.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous WDAEs up to week 12 Outcome type: dichotomous
Identification
Notes	Sponsorship source: grant from Bristol Myers Squibb US Pharmaceuticals Research Division, Princeton, New Jersey Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Low risk	Pravastatin sodium (20 mg) and matching placebo were given as identical‐appearing tablets.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	This study was supported by a grant from Bristol Myers Squibb US Pharmaceuticals Research Division, Princeton, New Jersey.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (182−179) * 100/182 = 1.6% participants were not included in the efficacy analysis. Placebo: (63−60) * 100/63 = 4.8% participants were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in the placebo and pravastatin arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment of WDAEs was not reported.

Jones 1991.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 5 mg/day Age: 52.3 n: 22 Male: 12 Female: 10 LDL cholesterol: 5.92 mmol/L (229 mg/dL) Pravastatin 10 mg/day Age: 54.2 n: 22 Male: 10 Female: 12 LDL cholesterol: 6.35 mmol/L (245.5 mg/dL) Pravastatin 20 mg/day Age: 54.1 n: 19 Male: 10 Female: 9 LDL cholesterol: 5.92 mmol/L (229 mg/dL) Pravastatin 40 mg/day Age: 55.6 n: 20 Male: 9 Female: 11 LDL cholesterol: 5.43 mmol/L (210 mg/dL) Placebo Age: 52.9 n: 42 Male: 22 Female: 20 LDL cholesterol: 6.02 mmol/L (233 mg/dL) Included criteria: One hundred and fifty outpatients (75 men, 75 women), age 21 years to 70 years, with hypercholesterolemia, who were maintained on a low‐fat, low‐cholesterol diet (either American Heart Association Phase I [1987] or a more restrictive diet at the investigator’s discretion), received randomized therapy during this 8‐week, dose‐response study. Patients had primary hypercholesterolemia with a mean plasma LDL cholesterol ≥ the 85th percentile for age and sex (minimum value of 3.88 mmol/L; 150 mg/dL), and plasma triglycerides < 2.82 mmol/L (250 mg/dL). Excluded criteria: non‐type II hypercholesterolemia, premenopausal women (unless surgically sterilized), diabetes mellitus, individuals with impairment of hepatic function or significant renal or endocrine disease, obesity (> 40% above ideal body weight), uncontrolled congestive heart failure or hypertension (> 160/100 mmHg), myocardial infarction within 6 months, or severe or unstable angina pectoris; patients who consumed more than 10 alcoholic drinks per week or who used corticosteroids, androgens, or estrogens (except replacement doses of ≤ 1.25 conjugated estrogens), fish oil supplements, or anticoagulants Pretreatment: No significant differences were detected among the five treatment groups.
Interventions	Intervention characteristics Pravastatin 5 mg/day (duration 8 weeks) Pravastatin 10 mg/day (duration 8 weeks) Pravastatin 20 mg/day (duration 8 weeks) Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 8 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: USA Setting: outpatient
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome reported
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Placebo: (42−36) * 100/42 = 14.3% participants were not included in the efficacy analysis. Pravastatin 5 mg/day: (22−16) * 100/22 = 27.3% participants were not included in the efficacy analysis. Pravastatin 10 mg/day: (22−18) * 100/22 = 18.2% participants were not included in the efficacy analysis. Pravastatin 20 mg/day: All participants were included in the efficacy analysis. Pravastatin 40 mg/day: (20−18) * 100/22 = 9.1% participants were not included in the efficacy analysis. Absolute rate of loss to follow‐up is different and high in some arms of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Jukema 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 450 Age (years): 56.5 Total cholesterol: 6.02 mmol/L (232.8 mg/dL) LDL cholesterol: 4.30 mmol/L (166.3 mg/dL) HDL cholesterol: 0.93 mmol/L (36.0 mg/dL) Triglycerides: 1.77 mmol/L (156.8 mg/dL) Placebo n: 434 Age (years): 55.9 Total cholesterol: 6.05 mmol/L (234.0 mg/dL) LDL cholesterol: 4.31 mmol/L (166.7 mg/dL) HDL cholesterol: 0.93 mmol/L (36.0 mg/dL) Triglycerides: 1.80 mmol/L (159.4 mg/dL) Included criteria: Male patients < 70 years old undergoing coronary arteriography to assess anginal complaints and patients with a qualifying baseline lipid and lipoprotein measurement by the Lipid Reference Laboratories, with a total cholesterol level between 4.0 mmol/L and 8.0 mmol/L (155 mg/dL and 310 mg/dL) after about 4 weeks or more of dietary advice were included. If the patient has incurred a myocardial infarction, approximately 8 weeks must elapse before the qualifying cholesterol level is measured. Patients undergoing CABG or PTCA procedure must have qualifying cholesterol before procedure or while on stable therapy of antianginal drugs. Patients with at least one coronary stenosis ≥ 50% (visually assessed) in a major coronary artery were also included. Cineangiograms must be certified by the Center Coordinator and one other qualified cardiologist to ensure that the film is of a quality that permits quantitative coronary arteriography (QCA). The fact that the film has been certified will be recorded on the Case Report Form. Excluded criteria: age ≥ 70 years at entry at the initial coronary cinearteriography; inability or unwillingness to consent to and undergo a repeat coronary arteriography; noncompliance with recommended treatment; enrollment in another study protocol that includes a coronary cineangiogram and experimental drug therapy; fasting cholesterol < 4.0 mmol/L (155 mg/dL) or ≥ 8.00 mmol/L (310 mg/dL) or triglycerides ≥ 4.0 mmol/L (354 mg/dL) as determined by the CORE lipid laboratory; life‐threatening illnesses other than CAD in which life expectancy is less than the study duration or one of the following conditions – malignancy, cardiac valve disease requiring valve replacement, cardiomyopathy, previous CABG, previous PTCA within 1 year before randomization, cardiac pacemaker implant, clinical congestive heart failure after medical management requiring diuretics (ejection fraction < 0.3 if performed), complete A‐V block, complete left bundle branch block, Wolff‐Parkinson‐White syndrome; use of lipid‐lowering drugs ≤ 6 weeks before qualifying lipid measurement (≤ 12 weeks for fibrates or HMG‐CoA reductase inhibitor); history of poor response to other HMG‐CoA reductase inhibitor (< 15% decrease in total cholesterol at usual dose); immune disorder (systemic lupus, dysproteinemia, or major hypersensitivity or allergic disorders) or use of immunosuppressive therapy or corticosteroids; significant metabolic disease – renal disease (nephrotic syndrome, decreased renal function with serum creatinine ≥ 150 mmol/L [2.5 mg/dL]), hepatobiliary disease with aspartate aminotransferase or alanine aminotransferase > 1.5 times the normal upper limit, chronic or recurrent pancreatitis, severely overweight (Quetelet index > 30), muscle disorders, oral or insulin‐dependent diabetes mellitus (uncorrected hypothyroidism or hyperthyroidism [a euthyroid patient on stable replacement of thyroid hormone is acceptable]), treatment with chronic corticosteroids or androgens, porphyria; significant gastrointestinal disease or surgery that might interfere with drug absorption; excess ethanol consumption (> 3 drinks per day) (1 drink = 45 g of 40% liquor or equivalent) Pretreatment: The treatment groups were well balanced with respect to baseline characteristics.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 2 years) Placebo (duration 2 years)
Outcomes	Total cholesterol percentage change from baseline at month 2 Outcome type: continuous LDL cholesterol percentage change from baseline at month 2 Outcome type: continuous HDL cholesterol percentage change from baseline at month 2 Outcome type: continuous Triglycerides percentage change from baseline at month 2 Outcome type: continuous WDAEs up to year 2 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Co, Princeton, New Jersey Country: the Netherlands Setting: 11 hospitals in the Netherlands, seven university and four non‐university hospitals
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Mmethod of random sequence generation not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants and physicians were blinded to the result of randomization throughout the study.
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	The REGRESS study was sponsored by Bristol Myers Squibb Co.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Kayikcioglu 2003.

Study characteristics
Methods	Study design: randomized, single‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 47 n: 19 Male: 7 Female: 12 Total cholesterol: 193 mg/dL (4.99 mmol/L) LDL cholesterol: 121 mg/dL (3.13 mmol/L) HDL cholesterol: 43 mg/dL (1.11 mmol/L) Triglycerides: 135 mg/dL (1.52 mmol/L) Placebo Age: 45 n: 19 Male: 9 Female: 10 Total cholesterol: 193 mg/dL (4.99 mmol/L) LDL cholesterol: 121 mg/dL (3.13 mmol/L) HDL cholesterol: 45 mg/dL (1.16 mmol/L) Triglycerides: 133 mg/dL (1.50 mmol/L) Included criteria: 40 prospectively enrolled consecutive patients with the diagnosis of cardiac syndrome‐X Excluded criteria: previous myocardial infarction, congestive heart failure, diabetes mellitus, valvular heart disease (including mitral valve prolapse), overt cardiomyopathy, sinus node dysfunction or conduction disturbance (including left bundle branch block), impaired renal or liver functions, hyperlipidemia (LDL levels ≥ 160 mg/dL or triglycerides > 200 mg/dL), and thyroid disease Pretreatment: The clinical characteristics did not differ statistically among the treatment groups at baseline.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 3 months) Placebo (duration 3 months)
Outcomes	Total cholesterol percentage change from baseline at month 3 Outcome type: continuous LDL cholesterol percentage change from baseline at month 3 Outcome type: continuous HDL cholesterol percentage change from baseline at month 3 Outcome type: continuous Triglycerides percentage change from baseline at month 3 Outcome type: continuous
Identification
Notes	Sponsorship source: not reported Country: Turkey Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (20−19) * 100/20 = 5% were not included in the efficacy analysis. Placebo: (20−19) * 100/20 = 5% were not included in the efficacy analysis. Rate of loss to follow‐up is the same and low between the pravastatin arm and placebo arm.

Kim 2013.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 60 n: 28 Total cholesterol: 6.3 mmol/L (244 mg/dL) LDL cholesterol: 4.0 mmol/L (155 mg/dL) HDL cholesterol: 1.3 mmol/L (50.3 mg/dL) Triglycerides: 1.6 mmol/L (141.7 mg/dL) Pravastatin 40 mg/day Age: 60 n: 25 Total cholesterol: 6.2 mmol/L (240 mg/dL) LDL cholesterol: 3.8 mmol/L (147 mg/dL) HDL cholesterol: 1.4 mmol/L (54.1 mg/dL) Triglycerides: 1.6 mmol/L (141.7 mg/dL) Placebo Age: 57 n: 20 Total cholesterol: 6.3 mmol/L (244 mg/dL) LDL cholesterol: 4.1 mmol/L (159 mg/dL) HDL cholesterol: 1.4 mmol/L (54.1 mg/dL) Triglycerides: 1.6 mmol/L (141.7 mg/dL) Included criteria: Korean female patients between 30 years and 70 years of age who had type 2 diabetes mellitus and hypercholesterolemia; patients with a < 10‐year recorded history of type 2 diabetes mellitus diagnosed according to the American Diabetes Association criteria; a verified body mass index between 20 and 30 and a verified glycosylated hemoglobin level ≤ 9.0%; a fasting triglyceride level < 4.52 mmol/L and LDL cholesterol ≥ 3.36 mmol/L Excluded criteria: a history of ischemic heart disease or congestive heart failure (New York Heart Association Class II) in the preceding 3 months; biochemical evidence of renal impairment (estimated glomerular filtration rate (GFR) 60 mL/min/1.73 m2); liver enzyme levels 80 IU/L; severe diabetic complications such as dialysis, proliferative retinopathy or stroke, and being pregnant or nursing at the time of the study; concomitant use of medications such as angiotensin converting enzyme inhibitors or other angiotensin II receptor blockers at entry and the use of insulin; patients taking thiazolidinedione in the preceding 2 months Pretreatment: Significant differences were observed for DBP between the placebo and the 20 mg and 40 mg pravastatin groups.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 16 weeks) Pravastatin 40 mg/day (duration 16 weeks) Placebo (duration 16 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 16 Outcome type: dichotomous
Identification
Notes	Sponsorship source: unrestricted grant from Daiichi‐Sankyo, Co., Ltd. Korea and CJ Pharma Co., Ltd Country: Republic of Korea Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported.
Other bias	High risk	This study was supported by an unrestricted grant from DaiichiSankyo, Co., Ltd. Korea and CJ Pharma Co., Ltd.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Knipscheer 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 5 mg/day Age: 12.1 n: 18 Male: 7 Female: 11 Total cholesterol: 7.7 mmol/L (298 mg/dL) LDL cholesterol: 6.2 mmol/L (240 mg/dL) HDL cholesterol: 1.2 mmol/L (46 mg/dL) Triglycerides: 0.7 mmol/L (62 mg/dL) Pravastatin 10 mg/day Age: 12.0 n: 18 Male: 7 Female: 11 Total cholesterol: 7.6 mmol/L (294 mg/dL) LDL cholesterol: 6.1 mmol/L (236 mg/dL) HDL cholesterol: 1.1 mmol/L (43 mg/dL) Triglycerides: 0.8 mmol/L (70.9 mg/dL) Pravastatin 20 mg/day Age: 11.9 n: 18 Male: 7 Female: 11 Total cholesterol: 8.2 mmol/L (317 mg/dL) LDL cholesterol: 6.7 mmol/L (259 mg/dL) HDL cholesterol: 1.2 mmol/L (46 mg/dL) Triglycerides: 0.6 mmol/L (53.1 mg/dL) Placebo Age: 12.1 n: 18 Male: 4 Female: 14 Total cholesterol: 7.8 mmol/L (302 mg/dL) LDL cholesterol: 6.4 mmol/L (247 mg/dL) HDL cholesterol: 1.1 mmol/L (43 mg/dL) Triglycerides: 0.8 mmol/L (70.9 mg/dL) Included criteria: children with known heterozygous familial hypercholesterolemia; hypercholesterolemia had to be present in siblings, parents, or grandparents, or clinical manifestations of premature atherosclerosis had to be present before the age of 50 in first‐ or second‐degree relatives Excluded criteria: undergone major surgery in the past 3 months or when they used medication interfering with lipid metabolism (such as anticonvulsants, oral contraceptives, corticosteroids, fibric acid derivatives, or immunosuppressants); hepatic or renal dysfunction Pretreatment: Baseline characteristics were comparable with respect to sex, age, ethnicity, and lipoprotein profiles.
Interventions	Intervention characteristics Pravastatin 5 mg/day (duration 12 weeks) Pravastatin 10 mg/day (duration 12 weeks) Pravastatin 20 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Netherlands Country: the Netherlands Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome was reported.
Other bias	High risk	Bristol Myers Squibb Netherlands financially supported the trial.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Placebo: All participants were included in the efficacy analysis. Pravastatin 5 mg/day: All participants were included in the efficacy analysis. Pravastatin 10 mg/day: (18‐17) * 100/18 = 2.6% participants were not included in the efficacy analysis. Pravastatin 20 mg/day: All participants were included in the efficacy analysis. Absolute rate of loss to follow‐up is low in all trial arms.

Koh 2009.

Study characteristics
Methods	Study design: randomized, single‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 56 n: 42 Male: 15 Female: 27 Total cholesterol: 254 mg/dL (6.57 mmol/L) LDL cholesterol: 170 mg/dL (4.40 mmol/L) HDL cholesterol: 54 mg/dL (1.40 mmol/L) Triglycerides: 162 mg/dL (1.83 mmol/L) Placebo Age: 59 n: 42 Male: 16 Female: 26 Total cholesterol: 267 mg/dL (6.90 mmol/L) LDL cholesterol: 177 mg/dL (4.58 mmol/L) HDL cholesterol: 55 mg/dL (1.42 mmol/L) Triglycerides: 147 mg/dL (1.66 mmol/L) Included criteria: Patients with hypercholesterolemia (LDL cholesterol levels ≥ 130 mg/dL) were included. No participant had taken any lipid‐lowering agent, hormone replacement therapy, or antioxidant vitamin supplements during the 2 months preceding the study. Excluded criteria: overt liver disease, chronic renal failure, hypothyroidism, myopathy, uncontrolled diabetes, severe hypertension, stroke, acute coronary events, coronary revascularization within the preceding 3 months, or alcohol abuse Pretreatment: There were no significant differences between groups for any of the baseline measurements.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 2 months) Placebo (duration 2 months)
Outcomes	Total cholesterol percentage change from baseline at month 2 Outcome type: continuous LDL cholesterol percentage change from baseline at month 2 Outcome type: continuous HDL cholesterol percentage change from baseline at month 2 Outcome type: continuous Triglycerides percentage change from baseline at month 2 Outcome type: continuous
Identification
Notes	Sponsorship source: grants from established investigator award (2006‐1, 2007‐1), Gachon University Gil Medical Center Country: Republic of Korea Setting: a primary care setting in the Vascular Medicine and Atherosclerosis Unit
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Low risk	This study was partly supported by grants from established investigator award (2006‐1, 2007‐1), Gachon University Gil Medical Center. All authors declare no conflict of interest.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Placebo: (43−42) * 100/43 = 2.3% participants were not included in the efficacy analysis. Pravastatin: (43−42) * 100/43 = 2.3% participants were not included in the efficacy analysis. Rate of loss to follow‐up is the same and low in the pravastatin arm and the placebo arm of the trial.

Koh 2013.

Study characteristics
Methods	Study design: randomized, single‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 54 n: 53 Male: 21 Female: 32 Total cholesterol: 241 mg/dL (6.23 mmol/L) LDL cholesterol: 165 mg/dL (4.27 mmol/L) HDL cholesterol: 51 mg/dL (1.32 mmol/L) Triglycerides: 136 mg/dL (1.54 mmol/L) Placebo Age: 56 n: 53 Male: 22 Female: 31 Total cholesterol: 248 mg/dL (6.41 mmol/L) LDL cholesterol: 166 mg/dL (4.29 mmol/L) HDL cholesterol: 54 mg/dL (1.40 mmol/L) Triglycerides: 138 mg/dL (1.56 mmol/L) Included criteria: patients with hypercholesterolemia (LDL cholesterol levels ≥ 130 mg/dL) Excluded criteria: overt liver disease, chronic renal failure, hypothyroidism, myopathy, uncontrolled diabetes, severe hypertension, stroke, acute coronary events, coronary revascularization within the preceding 3 months, or alcohol abuse Pretreatment: There were no significant differences between groups for any of the baseline measurements.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 2 months) Placebo (duration 2 months)
Outcomes	Total cholesterol percentage change from baseline at month 2 Outcome type: continuous LDL cholesterol percentage change from baseline at month 2 Outcome type: continuous HDL cholesterol percentage change from baseline at month 2 Outcome type: continuous Triglycerides percentage change from baseline at month 2 Outcome type: continuous
Identification
Notes	Sponsorship source: grants from established investigator award (2008‐1, 2009‐1; K.K. Koh), Gachon University Gil Hospital and by the Intramural Research Program, National Center for Complementary and Alternative Medicine, National Institutes of Health (M.J. Quon) Country: Republic of Korea Setting: primary care setting in the Cardiology Clinic
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Low risk	This study was supported by grants from established investigator award (2008‐1, 2009‐1; K.K. Koh), Gachon University Gil Hospital and by the Intramural Research Program, National Center for Complementary and Alternative Medicine, National Institutes of Health (M.J. Quon).
Incomplete outcome data (attrition bias) All outcomes	Low risk	Placebo: (54−53) * 100/54 = 1.85% participants were not included in the efficacy analysis. Pravastatin 40 mg/day: (54−53) * 100/54 = 1.85% participants were not included in the efficacy analysis. Rate of loss to follow‐up is the same and low in both arms of the trial.

Krempf 1997.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day Age: 57.9 n: 43 Male: 28 Female: 15 Total cholesterol: 6.8 mmol/L (263 mg/dL) LDL cholesterol: 4.6 mmol/L (178 mg/dL) HDL cholesterol: 1.21 mmol/L (46.8 mg/dL) Triglycerides: 1.89 mmol/L (167 mg/dL) Placebo Age: 56.2 n: 43 Male: 22 Female: 21 Total cholesterol: 6.6 mmol/L (255 mg/dL) LDL cholesterol: 4.5 mmol/L (174 mg/dL) HDL cholesterol: 1.21 mmol/L (46.8 mg/dL) Triglycerides: 1.76 mmol/L (156 mg/dL) Included criteria: Patients between 40 years and 70 years old had a history of non‐insulin‐­dependent diabetes mellitus for ≥ 1 year and mild primary hypercholesterolemia according to the American consensus conference on dyslipidemia in patients with a high CHD risk (i.e. fasting total cholesterol levels between 5.2 mmol/L and 7.8 mmol/L [200 mg/dL to 300 mg/dL]). Glycosylated hemoglobin had to be < 200% of the upper limit of normal; no participant had triglyceride levels < 2.28 mmol/L (200 mg/dL). All participants were on a diabetic diet and received either an oral agent or insulin. Excluded criteria: homozygous familial hypercholesterolemia or with triglycerides > 5.7 mmol/L; patients taking lipid‐lowering agents, fish oil preparations, thiazide diuretics, or beta blockers; patients showing abnormal cardiac, hepatic, thyroid, or renal function; premenopausal women, unless surgically sterile or fitted with an intrauterine device for at least 6 months; patients with type I diabetes mellitus or any other disease which could pose a risk to the patient or interfere with the study objective Pretreatment: The treatment groups were comparable with respect to age, sex, weight, and blood pressure and hemoglobin A1c levels.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 16 weeks) Placebo (duration 16 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 7 Outcome type: continuous LDL cholesterol percentage change from baseline at week 7 Outcome type: continuous HDL cholesterol percentage change from baseline at week 7 Outcome type: continuous Triglycerides percentage change from baseline at week 7 Outcome type: continuous WDAEs up to week 16 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb company Country: France Setting: diabetic clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAEs were reported for each arm of the trial.
Other bias	High risk	This study was supported by Bristol Myers Squibb Company.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

LeQuanSang 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg BID = 40 mg/day Age: 52.3 n: 10 Male: 10 Female: 0 Total cholesterol: 7.6 mmol/L (294 mg/dL) LDL cholesterol: 5.6 mmol/L (217 mg/dL) HDL cholesterol: 1.44 mmol/L (55.7 mg/dL) Triglycerides: 1.24 mmol/L (110 mg/dL) Placebo Age: 49.8 n: 12 Male: 12 Female: 0 Total cholesterol: 7.74 mmol/L (299 mg/dL) LDL cholesterol: 5.74 mmol/L (222 mg/dL) HDL cholesterol: 1.36 mmol/L (52.6 mg/dL) Triglycerides: 1.41 mmol/L (125 mg/dL) Included criteria: 22 male participants with type II hypercholesterolemia, aged 35 years to 63 years plasma total cholesterol value > 6.5 mmol/L and triacylglycerol levels ≤ 2 mmol/L Excluded criteria: hypertension; diabetes; renal, hepatic, or metabolic diseases; CHD; clinical evidence of advanced atherosclerotic lesions of the abdominal aorta, cervical arteries, or lower‐limb arteries, which was confirmed by ultrasonic examination of carotid, abdominal aortic, and femoral arteries; patients who were receiving treatment with drugs capable of modifying lipid metabolism or platelet handling Pretreatment: platelet [Ca2+]i did not significantly vary with age, body mass index, or blood pressure. [Ca2+]j was, however, associated with some plasma lipid components.
Interventions	Intervention characteristics Pravastatin 20 mg twice a day = 40 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	Sponsorship source: research grant from Bristol Myers Squibb Country: France Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	This study was partially supported by a research grant from Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Lewis 2007.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 80 mg/day Age (years): 49.8 n: 163 Males: 77 Females: 86 Total cholesterol: 219 mg/dL (5.66 mmol/L) LDL cholesterol: 138.8 mg/dL (3.59 mmol/L) HDL cholesterol: 49.2 mg/dL (1.27 mmol/L) Triglycerides: 160 mg/dL (1.81 mmol/L) Placebo Age (years): 49.9 n: 163 Males: 92 Females: 71 Total cholesterol: 219.1 mg/dL (5.66 mmol/L) LDL cholesterol: 140.5 mg/dL (3.63 mmol/L) HDL cholesterol: 46.4 mg/dL (1.20 mmol/L) Triglycerides: 159 mg/dL (1.80 mmol/L) Included criteria: men and women, 18 years or older; those with any chronic, well‐compensated stable liver disease defined as follows – chronic hepatitis C documented by a positive viral load test, nonalcoholic fatty liver or nonalcoholic fatty liver disease documented by radiological imaging or liver biopsy, or any other chronic liver disease documented by the usual clinical parameters; those whose LDL cholesterol was > 100 mg/dL and whose serum TGs were < 400 mg/dL Excluded criteria: a pregnant or lactating female or a woman of childbearing potential who was unwilling to use an adequate method of contraception during the study; any of these laboratory abnormalities – alanine aminotransferase or aspartate aminotransferase levels > 5 times the upper limit of normal, a total bilirubin level above the normal limit, a serum creatinine level > 1.5 mg/dL, a creatine kinase level > 3 times the upper limit of normal, an albumin level below the lower limit of normal, a prothrombin time > 2 seconds, and a platelet count below the lower limit of normal; ascites, jaundice, or cirrhosis with a Child‐Pugh score > 5; a disorder affecting serum bilirubin (for example, hemolytic anemia, Gilbert’s disease, and sickle cell anemia); antiviral therapy for hepatitis B or C; prior lipid‐lowering medications or therapy for 8 weeks or more; cancer (other than basal cell carcinoma) or cancer chemotherapy; significant cardiovascular, cerebrovascular, renal, or thyroid disease or uncontrolled diabetes mellitus within 6 months prior to randomization Pretreatment: The treatment groups of the randomized participants were well balanced, and no statistically significant differences were observed.
Interventions	Intervention characteristics Pravastatin 80 mg/day (duration 36 weeks) Placebo (duration 36 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 4, 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4, 12 Outcome type: continuous Triglycerides percentage change from baseline at week 4, 12 Outcome type: continuous WDAEs up to week 12 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	The authors thank Eileen Kelly of Bristol Myers Squibb for her invaluable assistance with the manuscript preparation.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin 80 mg/day: (163−153) * 100/163 = 6.1% were not included in the efficacy analysis. Placebo: (163−154) * 100/163 = 5.5% were not included in the efficacy analysis. Rate of loss to follow‐up is low in both arms of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding for WDAE not reported

Ling 2005.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 65.7 n: 24 Male: 19 Female: 5 Total cholesterol: 5.46 mmol/L (211 mg/dL) LDL cholesterol: 3.41 mmol/L (132 mg/dL) HDL cholesterol: 1.16 mmol/L (44.9 mg/dL) Triglycerides: 1.93 mmol/L (171 mg/dL) Placebo Age: 64.2 n: 25 Male: 21 Female: 4 Total cholesterol: 5.60 mmol/L (217 mg/dL) LDL cholesterol: 3.54 mmol/L (137 mg/dL) HDL cholesterol: 1.21 mmol/L (46.8 mg/dL) Triglycerides: 1.98 mmol/L (175 mg/dL) Included criteria: participants had the diagnosis of CAD made by coronary angiogram (> 50%) stenosis, documented history of a previous myocardial infarction, or evidence of ischemia by a positive stress test or perfusion scan; LDL cholesterol > 3.0 mmol/L (116 mg/dL) and < 5.9 mmol/L (228 mg/dL) or HDL cholesterol < 0.9 mmol/L (35 mg/dL); not on statin therapy or off statin therapy for > 6 weeks; on stable antianginal medical therapy for ≥ 1 month Excluded criteria: age ≥ 90 years, SBP < 90 mmHg or > 170 mmHg or DBP > 100 mmHg, left ventricular ejection fraction < 40%, acute coronary syndrome (ACS) within 4 weeks, percutaneous coronary intervention within 6 months, had known significant left main disease or triple‐vessel disease with proximal left anterior descending coronary artery involvement, poorly controlled diabetes, hepatic dysfunction, hemodynamically significant valvular disease, second‐degree or third‐degree atrioventricular block or life‐threatening arrhythmia, asthma or reactive airway disease, or previous CABG Pretreatment: There was no statistically significant difference in the baseline characteristics analyzed, which included participant age, sex, height, weight, SBP and DBP, heart rate (HR), diabetes, hypertension, previous myocardial infarction, smoking, and various medication uses. Baseline lipid parameters were also comparable. There was a trend for more smokers in the placebo group, but this did not reach statistical significance (P = 0.08).
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 8 Outcome type: dichotomous
Identification
Notes	One out of 24 placebo participants and 1 out of 25 pravastatin participants are receiving fibrates at baseline. Sponsorship source: unrestricted research grant by Bristol Myers Squibb Country: Canada Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number table randomization in the study was performed by an independent observer and occurred in blocks.
Allocation concealment (selection bias)	Low risk	An independent observer performed the randomization.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	We kindly appreciate the support of this study by Bristol Myers Squibb.
Incomplete outcome data (attrition bias) All outcomes	High risk	Placebo: (25−21) * 100/25 = 16% were not included in the efficacy analysis. Pravastatin: (24−22) * 100/24 = 8.3% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is high in the placebo arm of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Blinding of outcome assessment for WDAEs was not reported.

Megnien 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 50 n: 13 Male: 12 Female: 1 Total cholesterol: 7.48 mmol/L (289 mg/dL) LDL cholesterol: 5.58 mmol/L (216 mg/dL) HDL cholesterol: 1.30 mmol/L (50.3 mg/dL) Triglycerides: 1.32 mmol/L (117 mg/dL) Placebo Age: 49 n: 15 Male: 14 Female: 1 Total cholesterol: 7.88 mmol/L(305 mg/dL) LDL cholesterol: 5.90 mmol/L(228 mg/dL) HDL cholesterol: 1.34 mmol/L(51.8 mg/dL) Triglycerides: 1.42 mmol/L(126 mg/dL) Included criteria: Twenty‐eight hypercholesterolaemic participants (26 men and 2 women) 35 years to 63 years of age were included in the study total blood cholesterol had to be > 6.2 mmol/L and plasma triglycerides be < 2 mmol/L (type IIa hypercholesterolaemia) Excluded criteria: blood pressure elevation > 160/95 mmHg on at least three outpatient visits or with a history of treatment for hypertension, CHD, and clinical evidence of advanced atherosclerotic lesions of the abdominal aorta, the cervical arteries, or the lower‐limb arteries Pretreatment: there were no significant differences between the two groups in age, sex ratio, body mass index, or blood pressure as well as in blood lipids
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 3 months) Placebo (duration 3 months)
Outcomes	Total cholesterol percentage change from baseline at month 3 Outcome type: continuous LDL cholesterol percentage change from baseline at month 3 Outcome type: continuous HDL cholesterol percentage change from baseline at month 3 Outcome type: continuous Triglycerides percentage change from baseline at month 3 Outcome type: continuous
Identification
Notes	Sponsorship source: not reported Country: France Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Pravastatin 40 mg/day: (15−13) * 100/15 = 13.3% participants were not included in the efficacy analysis. Placebo: (16−15) * 100/16 = 6.25% participants were not included in the efficacy analysis. Absolute rate of loss to follow‐up is high in the pravastatin arm of the trial.

Melani 2003.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day n: 64 Pravastatin 20 mg/day n: 64 Pravastatin 40 mg/day n: 64 Placebo n: 57 Included criteria: adult men or women with primary hypercholesterolemia plasma LDL cholesterol concentration ≥ 3.8 to ≤ 6.5 mmol/L, as calculated by the Friedewald equation, and triglyceride levels ≤ 4.0 mmol/L Excluded criteria: prohibited concomitant illnesses and procedures included congestive heart failure (defined as New York Heart Association Class III or IV heart failure), uncontrolled cardiac arrhythmias, history of unstable or severe peripheral artery disease (within 3 months of study entry), unstable angina pectoris, myocardial infarction, coronary bypass surgery, or angioplasty (within 6 months of study entry); uncontrolled or newly diagnosed (within 1 month of study entry) diabetes mellitus, active or chronic hepatic hepatobiliary disease, known impairment of renal function, known coagulopathy, and unstable endocrine disease; individuals receiving immunosuppressant drugs or corticosteroids Pretreatment: The treatment groups were comparable with regard to baseline lipid parameters and demographics.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 12 weeks) Pravastatin 20 mg/day (duration 12 weeks) Pravastatin 40 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous WDAEs up to week 12 Outcome type: dichotomous
Identification
Notes	Taking the data from Figure 2, the N for placebo should be 57 and that for the 3 pravastatin groups can be assumed to be 64 each. It is not perfect but will not introduce any bias. WDAE numbers are for pooled pravastatin doses not 40 mg/day. Sponsorship source: This study was conducted by Schering‐Plough Research Institute, Kenilworth, New Jersey, USA, on behalf of Merck/Schering‐Plough Pharmaceuticals, North Wales, Pennsylvania, USA Country: USA Setting: clinical multicenter
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated randomization schedule with treatment codes in blocks of eight
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	This study was conducted by Schering‐Plough Research Institute, Kenilworth, New Jersey, USA, on behalf of Merck/Schering‐Plough Pharmaceuticals, North Wales, Pennsylvania, USA.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Pravastatin: (205−192) * 100/205 = 6.3% were not included in the efficacy analysis. Placebo: (65−57) * 100/65 = 12.3% were not included in the efficacy analysis. Absolute rate of loss to follow‐up was moderate and different between the two arms of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of WDAE outcome assessment was not reported.

Mercuri 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 151 Age (years): 54.9 Males: 85 Females: 66 Total cholesterol: 6.72 mmol/L (259.9 mg/dL) LDL cholesterol: 4.66 mmol/L (180.2 mg/dL) HDL cholesterol: 1.35 mmol/L (52.2 mg/dL) Triglycerides: 1.56 mmol/L (138.2 mg/dL) Placebo n: 154 Age (years): 55.1 Males: 79 Females: 75 Total cholesterol: 6.80 mmol/L (263 mg/dL) LDL cholesterol: 4.71 mmol/L (182.1 mg/dL) HDL cholesterol: 1.38 mmol/L (53.4 mg/dL) Triglycerides: 1.55 mmol/L (137.3 mg/dL) Included criteria: men and women, 45 years to 65 years old, with moderately elevated LDL cholesterol levels (three base­line determinations of LDL cholesterol between 3.88 mmol/L and 6.4 7 mmol/L and triglyceride levels < 2.82 mmol/L); free of symptoms or signs of CAD; and at least one carotid artery lesion detected by quantitative B‐mode ultrasound imaging Excluded criteria: persistent liver function abnormalities, other serious medical conditions, regular use of lipid‐lowering agents, anticoagulants, and calcium antagonists Pretreatment: Treatment groups were not different for any of the demographic, clinical, laboratory or ultrasonographic criteria.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 3 years) Placebo (duration 3 years)
Outcomes	Total cholesterol percentage change from baseline at month 3 Outcome type: continuous LDL cholesterol percentage change from baseline at month 3 Outcome type: continuous HDL cholesterol percentage change from baseline at month 3 Outcome type: continuous Triglycerides percentage change from baseline at month 3 Outcome type: continuous WDAEs up to year 3 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb S. p. A. Italy and a grant from Italian National Research Council Country: Italy Setting: lipid clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization was not reported
Allocation concealment (selection bias)	Low risk	Participant randomization into one of the two treatment arms was performed by an independent coordinating and analysis center.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants were double blindly randomized. Lipid parameter measurements were unlikely to be influenced by methods of blinding.
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	Unclear risk	Funded through independent research grants provided by Bristol Myers Squibb S.p.A. Italy and in part by a grant from the Italian National Research Council (C.N.R. Progetto Finalizzato, Invecchiamento)
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Meyers 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 49.5 n: 172 Male: 0 Female: 172 Total cholesterol: 306.5 mg/dL (7.93 mmol/L) LDL cholesterol: 227.3 mg/dL (5.88 mmol/L) HDL cholesterol: 45.0 mg/dL (1.16 mmol/L) Triglycerides: 171.0 mg/dL (1.93 mmol/L) Placebo Age: 48.1 n: 58 Male: 0 Female: 58 Total cholesterol: 295.5 mg/dL (7.64 mmol/L) LDL cholesterol: 214.4 mg/dL (5.54 mmol/L) HDL cholesterol: 46.1 mg/dL (1.19 mmol/L) Triglycerides: 175.2 mg/dL (1.98 mmol/L) Included criteria: Caucasian women, 18 years to 60 years of age, with an LDL cholesterol value ≥ 190 mg/dL and triglyceride levels < 400 mg/dL with a variation < 12% in LDL cholesterol values between samples were included. Participants with LDL cholesterol between 160 mg/dL and 189 mg/dL (and variation < 12%) were enrolled if they had a personal history of myocardial infarction, angiographically documented coronary atherosclerosis, a history of coronary bypass surgery or coronary angioplasty at least 6 months prior to the study, or angina pectoris as defined by the Rose angina questionnaire 10 or had at least two of the accepted atherosclerosis risk factors, including hypertension, cigarette smoking, obesity, family history of CHD at age < 55 years, or diabetes mellitus. Any participant whose initial LDL cholesterol values were 12% to 18% apart was allowed a third qualify­ing laboratory evaluation. If that result was ≤ 12% from either of the previous determinations, the participant was accepted for enrollment. Excluded criteria: thyroid, liver, or renal dysfunction; uncontrolled hypertension (blood pressure > 160/100 mmHg); uncontrolled diabetes mellitus (fasting blood sugar > 140 mg/dL); musculoskeletal diseases or ele­vated creatine kinase; current use of thiazides, beta‐blockers, anticoagulants, corticosteroids, estrogens beyond replacement doses, andro­gens, fish oil, aluminum‐containing antacids, or immunosuppressives; chemotherapy for malignancy; or history of heart failure malabsorption, alcohol abuse, pancreatitis, dysproteinemia, porphyria, or lupus erythematosus Pretreatment: No differences at baseline were noted between groups in any other historical or physical parameters.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6, 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6, 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6, 12 Outcome type: continuous Triglycerides percentage change from baseline at week 6, 12 Outcome type: continuous WDAEs up to week 12 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Company, Princeton, New Jersey Country: USA Setting: outpatient and lipid‐screening clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random block design
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	The study was supported by a grant from Bristol Myers Squibb Company, Princeton, New Jersey. A 3:1 active to placebo allocation scheme was used to ensure an adequate sample yet, for ethical reasons, limit exposure of the hypercholesterolemia participants to placebo.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (172−169) * 100/172 = 1.7% participants were not included in the efficacy analysis. Placebo: (58−56) * 100/58 = 3.4% participants were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in both arms of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Blinding of WDAE outcome assessment was not reported.

Morris 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 50.6 n: 45 Male: 44 Female: 1 Total cholesterol: 5.9 mmol/L (228 mg/dL) LDL cholesterol: 3.9 mmol/L (151 mg/dL) HDL cholesterol: 1.1 mmol/L (42.5 mg/dL) Triglycerides: 1.7 mmol/L (151 mg/dL) Placebo Age: 52.1 n: 53 Male: 49 Female: 4 Total cholesterol: 6.2 mmol/L (240 mg/dL) LDL cholesterol: 4.3 mmol/L (166 mg/dL) HDL cholesterol: 1.1 mmol/L (42.5 mg/dL) Triglycerides: 1.8 mmol/L (159 mg/dL) Included criteria: 217 males and females aged 18 years to 70 years with two or more major risk factors for CHD (male sex, smoking more than 10 per day, hypertension, family history of CHD, or personal history of CHD) were opportunistically screened for moderate primary hypercholesterolemia (5.2 mmol/L to 6.7 mmol/L); mean cholesterol in the range of 5.2 mmol/L to 6.7 mmol/L, with triglycerides < 4.0 mmol/L and LDL > 3.4 mmol/L Excluded criteria: not reported Pretreatment: No differences were seen between the groups.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 26 weeks) Placebo (duration 26 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous WDAEs up to week 26 Outcome type: dichotomous
Identification
Notes	Sponsorship source: ER Squibb & Sons Pty (NZ) Ltd Country: New Zealand Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	ER Squibb & Sons Pty (NZ) Ltd. supported the trial.
Incomplete outcome data (attrition bias) All outcomes	Low risk	(98‐95)/98 * 100 = 3.1% were not included in the efficacy analysis. The rate of loss to follow‐up was not reported separately for the placebo and pravastatin groups, but attrition is low overall.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAE was not reported.

O'Callaghan 1994.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age (years): 56 n: 12 Males: 6 Females: 6 Total cholesterol: 7.1 mmol/L (274.6 mg/dL) LDL cholesterol: 4.8 mmol/L (185.6 mg/dL) HDL cholesterol: 1.13 mmol/L (43.7 mg/dL) Triglycerides: 2.50 mmol/L (221.4 mg/dL) Placebo Age (years): 62 n: 12 Males: 6 Females: 6 Total cholesterol: 7.2 mmol/L (278.8 mg/dL) LDL cholesterol: 4.9 mmol/L (189.5 mg/dL) HDL cholesterol: 1.10 mmol/L (42.5 mg/dL) Triglycerides: 2.64 mmol/L (233.8 mg/dL) Included criteria: The participants were required to have a DBP of < 95 mmHg treated with either captopril alone or captopril plus felodipine and total cholesterol level of > 5.5 mmol/L and a triglyceride level of < 6.0 mmol/L. Excluded criteria: none reported Pretreatment: Baseline demographic characteristics of the two treatment groups were similar.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 15 weeks) Placebo (duration 15 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 4 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4 Outcome type: continuous HDL cholesterol percentage change from baseline at week 4 Outcome type: continuous Triglycerides percentage change from baseline at week 4 Outcome type: continuous WDAEs up to week 6 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb, Noble Park, Victoria, Australia Country: Australia Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	This study was supported by a grant from Bristol Myers Squibb, Noble Park, Victoria, Australia.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	All participants were included in the efficacy analysis except one participant from an unknown group.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAE was not reported.

Pravastatin Multicenter Study Group II 1993.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg /day n: 63 Male: 47 Female: 16 Age: 50.8 Total cholesterol: 7.9 mmol/L (305 mg/dL) LDL cholesterol: 6.1 mmol/L (236 mg/dL) HDL cholesterol: 1.1 mmol/L (43 mg/dL) Triglycerides: 1.3 mmol/L (115 mg/dL) Pravastatin 80 mg/day n: 63 Male: 45 Female: 18 Age: 51.6 Total cholesterol: 8.0 mmol/L (309 mg/dL) LDL cholesterol: 6.1 mmol/L (236 mg/dL) HDL cholesterol: 1.2 mmol/L (46 mg/dL) Triglycerides: 1.4 mmol/L (124 mg/dL) Placebo n: 60 Male: 37 Female: 23 Age: 52.8 Total cholesterol: 8.0 mmol/L (309 mg/dL) LDL cholesterol: 6.0 mmol/L (232 mg/dL) HDL cholesterol: 1.2 mmol/L (46 mg/dL) Triglycerides: 1.3 mmol/L (115 mg/dL) Included criteria: patients with primary hypercholesterolemia (plasma LDL cholesterol level of 4.14 mmol/L [160 mg/dL] and > the 90th percentile for age and sex and plasma triglyceride level < 2.82 mmol/L after diet) Excluded criteria: premenopausal women unless surgically sterile; patients with non‐type II hyperlipoproteinemia, diabetes mellitus (fasting blood glucose concentrations > 140 mg/dL), impaired hepatic or renal function, severe or unstable angina, excessive obesity, or uncontrolled hypertension; patients who consumed > 10 alcoholic drinks per week; patients with hypersensitivity to cholestyramine; patients receiving corticosteroids, androgens, or estrogens except as continuous, stable replacement; patients taking anticoagulants, theophylline, barbiturates, or quinidine or those regularly taking aluminum‐containing antacids Pretreatment: The participant characteristics did not differ significantly among the five treatment groups.
Interventions	Intervention characteristics Pravastatin 40 mg /day (duration 24 weeks) twice‐daily: 20 Pravastatin 80 mg/day (duration 24 weeks) twice‐daily : 40 Placebo (duration 24 weeks) twice‐daily : 0
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAE up to week 8 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb Co, New Brunswick, New Jersey. Country: USA Setting: lipid research clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	This work was supported by a grant from Bristol Myers Squibb Co, New Brunswick, New Jersey.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin 20 mg BID: (63−62) * 100/63 = 1.6% were not included in the efficacy analysis. Pravastatin 40 mg BID: (63−61) * 100/63 = 3.2% were not included in the efficacy analysis. Placebo: (60−57) * 100/60 = 5% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in all trial arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Blinding of outcome assessment was not reported for WDAEs.

Ritter 1993.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 53.7 n: 79 Male: 66 Female: 13 Total cholesterol: 6.9 mmol/L (267 mg/dL) Placebo Age: 54.6 n: 75 Male: 61 Female: 14 Total cholesterol: 6.8 mmol/L (263 mg/dL) Included criteria: a fasting total cholesterol between 5.2 mmol/L and 7.8 mmol/L (based on the mean of two estimations performed 1 week apart); two or more of the following additional CHD risk factors – male gender, cigarette smoking, hypertension and personal or family history of CHD Excluded criteria: premenopausal women (unless surgically sterilized); concurrent use of drugs known to influence lipid metabolism (e.g. regular corticosteroids, immunosuppressive drugs, lipid‐lowering agents, estrogens and progestogens [except postmenopausal hormone replacement therapy], and androgens); history of drug or alcohol abuse; unstable angina pectoris; uncontrolled heart failure or hypertension; acute medical or surgical condition within the previous 3 months; homozygous familial hypercholesterolemia; secondary hyperlipidemia of any cause; type I, III, IV, or V hyperlipidemia; fasting serum triglyceride > 4.0 mmol/L; chronic pancreatitis; or excessive obesity (> 40% above ideal body weight) Pretreatment: The two groups were well matched as regards baseline cholesterol concentration and other characteristics, including the distribution of CHD risk factors.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 26 weeks) Placebo (duration 26 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous WDAEs up to week 26 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: UK Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	One of the authors is associated with Bristol Myers Squibb Pharmaceuticals.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (79−75) * 100/79 = 5.1% participants were not included in the efficacy analysis. Placebo: (75−70) * 100/75 = 6.7% participants were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in the placebo and pravastatin arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of WDAE outcome assessment was not reported.

Rosenson 2003.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trials
Participants	Baseline characteristics Pravastatin 40 mg/day n: 45 Total cholesterol: 274 mg/dL (7.09 mmol/L) LDL cholesterol: 193 mg/dL (4.99 mmol/L) HDL cholesterol: 49 mg/dL (1.27 mmol/L) Triglycerides: 162 mg/dL (1.83 mmol/L) Pravastatin 80 mg/day n: 277 Total cholesterol: 275 mg/dL (7.11 mmol/L LDL cholesterol: 191 mg/dL (4.94 mmol/L) HDL cholesterol: 49 mg/dL (1.27 mmol/L) Triglycerides: 174 mg/dL (1.96 mmol/L) Pravastatin 160 mg/day n: 272 Total cholesterol: 274 mg/dL (7.09 mmol/L) LDL cholesterol: 191 mg/dL (4.94 mmol/L) HDL cholesterol: 48 mg/dL (1.24 mmol/L) Triglycerides: 173 mg/dL (1.95 mmol/L) Placebo n: 162 Total cholesterol: 271 mg/dL (7.01 mmol/L) LDL cholesterol: 187 mg/dL (4.84 mmol/L) HDL cholesterol: 48 mg/dL (1.24 mmol/L) Triglycerides: 182 mg/dL (2.05 mmol/L) Included criteria: study 1 – men and women, 24 years to 65 years of age, with primary hypercholesterolemia (LDL cholesterol ≥ 160 mg/dL and triglycerides ≤ 500 mg/dL) and no history of cardiovascular disease; study 2 – patients ≥ 18 years if they were otherwise healthy men, or women who were not pregnant, lactating, or of childbearing potential, with primary hypercholesterolemia (LDL cholesterol ≥ 160 mg/dL and triglycerides < 400 mg/dL) Excluded criteria: study 1 and 2 – cardiovascular disease, symptomatic cerebrovascular disease, and the current use of any cholesterol‐lowering agents and non‐study drugs that affect lipid metabolism Pretreatment: not reported in studies 1 and 2
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 6 weeks) Bedtime dosing Pravastatin 80 mg/day (duration 6 weeks) Bedtime dosing Pravastatin 160 mg/day (duration 6 weeks) Bedtime dosing Placebo (duration 6 weeks) Bedtime dosing
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous
Identification
Notes	Sponsorship source: grants from Bristol Myers Squibb Company, Princeton, New Jersey Country: USA Setting: study 1 – not reported; study 2 – clinical sites
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAEs were not reported for each arm separately.
Other bias	High risk	This report was supported by grants from Bristol Myers Squibb Company, Princeton, New Jersey.
Incomplete outcome data (attrition bias) All outcomes	Low risk	(788‐756)/788 * 100 = 4.1% were not included in the efficacy analysis.

Rosenson 2004.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day n: 12 Male: 8 Female: 4 Age: 51 Total cholesterol: 5.92 mmol/L (229 mg/dL) LDL cholesterol: 4.18 mmol/L (162 mg/dL) HDL cholesterol: 1.02 mmol/L (39.4 mg/dL) Triglycerides: 1.73 mmol/L (153 mg/dL) Placebo n: 14 Male: 10 Female: 4 Age: 51 Total cholesterol: 6.13 mmol/L (237 mg/dL) LDL cholesterol: 4.20 mmol/L (162 mg/dL) HDL cholesterol: 1.12 mmol/L (43.3 mg/dL) Triglycerides: 1.64 mmol/L (145 mg/dL) Included criteria: healthy non‐smoking adults (aged 35 years to 65 years) with moderate (LDL cholesterol 3.4 mmol/L to 4.9 mmol/L) hypercholesterolemia Excluded criteria: inflammatory disorders, using antioxidant vitamins, anti‐inflammatory agents, or lipid‐altering agents Pretreatment: not reported
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: not reported Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blind, placebo‐controlled; lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAE outcome was not reported.
Other bias	Unclear risk	Source of funding was not reported.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Rubenfire 1991.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg BID = 20 mg/day n: 57 Total cholesterol: 6.85 mmol/L (265 mg/dL) LDL cholesterol: 5.17 mmol/L (200 mg/dL) HDL cholesterol: 1.11 mmol/L (42.9 mg/dL) Triglycerides: 1.64 mmol/L (145 mg/dL) Placebo n: 25 Total cholesterol: 6.75 mmol/L (261 mg/dL) LDL cholesterol: 5.02 mmol/L (194 mg/dL) HDL cholesterol: 1.14 mmol/L (44.1 mg/dL) Triglycerides: 1.72 mmol/L (152 mg/dL) Included criteria: men and postmenopausal or surgically sterile women between the ages of 21 years and 75 years with a diagnosis of primary hypercholesterolemia by their referring physician; initial eligibility based on a medical history and on screening or previously available plasma cholesterol values (i.e. > 6.98 mmol/L [> 270 mg/dL]) Excluded criteria: plasma triglyceride levels > 2.82 mmol/L; homozygous familial hypercholesterolemia; type I, III, IV, or V hyperlipoproteinemia; uncontrolled thyroid disease; uncontrolled hypertension (SBP > 160 mm Hg or DBP > 100 mm Hg); uncontrolled non‐insulin‐dependent diabetes or insulin‐dependent diabetes (fasting blood glucose value > 7.77 mmol/L); excessive obesity (> 30% over ideal body weight); excessive alcohol consumption (> 3 drinks per day); significant hepatobiliary or renal disease (aspartate aminotransferase, alanine aminotransferase, or lactate dehydrogenase levels > 2 times or creatinine > 1.5 times, the upper limit of normal); and treatment with medications known to alter lipid metabolism (i.e. lipid‐lowering agents, steroids, and fish oil supplements). Participants with known atherosclerosis were excluded only if they had suffered a significant vascular event (e.g. stroke, myocardial infarction, vascular surgery) in the preceding 3 months. Pretreatment: The baseline pretreatment plasma lipid, lipoprotein cholesterol, and apolipoprotein concentrations were not significantly different between the two participant groups the baseline dietary values and body weights were not significantly different.
Interventions	Intervention characteristics Pravastatin 10 mg twice‐daily = 20 mg/day (duration 16 weeks) Placebo (duration 16 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 16 Outcome type: dichotomous
Identification
Notes	Sponsorship source: grants from the Squibb Institute for Medical Research, E. R. Squibb and Sons, Princeton, New Jersey, and the Sinai Hospital Guild Country: USA Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	Unclear risk	This study was supported in part by grants from the Squibb Institute for Medical Research, E. R. Squibb and Sons, Princeton, New Jersey, and the Sinai Hospital Guild.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Rustemeijer 1997.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 52 n: 24 Male: 14 Female: 10 Total cholesterol: 7.3 mmol/L (282 mg/dL) LDL cholesterol: 5.0 mmol/L (193 mg/dL) HDL cholesterol: 1.3 mmol/L (50.3 mg/dL) Triglycerides: 1.9 mmol/L (168 mg/dL) Placebo Age: 55 n: 25 Male: 12 Female: 13 Total cholesterol: 7.3 mmol/L (282 mg/dL) LDL cholesterol: 5.0 mmol/L (193 mg/dL) HDL cholesterol: 1.3 mmol/L (50.3 mg/dL) Triglycerides: 2.1 mmol/L (186 mg/dL) Included criteria: men and postmenopausal or surgically sterile women, or women with an IUD for at least 6 months aged between 18 years and 70 years, known to have hypercholesterolemia – defined as total cholesterol ≥ 6.5 mmol/L and ≤ 8.0 mmol/L – and diabetes mellitus, insulin dependent or non‐insulin dependent, for at least 2 years, with adequate control as indicated by glycosylated hemoglobin ≤ 10% Excluded criteria: primary hyperlipidemias; use of alcohol or drugs which influenced carbohydrate or lipid metabolism; patients with renal, hepatobiliary, or endocrine diseases; pancreatitis, porphyria; excessive obesity (Quetelet index ≥ 30), malignancy or with recent signs or symptoms of cardiovascular disease Pretreatment: no difference in baseline characteristics between the groups and in medication used, dietary compliance, alcohol intake, smoking habits, risk factors for CHD, complications of diabetes mellitus, blood pressure, or glycosylated hemoglobin levels
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 24 weeks) Bedtime dosing Placebo (duration 24 weeks) Bedtime dosing
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous WDAEs up to week 24 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Bristol Myers Squibb B. V., the Netherlands Country: the Netherlands Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	A grant from the Bristol Myers Squibb B. V., the Netherlands
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: all participants were included in the efficacy analysis. Placebo: (25−24) * 100/25 = 4% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is low in both trial arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Santinga 1994.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 70.3 n: 93 Male: 33 Female: 60 Total cholesterol: 272.2 mg/dL (7.04 mmol/L) LDL cholesterol: 199.0 mg/dL (5.15 mmol/L) HDL cholesterol: 50.4 mg/dL (1.30 mmol/L) Triglycerides: 142.6 mg/dL (1.61 mmol/L) Placebo Age: 70.8 n: 48 Male: 14 Female: 34 Total cholesterol: 278.7 mg/dL (7.21 mmol/L) LDL cholesterol: 207.9 mg/dL (5.38 mmol/L) HDL cholesterol: 45.9 mg/dL (1.19 mmol/L) Triglycerides: 155.5 mg/dL (1.76 mmol/L) Included criteria: men and women 65 years of age or older with primary (type II) hypercholesterolemia. After ≥ 6 weeks on a low fat, low cholesterol diet, the mean of two consecutive determinations of fasting plasma LDL cholesterol concentration was required to be > the 95th percentile for the corresponding age and gender group of the US population (165 mg/dL for men or > 170 mg/dL for women), and the mean triglyceride concentration in the same specimens was to be < 250 mg/dL. Women receiving a stable dose of conjugated estrogens were eligible. Excluded criteria: homozygous familial hypercholesterolemia; type I, III, IV, or V hyperlipoproteinemia; significant endocrine, renal, hepatic, metabolic, or cardiovascular disease; those taking medication (e.g. corticosteroids, thiazide diuretics, beta‐adrenergic blockers) that might affect lipid levels Pretreatment: No significant differences in the demographic characteristics between the two treatment groups were detected by a chi‐square test of association or a linear model adjusted for numerical differences among investigators.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 16 weeks) Bedtime dosing Placebo (duration 16 weeks) Bedtime dosing
Outcomes	Total cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 4, 8, 12 Outcome type: continuous Triglycerides percentage change from baseline at week 4, 8, 12 Outcome type: continuous WDAEs up to week 16 Outcome type: dichotomous
Identification
Notes	Sponsorship source: This study was supported by a grant from the Bristol Myers Squibb Company, Princeton, New Jersey. Country: USA Setting: medical clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	High risk	This study was supported by a grant from the Bristol Myers Squibb Company, Princeton, New Jersey.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Pravastatin: (94‐93) * 100/94 = 1.1% were not included in the efficacy analysis. Placebo: all participants were included in the efficacy analysis. Absolute rate of loss to follow‐up is low in the two trial arms.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Saxenhofer 1990.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 5 mg BID Age: 48 Males: 6 Females: 4 n: 10 Total cholesterol: 9.4 mmol/L (363 mg/dL) LDL cholesterol: 6.4 mmol/L (247 mg/dL) HDL cholesterol: 1.3 mmol/L (50.3 mg/dL) Triglycerides: 2.1 mmol/L (186 mg/dL) Pravastatin 10 mg BID Age: 54 Males: 9 Females: 1 n: 10 Total cholesterol: 9.3 mmol/L (360 mg/dL) LDL cholesterol: 6.4 mmol/L (247 mg/dL) HDL cholesterol: 1.2 mmol/L (46.4 mg/dL) Triglycerides: 2.3 mmol/L (204 mg/dL) Placebo Age: 52 Males: 6 Females: 4 n: 10 Total cholesterol: 8.1 mmol/L (313 mg/dL) LDL cholesterol: 5.6 mmol/L (217 mg/dL) HDL cholesterol: 1.2 mmol/L (46.4 mg/dL) Triglycerides: 2.0 mmol/L (177 mg/dL) Included criteria: Thirty patients with hyperlipoproteinemia type IIa (N = 22) and type llb (N = 8) were studied. Patients who, after ≥ 4 weeks on a low‐fat, low‐cholesterol diet (30% fat, unsaturated/saturated ratio = 2/1; cholesterol < 300 mg daily), had two consecutive fasting total plasma cholesterol values > 6.8 mmol/L and plasma total triglyceride < 4 mmol/L were included. Excluded criteria: homozygous familial hypercholesterolemia; hypo‐ or hyperthyroidism; nephrotic syndrome or other renal disease; impaired hepatic function; diabetes mellitus; other causes of secondary hypercholesterolemia; symptomatic heart disease; myocardial infarction within the preceding 6 months; history of alcohol or drug abuse; concomitant treatment with corticosteroids, estrogens, androgens, quinidine, anticoagulants (antiplatelet drugs were permitted), theophylline, barbiturates, or regular use of antacids; premenopausal women, unless they were surgically sterilized Pretreatment: There was no significant difference between the three dose groups with respect to baseline lipids. Group II had a different male to female ratio.
Interventions	Intervention characteristics Pravastatin 5 mg twice‐daily (duration 16 weeks) Pravastatin 10 mg twice‐daily (duration 16 weeks) Placebo (duration 16 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: Source of funding was not reported. Country: Switzerland Setting: outpatient clinic
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blind placebo‐controlled trial, placebo being supplied as tablets of identical appearance
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAE outcome was not reported.
Other bias	Unclear risk	Source of funding was not reported.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Smit 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 50 n: 13 Male: 7 Female: 6 Total cholesterol: 5.9 mmol/L (228 mg/dL) LDL cholesterol: 3.9 mmol/L (151 mg/dL) HDL cholesterol: 1.16 mmol/L (44.9 mg/dL) Triglycerides: 1.76 mmol/L (156 mg/dL) Placebo Age: 49 n: 14 Male: 11 Female: 3 Total cholesterol: 6.1 mmol/L (236 mg/dL) LDL cholesterol: 4.1 mmol/L (159 mg/dL) HDL cholesterol: 1.21 mmol/L (46.8 mg/dL) Triglycerides: 1.69 mmol/L (150 mg/dL) Included criteria: Thirty‐three patients (22 women, 11 men), age 51 with symptomatic gallstone disease were included. All participants had one or more gallbladder stones at ultrasonographic examination but no abnormalities of liver and bile ducts. Liver biochemistry was normal. All participants had radiolucent gallstones in a gallbladder opacifying on oral cholecystography, performed after oral contrast (3 g Cholebrine; Dagra, Diemen, the Netherlands). Gallstones occupied < 30% of fasting gallbladder volume. The participants had no previous operations of the gastrointestinal tract or concomitant diseases, nor did they use any potentially interfering medication. Serum cholesterol levels were ≥ 5.0 mmol/L. Excluded criteria: not reported Pretreatment: Relevant baseline characteristics of the pravastatin group (13 participants) and placebo group (14 participants) were comparable.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 3 weeks) Placebo (duration 3 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 3 Outcome type: continuous LDL cholesterol percentage change from baseline at week 3 Outcome type: continuous HDL cholesterol percentage change from baseline at week 3 Outcome type: continuous Triglycerides percentage change from baseline at week 3 Outcome type: continuous
Identification
Notes	Sponsorship source: Bristol Myers Squibb, Woerden, the Netherlands (for supply of study drug and placebo) Country: the Netherlands Setting: hospital
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Pravastatin 40 mg/day: (16−13) * 100/16 = 18.75% were not included in the efficacy analysis. Placebo: (17−14) * 100/17 = 17.6% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is high in the two trial arms.

Spitalewitz 1993.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day n: 13 Male: 7 Female: 6 Age (years): 48 Total cholesterol: 301 mg/dL (7.78 mmol/L) LDL cholesterol: 222 mg/dL (5.74 mmol/L) HDL cholesterol: 44 mg/dL (1.14 mmol/L) Triglycerides: 221 mg/dL (2.5 mmol/L) Placebo n: 8 Male: 5 Female: 3 Age (years): 55 Total cholesterol: 281 mg/dL (7.27 mmol/L) LDL cholesterol: 208 mg/dL (5.38 mmol/L) HDL cholesterol: 43 mg/dL (1.11 mmol/L) Triglycerides: 182 mg/dL (2.05 mmol/L) Included criteria: Patients between the ages of 18 years and 70 years who had nephrotic syndrome (defined for this study as a urinary protein excretion ≥ 2.0 g/24 hours and a serum albumin of < 3.5 g/dL) and the mean of two plasma total cholesterol levels (obtained 1 week apart) above the 75th percentile for the US population by age and gender and > 220 mg/dL and patients on thiazide diuretics, beta‐adrenergic blockers, and other drugs known to affect blood lipids were included only if the drug and dose had been constant for 8 weeks prior to the study and was expected to be constant throughout the study period. Excluded criteria: patients with a serum creatinine ≥ 5.0 mg/dL, poorly controlled diabetes mellitus (fasting glucose > 200 mg/dL), taking > 60 mg of prednisone (or its equivalent) or other immunosuppressive agents, thyroid disease, primary hyperlipidemia, significant liver disease, excessive obesity (> 40% above ideal body weight), or excessive ethanol consumption Pretreatment: There were no renal diagnostic differences between the two treatment groups.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 24 weeks) Placebo (duration 24 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 4, 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4, 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 4, 8 Outcome type: continuous Triglycerides percentage change from baseline at week 4, 8 Outcome type: continuous
Identification
Notes	Sponsorship source: supported in part by a grant from Bristol Myers Squibb, Co. Inc. Country: USA Setting: hypertension and renal clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation was not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAE outcome was not reported.
Other bias	Unclear risk	Supported in part by a grant from Bristol Myers Squibb, Co. Inc.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Vanhanen 1995.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age: 53 n: 9 Male: 5 Female: 4 Total cholesterol: 7.11 mmol/L (275 mg/dL) LDL cholesterol: 4.44 mmol/L (172 mg/dL) HDL cholesterol: 1.22 mmol/L (47.2 mg/dL) Triglycerides: 2.09 mmol/L (185 mg/dL) Placebo Age: 57 n: 9 Male: 5 Female: 4 Total cholesterol: 7.43 mmol/L (287 mg/dL) LDL cholesterol: 4.75 mmol/L (184 mg/dL) HDL cholesterol: 1.16 mmol/L (44.9 mg/dL) Triglycerides: 2.20 mmol/L (195 mg/dL) Included criteria: Hyperlipidemic participants were advised to use a low cholesterol (< 300 mg/day) and low fat diet and placebo for 4 weeks. Excluded criteria: not reported Pretreatment: not reported
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 16 weeks) Placebo (duration 16 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	Sponsorship source: not reported Country: Finland Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	No WDAE outcome reported
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

Vincenzi 2014.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age (years): 42.6 n: 30 Males: 8 Females: 22 Total cholesterol: 179.63 mg/dL (4.65 mmol/L) LDL cholesterol: 106.22 mg/dL (2.75 mmol/L) HDL cholesterol: 48.1 mg/dL (1.24 mmol/L) Triglycerides: 174.4 mg/dL (1.97 mmol/L) Placebo Age (years): 44.5 n: 30 Males: 14 Females: 16 Total cholesterol: 178.2 mg/dL (4.61 mmol/L) LDL cholesterol: 104.03 mg/dL (2.69 mmol/L) HDL cholesterol: 45.36 mg/dL (1.17 mmol/L) Triglycerides: 145.46 mg/dL (1.64 mmol/L) Included criteria: 81 male and female outpatients between the ages of 18 years and 68 years with the diagnosis of schizophrenia, any subtype, or schizoaffective disorder, any subtype Excluded criteria: inability to provide informed consent; participation in other research studies; unstable psychiatric illness; current alcohol or substance abuse; current treatment with insulin; pregnancy; untreated thyroid disease; significant medical illness, including severe cardiovascular, hepatic, or renal disease (serum creatinine ≥ 1.5 mg/dL); anemia (hemoglobin < 11.0 mg/dL), history of severe head injury; or untreated muscle disease. Patients treated with the following medications known to affect glucose tolerance were also excluded – anti‐inflammatory drugs (including aspirin and ibuprofen), thiazide diuretics, agents that induce weight loss, or St. John's Wort. Similarly, patients treated with colchicine, azole antifungals, macrolide antibiotics, HIV protease inhibitors, or a known hypersensitivity to pravastatin or to any of its components were excluded. Pretreatment: There were no statistically significant differences between the placebo and pravastatin groups by age, gender, race, ethnicity, diagnosis, or marital status at diagnosis.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6, 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6, 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6, 12 Outcome type: continuous Triglycerides percentage change from baseline at week 6, 12 Outcome type: continuous WDAEs up to 12 weeks Outcome type: dichotomous
Identification
Notes	Sponsorship source: the Stanley Medical Research Institute Country: USA Setting: outpatient
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	Low risk	This work was supported by the Stanley Medical Research Institute (Grant number 09T‐1296).
Incomplete outcome data (attrition bias) All outcomes	High risk	Pravastatin: (30−24) * 100/30 = 20% were not included in the efficacy analysis. Placebo: (30−25) * 100/30 = 17% were not included in the efficacy analysis. Absolute rate of loss to follow‐up was high in both arms of the trial.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Wiklund 1990.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg BID = 20 mg/day n: 40 Age: 51.6 Male: 18 Female: 22 Total cholesterol: 9.9 mmol/L (383 mg/dL) LDL cholesterol: 8.1 mmol/L (313.2 mg/dL) HDL cholesterol: 1.23 mmol/L (47.6 mg/dL) Triglycerides: 1.74 mmol/L (154 mg/dL) Placebo n: 40 Age: 50.1 Male: 22 Female: 18 Total cholesterol: 9.99 mmol/L (386 mg/dL) LDL cholesterol: 8.2 mmol/L (317.1 mg/dL) HDL cholesterol: 1.24 mmol/L (48 mg/dL) Triglycerides: 1.67 mmol/L (148 mg/dL) Included criteria: 120 patients with heterozygous familial hypercholesterolemia Excluded criteria: Premenopausal women and patients with diabetes, hepatic dysfunction, severe hypertension, or excessive obesity were excluded from the study. Pretreatment: none
Interventions	Intervention characteristics Pravastatin 10 mg twice‐daily = 20 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 6 Outcome type: continuous LDL cholesterol percentage change from baseline at week 6 Outcome type: continuous HDL cholesterol percentage change from baseline at week 6 Outcome type: continuous Triglycerides percentage change from baseline at week 6 Outcome type: continuous WDAEs up to week 6 Outcome type: dichotomous
Identification
Notes	Sponsorship source: Swedish Heart‐Lung Foundation, the Swedish Medical Research Council (projects 4531, 7137, and 7142), King Gustaf V and Queen Victoria Foundation, a grant from E. R. Squibb and Sons Country: Sweden Setting: lipid clinics
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blind fashion; lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	Unclear risk	This study was supported by the Swedish‐Heart‐Lung Foundation, the Swedish medical Research Council (projects 4531, 7137, and 7142), the King Gustaf V and Queen Victoria Foundation, and by a grant from E. R. Squibb and Sons.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Wiklund 1996.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 40 mg/day Age (years): 52.5 n: 64 Males: 42 Females: 22 Total cholesterol: 7.24 mmol/L (280 mg/dL) LDL cholesterol: 5.19 mmol/L (200.7 mg/dL) HDL cholesterol: 1.25 mmol/L (48.3 mg/dL) Triglycerides: 1.80 mmol/L (159.4 mg/dL) Placebo Age (years): 53.1 n: 69 Males: 42 Females: 27 Total cholesterol: 7.24 mmol/L (280 mg/dL) LDL cholesterol: 5.12 mmol/L (198.0 mg/dL) HDL cholesterol: 1.20 mmol/L (46.4 mg/dL) Triglycerides: 1.81 mmol/L (160.3 mg/dL) Included criteria: 290 patients from 21 years to 70 years of age. Enrollment was limited to ambulatory men and women with no childbearing potential, primary hypercholesterolemia (total plasma cholesterol of ≥ 6.0 mmol/L or ≥ 90th percentile for age and sex and triglyceride concentrations < 4.0 mmol/L) despite dietary intervention. Excluded criteria: Patients with homozygous familial hypercholesterolemia or type I, III, IV, or V hyperlipoproteinemia were excluded, as were patients with significant cardiovascular, renal, hepatic, metabolic, or gastrointestinal disease or other diseases, except for CHD, expected to limit life span to < 5 years patients receiving treatment with corticosteroids, estrogens except as replacement therapy, androgens, quinidine, coumarin anticoagulants, theophylline, barbiturates, aluminum‐containing antacids, cyclosporine, or fish oil preparations. Pretreatment: Baseline characteristics and medical histories of cardiovascular events did not differ significantly.
Interventions	Intervention characteristics Pravastatin 40 mg/day (duration 12 weeks) Placebo (duration 12 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 12 Outcome type: continuous LDL cholesterol percentage change from baseline at week 12 Outcome type: continuous HDL cholesterol percentage change from baseline at week 12 Outcome type: continuous Triglycerides percentage change from baseline at week 12 Outcome type: continuous
Identification
Notes	Sponsorship source: Swedish Medical Research Council (45311 7137.8708), the Swedish Heart and Lung Foundation, the King Gustav V and Queen Victoria Foundation, and the Bristol Myers Squibb Company, Princeton, New Jersey Country: Sweden and Finland Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by method of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAE outcome was not reported.
Other bias	Unclear risk	Grants from the Swedish Heart and Lung Foundation, the Swedish Medical Research Council (grants no. 4531, 7137) and the Bristol Myers Squibb Company, Princeton, New Jersey
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	(290‐266)/290 * 100 = 8.3% were not included in the efficacy analysis overall treatments. The rate of loss to follow‐up was moderate and not reported separately for each arm of the trial.

Yasushi 1987.

Study characteristics
Methods	Study design: randomized, double‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 10 mg/day n: 41 Males: 9 Females: 32 Total cholesterol: 286 mg/dL (7.40 mmol/L) LDL cholesterol: 202 mg/dL (5.22 mmol/L) HDL cholesterol: 51 mg/dL (1.32 mmol/L) Triglycerides: 173 mg/dL (1.95 mmol/L) Pravastatin 20 mg/day n: 44 Males: 14 Females: 30 Total cholesterol: 276 mg/dL (7.14 mmol/L) LDL cholesterol: 189 mg/dL (4.89 mmol/L) HDL cholesterol: 54 mg/dL (1.40 mmol/L) Triglycerides: 163 mg/dL (1.84 mmol/L) Placebo n: 23 Males: 7 Females: 16 Total cholesterol: 286 mg/dL (7.40 mmol/L) LDL cholesterol: 203 mg/dL (5.25 mmol/L) HDL cholesterol: 49 mg/dL (1.27 mmol/L) Triglycerides: 170 mg/dL (1.92 mmol/L) Included criteria: One‐hundred and eighteen hypercholesterolemic patients aged 30 to 80 were given pravastatin at the dose of 0, 10, and 20 mg under double‐blind conditions. They maintained their usual diet and activities during the study, but alcohol was prohibited. Excluded criteria: none reported Pretreatment: There was no significant difference between the participants in the three groups with respect to baseline serum cholesterol and triglyceride levels as well as age and sex.
Interventions	Intervention characteristics Pravastatin 10 mg/day (duration 8 weeks) Pravastatin 20 mg/day (duration 8 weeks) Placebo (duration 8 weeks)
Outcomes	Total cholesterol percentage change from baseline at week 4, 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 4, 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 4, 8 Outcome type: continuous Triglycerides percentage change from baseline at week 4, 8 Outcome type: continuous WDAEs up to week 8 Outcome type: dichotomous
Identification
Notes	Sponsorship source: not reported Country: Japan Setting: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by methods of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	Low risk	WDAE outcome was reported.
Other bias	Unclear risk	Source of funding not reported
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Placebo: (23−21) * 100/23 = 8.7% were not included in the efficacy analysis. Pravastatin 10 mg/day: (40−39) * 100/40 = 2.5% were not included in the efficacy analysis. Pravastatin 20 mg/day: (44−39) * 100/44 = 11.4% were not included in the efficacy analysis. Absolute rate of loss to follow‐up is moderate in two trial arms and low in another.
Blinding of outcome assessment (detection bias) WDAE	Unclear risk	Method of blinding of outcome assessment for WDAEs was not reported.

Zambon 1994.

Study characteristics
Methods	Study design: randomized, single‐blind, placebo‐controlled trial
Participants	Baseline characteristics Pravastatin 20 mg/day Age: 52 n: 12 Male: 9 Female: 3 Total cholesterol: 8.39 mmol/L (324 mg/dL) LDL cholesterol: 6.01 mmol/L (232 mg/dL) HDL cholesterol: 0.98 mmol/L (37.9 mg/dL) Triglycerides: 3.29 mmol/L (291 mg/dL) Placebo Age: 57 n: 12 Male: 6 Female: 6 Total cholesterol: 6.85 mmol/L (265 mg/dL) LDL cholesterol: 4.88 mmol/L (189 mg/dL) HDL cholesterol: 1.06 mmol/L (41.0 mg/dL) Triglycerides: 2.84 mmol/L (252 mg/dL) Included criteria: Twenty‐four patients with combined hyperlipidemia were included. All lipid‐lowering drugs and other drugs known to affect lipid metabolism were discontinued ≥ 4 weeks before the start of the study. None of the patients had diabetes mellitus or hepatic, thyroid, or renal disease; none had a stroke, myocardial infarction, or major surgery during the previous 6 months. Pravastatin was given nightly. Excluded criteria: none Pretreatment: Serum total cholesterol, LDL cholesterol, and triglycerides baseline values were greater in the pravastatin group compared to the placebo group.
Interventions	Intervention characteristics Pravastatin 20 mg/day (duration 16 weeks) Nightly Placebo (duration 16 weeks) Nightly
Outcomes	Total cholesterol percentage change from baseline at week 8 Outcome type: continuous LDL cholesterol percentage change from baseline at week 8 Outcome type: continuous HDL cholesterol percentage change from baseline at week 8 Outcome type: continuous Triglycerides percentage change from baseline at week 8 Outcome type: continuous
Identification
Notes	Sponsorship source: CNR‐Italy, PF INV Country: Italy Setting: outpatient
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Lipid parameter measurements unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)	Low risk	Lipid parameters were measured in a remote laboratory.
Selective reporting (reporting bias)	Low risk	LDL cholesterol outcome was reported.
Selective Reporting (reporting bias) WDAE	High risk	WDAE outcome was not reported.
Other bias	Unclear risk	This work was supported in part by CNR‐Italy, PF INV.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants were included in the efficacy analysis.

ACE: angiotensin‐converting enzyme; ACS: acute coronary syndrome; ALAT: alanine transaminase ; ASP: aspartate aminotransferase; CABG: coronary artery bypass grafting; CAD: coronary artery disease; CHD: coronary heart disease; CNS: central nervous system; DBP: diastolic blood pressure; GFR: glomerular filtration rate; HbA1c: glycated hemoglobin; HDL: high‐density lipoprotein; HMG‐CoA: 3‐hydroxy‐3‐methyl‐glutaryl coenzyme A; HR: heart rate; LDL: low‐density lipoprotein; PTCA: percutaneous transluminal coronary angioplasty; QCA: quantitative coronary arteriography; SBP: systolic blood pressure; SD: standard deviation; TC: total cholesterol; UA: unstable angina; ULN: upper limit of normal; WDAE: withdrawal due to adverse effects.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Aberg 2017	Placebo‐controlled trial but no placebo data reported; use of retroviral drugs is a confounding factor
Aguilar Salinas 1997	Pharmacokinetic study of 80 hours
Baker 2012	Confounding factors
Beigel 1993	Dose not specified
Bennett 2004	No placebo group for the pravastatin portion of the trial
Bonnet 2007	Confounding due to protease inhibitors
Castelao 1995	Confounding factors
Cobbaert 1992	Dosing was not constant
Cueto Manzano 2013	Lipid data were combined for all cross‐over periods.
Cutler 1995	Lipid data were combined for both cross‐over periods.
Dangas 1999a	Dose not specified
Dangas 1999b	Dose not specified
Dangas 2000	Dose not specified
Dobs 1995	Ineligible measurement
Eckernas 1993	Incomplete washout
Ehrenberg 1999	Lipid data were combined for all cross‐over periods.
EUCTR2004 003235 31 AT	Trial not completed
EUCTR2006 000515 15 GR 2006	No washout period of all previous lipid‐lowering drugs
Ferrari 1993	Incomplete washout
Fogari 1992	Number of participants was not reported for the placebo and pravastatin groups separately.
Gengo 1995	Lipid data were combined for both cross‐over periods.
Glorioso 1999	Number of participants was not reported for the placebo and pravastatin groups separately.
Harrison 1994	Cross‐over trial cannot calculate change from baseline for separate placebo and pravastatin groups.
Hunninghake 1990a	Variable N
Hunninghake 1990b	Variable N
Hunninghake 1993	Lipid data combined for all doses
Imai 1999	Lipid data combined for all doses
Isaacsohn 2001	No placebo group for the pravastatin portion of the trial
Ismail 1990	Lipid data combined for all pravastatin doses
Jay 1990	More participants with lipid data than randomized
Kesteloot 1992	Dosing not constant
Kesteloot 1997	Dosing not constant
Klausen 1993	Confounding factors
Knopp 1996	Lipid data were combined for all cross‐over periods.
Koeijvoets 2005	No placebo lipid change data
Kool 1995	Lipid data were combined for both cross‐over periods.
Kostis 1992	No parallel placebo group
Kostis 1994	Ineligible measurement
Lye 1995	Confounding factors
Mallon 2006	Confounding due to protease inhibitors
Mostaza 2000	Lipid data were combined for all cross‐over periods.
Nakamura 1996	No placebo lipid change data
NCT00177580 2005	Trial not completed
NCT00221754 2005	Confounding due to protease inhibitors
NCT00227500 2005	Confounding due to protease inhibitors
NCT00245388 2005	Trial not completed
Owens 1991	Dose not specific
Paradisi 2012	Median change
Partinen 1994	Lipid data were combined for both cross‐over periods.
Reid 2005	Median change
Rosengarten 2007	Sequential dosing
Rosenson 2005	Combined data for all statins
Ruscica 2014	No placebo group for the pravastatin portion of the trial
Santos 2000	No washout period of all previous lipid‐lowering drugs
Schaefer 2005	Sequential dosing
Solheim 2001	Median change
Stein 2004	Confounding due to protease inhibitors
Straznicky 1995	Lipid data were combined for all cross‐over periods.
Thompson 2002	Combined data for all statins
Thompson 2004	Confounding factors
Torrecilla 2009	Lipid data were combined for all cross‐over periods.
Toto 2000	No placebo lipid change data
Vega 1990	No placebo lipid change data
Yoshimura 1994	Confounding factors
Zanchetti 2004	Treatment period too long
Zhang 1995	Lipid data were combined for all cross‐over periods.

Differences between protocol and review

We carried out a sensitivity analysis for total cholesterol and LDL cholesterol to determine if imputing standard deviations has an important impact on the review. This was added to the review but was not mentioned in the protocol.

All sensitivity analyses mentioned in the protocol are analyzed as subgroup analyses in the review.

Random‐effects meta‐analysis at a certain heterogeneity threshold was mentioned in the protocol but not in the review.

Confounding factors, inappropriate dosing, no appropriate placebo data, combined cross‐over data, inappropriate outcomes (medians), inadequate dietary baseline period, unknown n are reasons we will exclude. This sentence was added to the review but is not mentioned in the protocol.

Subgroup analyses were conducted for LDL cholesterol and not secondary outcomes for which the heterogeneity was high (HDL and total cholesterol) to analyze the subgroups based on the following factors. This sentence was added to the review but is not mentioned in the protocol.

Additional subgroup analysis for Bristol‐Myers Squibb versus non‐Bristol‐Myers Squibb funded trials was conducted that was not listed in the protocol.

Contributions of authors

Review authors SPA, NA, and JMW contributed to the design of the protocol. SPA, NA, ST contributed to trial eligibility. SPA and NA extracted the data. JMW and SPA wrote the discussion and overall completeness and applicability of evidence sections. SPA wrote the Agreements and disagreements with other studies or reviews section.

Sources of support

Internal sources

• UBC, Canada

  Department of Anesthesiology, Pharmacology & Therapeutics, University of BC, Canada.
Office space

External sources

• Therapeutics Initiative, Canada

  BC Ministry of Health grant to the Therapeutics Initiative, Canada.
Salary support

Declarations of interest

Stephen P Adams has no known financial, political, personal, intellectual, or religious interests that could appear to influence the work reported in this systematic review.

Nima Alaeiilkhchi has no known financial, political, personal, intellectual, or religious interests that could appear to influence the work reported in this systematic review.

Sara Tasnim has no known financial, political, personal, intellectual, or religious interests that could appear to influence the work reported in this systematic review.

James M Wright has no known financial, political, personal, intellectual, or religious interests that could appear to influence the work reported in this systematic review. He is the Co‐ordinating Editor of the Cochrane Hypertension Group. He has played no role in the editorial process for this review.

These authors should be considered joint first author

New