Statins for the primary prevention of venous thromboembolism

Abstract

Background

Venous thromboembolism (VTE) involves the formation of a blood clot in a vein, and includes deep venous thrombosis (DVT) or pulmonary embolism (PE). The annual incidence for VTE varies from 0.75 to 2.69 per 1000 individuals, with about 40 million people worldwide impacted by VTE. Statins, 3‐hydroxy‐3‐methylglutaryl (HMG)‐coenzyme A (CoA) reductase inhibitors, inhibit cholesterol biosynthesis and display several vascular‐protective effects, including antithrombotic properties. However, the potential role of statins in the primary prevention of VTE is still not clear.

Objectives

To evaluate the benefits and risks of statins in preventing venous thromboembolism (VTE) in individuals with no prior history of VTE.

Search methods

We used standard Cochrane search methods. The search was last updated on 13 March 2023.

Selection criteria

We included randomized controlled trials (RCTs) comparing statins with any control intervention (including placebo and usual care) in healthy individuals or participants with conditions other than VTE. There were no restrictions on the dose, duration, route, or timing of statins.

Data collection and analysis

We used standard Cochrane methods. Our primary outcomes were VTE, DVT, and PE. Our secondary outcomes were serious adverse events, adverse events, and mortality. We used the trial sequential analysis (TSA) method to judge whether the evidence was sufficient, and we used the GRADE approach to assess the certainty of the evidence for each outcome.

Main results

We included 27 RCTs involving 122,601 adults (aged 18 years and above) who were healthy, had various medical conditions (e.g. hypercholesterolemia), or were at risk for cardiovascular disease. Both males and females were included in all studies. Two studies focused solely on participants over 60 years of age. We deemed four studies to have a low risk of bias overall, while 19 were at high risk of bias, and four were unclear.

The 27 studies compared use of statins versus placebo or usual care in individuals who had never experienced VTE. The statins used in the studies were atorvastatin, rosuvastatin, pravastatin, lovastatin, fluvastatin, and simvastatin. Twenty‐three studies followed up participants for over a year, with six of those extending follow‐ups for over five years. Twenty‐five studies were based in hospitals, and 24 studies were funded by industry. Only one study used VTE as a primary endpoint.

The median incidence of VTE in the statins group was 0.72% (ranging from 0% to 10.53%), and in the control group it was 0.89% (ranging from 0% to 6.83%). Our pooled analysis of the 27 studies showed that, relative to control groups, statins may slightly reduce the overall incidence of VTE (odds ratio (OR) 0.86, 95% confidence intervals (CI) 0.76 to 0.98; 27 studies, 122,601 participants; low‐certainty evidence). Of the statins we evaluated, only rosuvastatin seemed to be associated with a reduced incidence of VTE, albeit the reduction in incidence was very small. The evidence did not clearly indicate a difference between groups in the incidence of DVT (OR 0.70, 95% CI 0.41 to 1.18; six studies, 40,305 participants; low‐certainty evidence), PE (OR 0.83, 95% CI 0.46 to 1.52; five studies, 28,427 participants; low‐certainty evidence), or myopathy (OR 1.10, 95% CI 0.83 to 1.45; 10 studies, 75,551 participants; low‐certainty evidence). Nonetheless, statin use might slightly decrease the incidence of any serious adverse event (OR 0.95, 95% CI 0.91 to 0.99; 13 studies, 67,020 participants; low‐certainty evidence) and any death (OR 0.90, 95% CI 0.86 to 0.95; 24 studies, 116,761 participants; low‐certainty evidence), compared to control.

Authors' conclusions

Using statins for the primary prevention of VTE may slightly reduce the incidence of VTE and all‐cause mortality. However, this effect is likely too weak to be considered significant. Statin use may not decrease the occurrence of DVT and PE. The current evidence is insufficient to draw strong conclusions because of the risk of bias in the studies, imprecision in the effect estimates, and potential publication bias. More evidence from well conducted and fully reported RCTs is needed to assess the preventive effects of different types of statins, as well as the effects of different dosages and treatment durations in various populations.

Plain language summary

Does taking statins prevent blood clots from forming in the veins of people who have not previously had a blood clot?

Key messages

‐ Statin use for the primary prevention of venous thromboembolism (VTE; a condition in which a blood clot, or embolism, forms in the vein) may slightly decrease the incidence of VTE and death from any cause, but the reduction may be too small to be important.

‐ Statins may make no difference to the likelihood of experiencing deep vein thrombosis (DVT; a blood clot in the lower legs), pulmonary embolism (PE; a blood clot in the lungs), or myopathy (a condition affecting the skeletal muscles).

‐ Available evidence was limited, and we are not certain about its reliability. Future prospective studies should be well‐planned and conducted. They should involve a large number of people and should be conducted over a reasonable duration of at least a year.

What is venous thromboembolism (VTE)?

VTE is a condition in which a blood clot (embolism) forms in the veins. Blood flow through the affected vein is reduced by the clot, causing swelling and pain. VTE most commonly occurs in the 'deep veins' in the lower legs, thighs, or pelvis, and is referred to as DVT. If the blood clot breaks loose and travels to the lungs, it is known as PE. Annually, from 100,000 individuals, 57 are diagnosed with VTE, 35 with DVT, and 21 with PE, worldwide.

Statins, which include atorvastatin, fluvastatin, lovastatin, pitavastatin, pravastatin, rosuvastatin, and simvastatin, are the most frequently used cholesterol‐lowering drugs. These statins potentially lower cholesterol levels and consequently may have the potential to decrease the incidence of VTE.

What did we want to find out?

We aimed to discover the benefits and risks of statins for the primary prevention of VTE (first diagnosis) in individuals with no prior history of VTE.

What did we do?

We searched for randomized controlled trials (RCTs) comparing statins use with placebo (i.e. fake or 'dummy' pill) or usual care (routine care such as changing diet), to assess if the statins made a difference in the number of individuals who developed a first‐time VTE or experienced any side effects.

RCTs are experimental studies where participants are randomly assigned to two or more treatment groups. This random method of allocating people to groups helps reduce the risk of bias by ensuring that the groups are similar, and that neither the investigators nor the participants know who is in which group.

What did we find?

Our review encompasses 27 RCTs comprising 122,601 adults (over 18 years of age); two studies focused on participants older than 60 years. While one study involved a healthy population, the participants in the remaining 26 studies had various diseases. All studies involved both male and female participants. Two studies engaged participants from primary care centers, while the remaining 25 incorporated participants from hospitals.

The statins included were atorvastatin, rosuvastatin, pravastatin, lovastatin, fluvastatin, and simvastatin. Most of the studies (23 of the 27) continued their follow‐up with participants for more than a year. Similarly, most of the studies (24 of the 27) received funding from commercial companies.

We pooled the results of individual studies and found that statins may slightly reduce the number of VTE cases. The number of VTE cases prevented by statins was small. However, statins may not reduce the number of blood clots in the leg or lung, identified as DVT and PE, respectively. We did not find any evidence of a difference in the quantity of less severe side effects, such as myopathy. There is a possibility that statins might reduce the incidence of death from any cause or reduce the number of severe side effects.

What are the limitations of the evidence?

We are not very confident in the results due to concerns about the methods used in some of the studies. The low number of blood clots makes it difficult to detect effects. Differences in the general health of participants, as well as the dosage and types of statins used across studies, introduce further complexity. The exclusion of studies that did not publish the incidence of VTE could mean that relevant studies have been missed, which might affect the conclusions of this review.

How up‐to‐date is this evidence?

The evidence is current up to 13 March 2023.

Summary of findings

Summary of findings 1. Statins versus placebo or usual care for the primary prevention of venous thromboembolism.

Population: people without previous venous thromboembolism Comparison: statins versus placebo or usual care Settings: community or hospital Intervention: statins Control: placebo or usual care Follow‐up: 90 days to 6.1 years
Outcomes#	Anticipated absolute effects* 95% CI		Relative effect (95% CI)	Number of participants (RCTs)	Certainty of the evidence (GRADE)	Comments
	Assumed risk with placebo	Corresponding risk with statin
All cases of venous thromboembolism Follow‐up time point: 9 weeks to 6.1 years	9 per 1000	8 per 1000 (7 to 9)	OR 0.86 (0.75 to 0.99)	122,601 (27 RCTs)	⊕⊕⊝⊝ Lowa,b
Deep vein thrombosis Follow‐up time point: 90 days to 5.6 years	5 per 1000	4 per 1000 (2 to 6)	OR 0.70 (0.41 to 1.18)	40,305 (6 RCTs)	⊕⊕⊝⊝ Lowc,d
Pulmonary embolism Follow‐up time point: 1 to 5.1 years	3 per 1000	2 per 1000 (1 to 4)	OR 0.83 (0.46 to 1.52)	28,427 (5 RCTs)	⊕⊕⊝⊝ Lowa,c
Any serious adverse event Follow‐up time point: 1 to 5.2 years	214 per 1000	207 per 1000 (200 to 216)	OR 0.95 (0.91 to 0.99)	67,020 (13 RCTs)	⊕⊕⊝⊝ Lowa,b
Myopathy Follow‐up time point: 1.5 to 6.1 years	2 per 1000	3 per 1000 (2 to 3)	OR 1.10 (0.83 to 1.45)	75,551 (10 RCTs)	⊕⊕⊝⊝ Lowa,c
Mortality ‐ all causes Follow‐up time point: 90 days to 6.1 years	108 per 1000	98 per 1000 (94 to 102)	OR 0.90 (0.86 to 0.95)	116,761 (24 RCTs)	⊕⊕⊝⊝ Lowa,b
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). #All outcomes were reported at the longest follow‐up time point. CI: confidence interval; OR: odds ratio; RCTs: randomized controlled trials
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 the effect

^aWe downgraded the certainty of the evidence by one level as some of the included studies were at high risk of bias.

^bWe downgraded the certainty of the evidence by one level as there was possible publication bias.

^cWe downgraded the certainty of the evidence by one level for imprecision as the total number of events is less than 300.

^dWe downgraded the certainty of the evidence by one level for inconsistency as there was high statistical heterogeneity across studies.

Background

Description of the condition

Venous thromboembolism (VTE) is a condition in which a blood clot (thrombus) forms in an intact vein when red blood cells, fibrin, and, to a lesser extent, platelets and leucocytes (white blood cells), combine. The clot reduces blood flow through the affected vein, leading to swelling and pain. VTE most commonly occurs in the ‘deep veins’ of the lower legs, thighs, or pelvis. Hence, it is usually referred to as deep vein thrombosis (DVT). An embolism is formed if a part or all of the blood clot breaks off from the site where it was created and travels through the venous system. If the clot lodges in the lungs, it results in a pulmonary embolism (PE). The risk factors for VTE are complex, including hospitalisation for active cancer, surgery or acute illness, neurological disease with leg paresis, nursing‐home confinement, trauma or fracture, superficial vein thrombosis, non‐O blood groups, factor V Leiden mutation, advanced age, long‐haul travel, and in women, pregnancy and puerperium, oral contraception, and hormone therapy (Heit 2015). The risk factors for VTE are listed in Table 2.

1. Risk factors for venous thromboembolism.

General	High‐risk clinical situations	Diseases associated with a prothrombotic state	Inherited thrombophilia
Older age Obesity Family history of venous thromboembolism Oral contraceptive pill, hormone replacement, tamoxifen Venous insufficiency or varicose veins	Surgery (especially hip and knee surgery or major surgery for malignancy) Pregnancy/puerperium Acute medical illness Congestive cardiac and respiratory failure Trauma Central venous catheter Immobility, paresis Malignancy American Society of Anesthesiologists ≥ 3 Blood transfusion Cardiovascular disease Diabetes mellitus Hypertension Immobility Lung disease Peripheral artery disease	Myeloproliferative disorders Antiphospholipid syndrome Paroxysmal nocturnal haemoglobinuria Nephrotic syndrome Hyperviscosity syndrome Inflammatory bowel disease elevated D‐dimer	Factor V Leiden mutation Antithrombin, protein C, and protein S deficiency Prothrombin gene mutation (Factor II G20210A mutation)

VTE may occur in individuals with cancer or as a result of trauma, hospitalization, or surgery, which is referred to as provoked VTE. In contrast, unprovoked VTE is defined as an event that occurs without any known malignancy (diagnosed either before or within three months after the VTE), trauma, hospitalization, or surgery. Approximately two‐thirds of VTE episodes clinically manifest as DVT, and one‐third as isolated PE or PE coexisting with DVT (Oleksiuk‐Bójk 2023). Patients with untreated, symptomatic, proximal DVT (i.e. above the knee, affecting the femoral or iliofemoral veins) have a 50% likelihood of developing symptomatic PE within three months (Kacimi 2022; Moheimani 2011).

Studies have demonstrated that the annual incidence rates for VTE range from 0.57 to 2.69 per 1000 individuals (Alotaibi 2016; Arshad 2017; Heit 2016; Ho 2008; ISTH 2014; Lee 2017; Raju 2009; Tagalakis 2013), while the incidence rates for DVT vary from 0.35 to 1.00 per 1000 person‐years (number of individuals in the study multiplied by the time each individual is observed, typically in years), and for PE from 0.21 to 0.45 per 1000 person‐years (Alotaibi 2016; Arshad 2017; Ho 2008; Raju 2009; Tagalakis 2013) (Table 3). The prevalence of VTE is estimated to increase, for example, from 0.42% in 2013 to 0.57% by 2050 in the USA (Liu 2013), and from 0.33 in 2014 to 0.54 in 2018 per 1000 person‐years in Korea (Hwang 2022). VTE can cause complications like post‐thrombotic syndrome (PTS), PE, and death (Kacimi 2022; Moheimani 2011). It is the third most prevalent acute cardiovascular syndrome that can lead to severe morbidity and mortality (Cohen 2021). Reasonable estimates for the incremental cost of VTE are around USD $10,000 for admission to hospital for a VTE or around USD $17,000 for a PE in adjusted 2019 US dollars (Guntupalli 2023).

2. Incidence of venous thromboembolism, deep vein thrombosis, and pulmonary embolism.

Study	Region(s)	Venous thromboembolism (per 1000 persons)	Deep vein thrombosis (per 1000 persons)	Pulmonary embolism (per 1000 persons)
Alotaibi 2016	Canada	1.38	1	0.38
Arshad 2017	Norway	0.86	0.48	0.37
Heit 2016	Europe	1.04 to 1.83	‐	‐
Ho 2008	Western Australia	0.83	0.52	0.31
Hwang 2022	Korea	0.537	‐	‐
ISTH 2014	Western Europe, North America, Australia, and Southern Latin America	0.75 to 2.69	‐	‐
Lee 2017	Asia	0.11 to 0.88	‐	‐
Liu 2013	US, Europe, and Asia	‐	0.21 to 1.48	0.2 to 0.66
Raju 2009	Worldwide	0.57	0.35	0.21
Tagalakis 2013	Canada	1.22	0.78	0.45

Description of the intervention

The prevention of thrombosis can be achieved through pharmacological or physical means. Anti‐thrombotic drugs include low molecular weight heparin (LMWH), factor Xa inhibitor (fondaparinux), vitamin K antagonists (VKAs), and direct oral anticoagulants (DOACs) (Alpert 2001; Diuguid 2001; Li 2023; Song 2019). Mechanical prophylaxis, such as intermittent pneumatic compression (IPC) stockings or graduated compression stockings (GCS), is recommended for individuals with a higher‐than‐normal risk of bleeding or as an adjunct to the more efficacious pharmacological prophylaxis (Chapman 2009; Diuguid 2001). All are options in the primary prevention of VTE. The choice of which prophylaxis to use depends on an individual’s risk factors, the availability of recommended medication, and the clinical judgment of the treating doctor (Chapman 2009).

Statins, which are 3‐hydroxy‐3‐methylglutaryl (HMG)‐coenzyme A (CoA) reductase inhibitors (including atorvastatin, fluvastatin, lovastatin, pitavastatin, pravastatin, rosuvastatin, and simvastatin), are the most prevalent cholesterol‐lowering drugs available. Statins were introduced as a new group of cholesterol‐lowering drugs to provide a safe and effective means for patients with hypercholesterolemia to reduce their plasma cholesterol levels. The chemical structure of statins consists of two components: the pharmacophore, which is a dihydroxyheptanoic acid segment, and its moiety, which comprises a ring system with diverse substituents. Different statin groups have variations in their structure, which determines their water solubility and affects their absorption, distribution, metabolism, and excretion. Upon oral administration, all statins are effectively absorbed from the intestine, although extensive first‐pass metabolism occurs in the liver, subsequently reducing systemic bioavailability. Lovastatin, simvastatin, and pravastatin are derived from fungal metabolites, with elimination half‐lives ranging from one to three hours. Atorvastatin, rosuvastatin, fluvastatin, and pitavastatin, on the other hand, are entirely synthetic compounds. Their elimination half‐lives vary from one hour for fluvastatin to 19 hours for rosuvastatin. Lovastatin, simvastatin, atorvastatin, fluvastatin, and pitavastatin are relatively lipophilic compounds. These lipophilic statins are more likely to be metabolized by the cytochrome P(450) system, other than pitavastatin, which shows limited metabolism via this pathway. The liver and kidneys play vital roles in eliminating statins from the circulation system through the bile and, ultimately, into the feces.

Atorvastatin, lovastatin, and pravastatin are typically administered from 10 mg to 80 mg daily. Rosuvastatin is given from 5 mg to 40 mg daily, while fluvastatin can be administered from 20 mg to 80 mg each day. Similarly, simvastatin and rosuvastatin can be given from 5 mg to 80 mg daily. Statins are often recommended for adults aged 40 to 75 years who have no history of heart disease but are at risk due to factors such as diabetes, high blood pressure, or hyperlipidemia. Additionally, people with a known history of heart disease, stroke, or peripheral blood vessel disease, those with persistently high levels of blood LDL cholesterol (LDL‐C greater than 190 mg/dL or greater than 4.9 mmol/L), and those with a high risk of developing heart disease or stroke may benefit from statins. For participants with a five‐year risk of major vascular events lower than 10%, the estimated reduction in such events is around 11 per 1000 over five years for each 1.0 mmol/L reduction in LDL cholesterol (4.1% for statins or more intensive statin regimen versus 5.2% for control or less intensive regimen) (Mihaylova 2012).

However, statins may cause musculoskeletal symptoms, pose an increased risk of diabetes, and lead to higher rates of hemorrhagic stroke (Pinal‐Fernandez 2018). The most frequently reported side effects include muscle‐related issues, ranging from mild muscle aches (myalgia) to more severe conditions such as rhabdomyolysis, which can lead to kidney damage. These side effects can result in non‐compliance or discontinuation of therapy, impacting treatment outcomes. Statins can also affect liver function as they may cause elevated liver enzymes in some patients. Recent studies have indicated an association between statin use and an increased risk of developing type 2 diabetes, particularly in individuals already predisposed to the condition (Thakker 2016). This is thought to be due to statins' impact on insulin sensitivity and glucose metabolism. Other adverse events may include gastrointestinal symptoms such as nausea, diarrhea, and abdominal pain. These symptoms are typically mild but can contribute to non‐adherence in some patients.

How the intervention might work

The mechanism of action of statins in the prevention of VTE is not well‐established, and multiple mechanisms have been proposed (Wallace 2017). There are plausible biological links between statin therapy and the reduction of thrombotic risk, primarily through targeting blood coagulation, lipid metabolism, the immune system, the endothelium, and inflammation (Lippi 2013). Liu 2022 summarized the anti‐thrombotic impact of statins, which included cholesterol‐lowering effects, plaque stabilization, lipid‐lowering independent inhibitory effect on platelet activation and coagulation cascade; downregulating prothrombotic factors; promotion of endothelial nitric oxide synthase (eNOS), thrombomodulin expression, and fibrinolytic activity; and interference with the clotting system and coagulation cascade (Liu 2022). Statins inhibit cholesterol biosynthesis via the down‐regulation of HMG CoA reductase, a rate‐limiting enzyme for cholesterol biosynthesis (Ginter 2009; Istvan 2002). Statins reduce the cholesterol content of hepatocytes, thereby promoting enhanced expression of low‐density lipoprotein (LDL) receptors on the cell membrane, leading to increased receptor‐mediated endocytosis of LDL and decreased serum levels of LDL (Billings 2011; Orsi 2019). A 1 mmol/L reduction in LDL‐C could reduce the risk of VTE by an average of 37% (Zaccardi 2018). There is increasing evidence that statins modify the blood coagulation cascade at multiple levels, including down‐regulation of tissue factor (TF) expression, increased protein C activity, increased factor V and VII inactivation, and enhanced thrombomodulin expression (Ridker 2012; Riva 2015; Undas 2005; Wallace 2017).

Consequently, the benefits of statins may accrue not merely from their effects on lipid levels but also through their influence on thrombosis and inflammation (Albert 2001; Kaba 2004; Undas 2005). Statins may also have an active anti‐inflammatory component that contributes to decreased thrombus formation (Wallace 2017). A systematic review showed that statin therapy reduced interleukin 6 (IL‐ 6), C‐reactive protein (CRP) and monocyte chemoattractant protein‐1 (MCP‐1) (Rodriguez 2012). This suggests that the anti‐thrombotic effects are likely to be manifested through the anti‐inflammatory properties of statins (Rodriguez 2012). Studies have confirmed that statins may decrease VTE risk by inhibiting platelet activation and aggregation, which play key roles in the initiation of thrombus formation (Biedermann 2016; Moraes 2013; Pawelczyk 2015). Statins not only work by lowering cholesterol but also affect blood coagulation and inflammation through multiple mechanisms, thereby reducing the risk of VTE.

Why it is important to do this review

The ideal drug for primary prevention of VTE is efficacious, poses minimal bleeding risk, and is easy to administer. Statins meet the last two criteria, but their benefits and harms remain unclear (Ray 2003). Statins are commonly used in patients with cardiovascular risk factors for prophylaxis of cardiovascular disease (CVD). Therefore, if statins could be shown to downregulate hemostasis and prevent VTE episodes, they could become an alternative intervention to prevent VTE (Orsi 2019). While several observational studies have found evidence that statin use is associated with reduced risk of VTE (De Moreuil 2017; El‐Refai 2017; Grady 2000; Lacut 2004; Lassila 2014; Lötsch 2014; Ramcharan 2009; Skajaa 2019; Sørensen 2009), others have been unable to confirm this association (Al Harbi 2013; Frasco 2022; Ray 2001a; Shai 2014); see Table 4. The observational studies often failed to control for all confounding factors and cannot establish causal relationships. Three randomized controlled studies showed statin use might reduce the incidence of VTE (Glynn 2009; Kjekshus 2007; Yusuf 2016), but most studies (Ades 2018; Dinglas 2016; Fellström 2004; Fellström 2009; Freeman 2011; Talasaz 2023) did not find this.

3. Available evidence between statin use and the incidence of venous thromboembolism.

Study	Sample size	Venous thromboembolism (statin vs control)	Deep venous thrombosis (statin vs control)	Pulmonary embolism (statin vs control)
Al Harbi 2013	798	Hazard ratio 0.63, 95% CI 0.25 to 1.57	‐	‐
De Moreuil 2017	364	Odds ratio 0.33, 95% CI 0.12 to 0.90	‐	‐
El‐Refai 2017	170,459	‐	Hazard ratio 0.77, 95% CI 0.61 to 0.99	Hazard ratio 0.80, 95% CI 0.64 to 0.99
Frasco 2022	1384	Hazard ratio 0.92, 95% CI 0.39 to 2.21	‐	‐
Grady 2000	2763	Relative hazard 0.5, 95% CI 0.2 to 0.9	‐	‐
Herrington 2002	2763	Relative hazard 0.45, 95% CI 0.23 to 0.88	‐	‐
Huerta 2007	6550	‐	Odds ratio 0.70, 95% CI 0.50 to 0.97	‐
Lacut 2004	377	Odds ratio 0.42, 95% CI 0.23 to 0.76	‐	‐
Lacut 2008	1354	Odds ratio 0.53, 95% CI 0.37 to 0.78	‐	‐
Lassila 2014	8028	Rate ratio 0.60, 95% CI 0.36 to 1.00	‐	‐
Lötsch 2014	1434	Hazard ratio 0.43, 95% CI 0.19 to 0.98	‐	‐
Ramcharan 2009	10,452	Odds ratio 0.45, 95% CI 0.36 to 0.56	‐	‐
Ray 2001a	6550	Odds ratio 0.85, 95% CI 0.66 to 1.09	‐	‐
Ray 2001b	89,508	Hazard ratio 0.68, 95% CI 0.59 to 0.79	‐	‐
Shai 2014	3585	Hazard ratio 0.86, 95%CI 0.65 to 1.14	‐	‐
Skajaa 2019	601,011	Hazard ratio 0.95, 95% CI 0.92 to 0.97	‐	‐
Smeeth 2009	6199	Hazard ratio 1.02, 95% CI 0.88, 1.18	‐	‐
Sørensen 2009	64,064	Relative risk 0.74, 95% CI 0.63 to 0.85	‐	‐

Abbreviations: CI: confidence interval; vs versus

Similarly, findings from existing meta‐analyses provide conflicting evidence about the preventive role of statins (Kunutsor 2017; Miksza 2019; Rahimi 2012). Rahimi 2012 discovered statin therapy did not lessen the risk of VTE events (statins 465 (0.9%), control 521 (1.0%); odds ratio (OR) 0.89, 95% confidence interval (CI) 0.78 to 1.01, P = 0.08), while Kunutsor 2017 proposed that statin usage could lower the VTE risk (risk ratio (RR) 0.85, 95% CI 0.73 to 0.99; P = 0.04) when compared with placebo or no treatment. The benefits of statin use were not confirmed by recent randomized controlled trials (RCTs) (Ades 2018; Dinglas 2016; Talasaz 2023). Moreover, the meta‐analyses centered on the effectiveness of statins, neglecting to assess their potential detriments.

Consequently, uncertainty still exists about whether statins prevent VTE, prompting calls for more information to quantify the risk reduction for VTE with statin use (Gaertner 2016; Ray 2001a). In this first updated systematic review (Li 2014), we aimed to assess the benefits and risks of statins. We evaluated all available RCTs that used statins for the primary prevention of VTE to aid decision‐making for healthcare professionals and non‐professionals.

Objectives

To evaluate the benefits and risks of statins in preventing venous thromboembolism (VTE) in individuals with no prior history of VTE.

Methods

Criteria for considering studies for this review

Types of studies

We included RCTs that compared statins with any control intervention, including placebo or usual care. We did not include quasi‐randomized trials, where the allocation to different treatment regimens is not adequately concealed (e.g. allocation by alternation, date of birth, hospital number, day of the week), as foreknowledge of treatment allocation could lead to biased distribution and exaggerated treatment effects. We only included studies focused on the primary prevention of VTE, excluding studies that investigated secondary prevention and did not prospectively collect information on VTE.

Types of participants

We included studies involving participants who were healthy or had non‐VTE diseases. Participants with VTE risk factors were also included, provided they had not previously experienced VTE (Table 2). We excluded studies with participants, or subsets of participants, who had prior instances of VTE.

Types of interventions

We included studies that compared statins with any control intervention, including placebo or usual care. Eligible statins included atorvastatin, rosuvastatin, pravastatin, lovastatin, fluvastatin, pitavastatin, and simvastatin. There were no restrictions on the dose, duration, route, or timing of statins. We accepted concomitant interventions, provided they were administered to participants in both arms of the study. Usual care included a variety of lipid‐lowering therapies, such as diet, behavior modification, or antihyperlipidemic medication.

Types of outcome measures

We made decisions on exclusions only after contacting the study authors to verify whether any unreported or unpublished information was assessed. We classified the timing of the outcome assessment into three follow‐up durations: short‐term (≤ one year), medium‐term (one to five years), and long‐term (> five years). All outcomes were reported at the longest follow‐up time point.

Primary outcomes

• Any VTE (provoked or unprovoked, symptomatic or asymptomatic)

• DVT (provoked or unprovoked) located in any part of the human body, such as the lower legs, thighs, pelvis, or upper limbs

• PE (provoked or unprovoked)

The primary outcome of VTE could have been reported in any included study as either a primary, secondary, exploratory, or adverse event outcome. There was a range of appropriate assessment methods for VTE: detection by imaging via venous ultrasonography or venography for confirmation of DVT; angiography, computed tomography (CT), or ventilation and perfusion scan (V/Q) for confirmation of PE; or any other imaging approaches, such as magnetic resonance imaging (MRI) or any other emerging technologies for either DVT or PE.

Secondary outcomes

• Serious adverse events (SAEs) related to statin use; any adverse event that resulted in death or life‐threatening conditions, required inpatient hospitalization or prolonged existing hospitalization, or resulted in a persistent or significant disability or incapacity, or a congenital anomaly or birth defect

• Adverse events related to statin use

• Mortality (all‐cause mortality and mortality after PE)

We focused on VTE‐related outcomes as cardiovascular outcomes have been investigated in another Cochrane review (Taylor 2013).

Search methods for identification of studies

Electronic searches

The Cochrane Vascular Information Specialist conducted systematic searches of the following databases for RCTs, without any restrictions based on language, publication year, or status.

• Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web; searched 13 March 2023)

• Cochrane Central Register of Controlled Trials (CENTRAL; 2023, Issue 2) via the Cochrane Register of Studies Online (CRSO; searched 13 March 2023)

• MEDLINE Ovid (MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, MEDLINE Daily and MEDLINE; 1946 to 13 March 2023)

• Embase Ovid (1974 to 13 March 2023)

• CINAHL EBSCO (Cumulative Index to Nursing and Allied Health Literature; 1982 to 13 March 2023)

We developed search strategies for other databases based on the search strategy designed for MEDLINE. Where appropriate, they were combined with adaptations of the highly sensitive search strategy designed by Cochrane to identify RCTs and controlled clinical trials (as detailed in Chapter 4 of the Cochrane Handbook for Systematic Reviews of Interventions) (Lefebvre 2022).

We also searched the following trial registries.

• ClinicalTrials.gov (clinicaltrials.gov; searched 13 March 2023)

• World Health Organization International Clinical Trials Registry Platform (who.int/trialsearch; searched 13 March 2023)

The most recent searches were carried out on 13 March 2023. The search strategies are provided in Appendix 1.

Searching other resources

We examined the bibliographies of study reports identified in our search to locate other pertinent articles. Additionally, we searched the reference lists of available reviews, systematic reviews, and meta‐analyses. We reached out to the authors of relevant articles via emails to request data and identify other unpublished RCTs.

Data collection and analysis

Selection of studies

Two review authors (LL and ZXW) independently selected trials for inclusion. Any disagreements were resolved through discussion with a third review author (JHT). We documented the reasons for excluding all reports after full‐text assessment and presented a selection of these in a Characteristics of excluded studies table. We reached out to the study authors for all studies in which VTE was not reported or published, to verify the availability of these data. We depicted the study selection process in a PRISMA diagram (Page 2021).

Data extraction and management

Two review authors (LL and JHT) independently extracted relevant data from the included studies using a pre‐tested data collection form. We reported the following extracted information in the Characteristics of included studies table.

• Methods (study design, number of participants, post‐randomization exclusions, losses to follow‐up, intention‐to‐treat (ITT) analysis, and study duration)

• Participant characteristics (country, setting, age, sex, inclusion and exclusion criteria)

• Interventions (statins used, dose, comparator)

• Outcomes (outcomes evaluated, methods of measurement, and timing of assessment)

• Funding source for the study, and declarations of interest by the study authors

A third review author (ZXW or PZZ) cross‐checked the extracted data. If discrepancies arose, two review authors (LL and JHT) verified the data and established a consensus. In the event that they could not agree, the third review author (ZXW or PZZ) joined the discussion until a consensus was reached.

Assessment of risk of bias in included studies

Two review authors (LL and JHT) independently carried out a risk of bias assessment of all included studies, using the Cochrane risk of bias tool (RoB1), as outlined in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). This tool evaluates random sequence generation, allocation concealment methods, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and any other potential biases. Our bias assessment focused on blinding of outcome assessment, incomplete outcome data, and selective outcome reporting, particularly in regard to the primary outcome in the studies and the VTE outcomes.

We assigned a high, low, or unclear risk of bias to each domain. Any disagreements were resolved through discussion with a third review author (KHY). If clarification was necessary to assess the risk of bias, we contacted the study authors. We deemed studies to have a low risk of bias overall when all domains received a low risk of bias rating. We deemed studies to have a high risk of bias overall if one or more domains received a high risk of bias rating or if multiple domains were rated as having an unclear risk of bias. Finally, we deemed studies to have an unclear risk of bias overall if one or more domains were rated as having an unclear risk of bias, with none at a high risk of bias.

Measures of treatment effect

We defined the measures of treatment effects according to Chapter 6 in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022). For dichotomous outcomes, we expressed results as odds ratios (ORs) with 95% confidence intervals (CI).

Unit of analysis issues

We used the individual participant as the unit of analysis. Aggregate patient data (APD) from published articles was used for data analysis. For parallel RCTs, we considered the incidence of outcomes in each group. For cross‐over RCTs, we combined the incidence of outcomes in each group from the first period before cross‐over and the second period after cross‐over. For cluster‐randomized trials, wherein groups of individuals were randomized together to the same intervention, we combined the incidence of outcomes in each group of participants. The direct combination was employed for multi‐arm trials’ intervention groups.

Dealing with missing data

We attempted to contact all authors of the original studies, using provided emails, telephone numbers, or fax details where available, to request missing data. If we did not receive a response from the study’s authors within four weeks, we extracted all the available data from the publication. In the case of missing data due to participants dropping out or being lost to follow‐up, we conducted a primary analysis based on ITT analysis data, as described in Chapter 10 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022).

Assessment of heterogeneity

We evaluated studies for clinical heterogeneity by examining variability in participants, interventions, outcomes, and methodological heterogeneity by investigating variability in study design, outcome measurement tools, and risk of bias. We assessed statistical heterogeneity through a visual inspection of the forest plot to check for overlapping within CIs and by using the Chi² test for homogeneity with a 10% level of significance. We also applied the I² statistic to measure the degree of inconsistency among the studies, considering an I² greater than 50% to represent substantial heterogeneity (Deeks 2022). Potential sources of statistical heterogeneity were explored using subgroup analyses (Subgroup analysis and investigation of heterogeneity) and sensitivity analyses (Sensitivity analysis).

Assessment of reporting biases

We assessed reporting biases using funnel plots, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Page 2022).

Data synthesis

We used Review Manager Web for all statistical analyses and to generate figures (Review Manager 2020; RevMan 2019), employing an ITT analysis. We incorporated expected clinical, methodological, or both types of heterogeneity across studies using random‐effects models (see Assessment of heterogeneity). Data were represented as OR with 95% CI. We chose to use the Mantel‐Haenszel methods because of their superior statistical properties when few events occur (Deeks 2022).

Subgroup analysis and investigation of heterogeneity

We conducted the following subgroup analyses to investigate heterogeneity.

• Provoked status (unprovoked VTE, provoked VTE)

• Age (< 70 years, > 70 years)

• Sex (female, male)

• Disease (healthy population, hypercholesterolemia, heart failure, chronic kidney disease, diabetes mellitus, aortic valve stenosis, cerebrovascular disease, vascular disease, other diseases)

• Type of statin (atorvastatin, rosuvastatin, pravastatin, lovastatin, fluvastatin, pitavastatin, simvastatin)

• Dose (different doses for each statin)

• Treatment duration (≤ one year, one to five years, > five years)

We compared subgroup differences using the formal test for subgroup differences in RevMan Web (Review Manager 2020).

Sensitivity analysis

For VTE, we integrated unpublished data (Kunutsor 2017; Rahimi 2012) in meta‐analyses and performed sensitivity analysis excluding these unpublished data. We conducted sensitivity analysis excluding studies at overall high risk of bias, to explore the impact on the results of the meta‐analysis. We also carried out a sensitivity analysis that excluded studies with more than 15% missing outcome data.

For other outcomes, we conducted sensitivity analyses to explore whether the omission of particular studies affected the stability of the treatment effect estimates when there was heterogeneity in the meta‐analysis.

Trial sequential analysis

When meta‐analyses are updated with new trials, type I errors may result due to sparse data and repeated significance testing (Brok 2008). We utilized trial sequential analysis (TSA) to determine if the evidence identified for this review was sufficient. Trial sequential monitoring boundaries were established using TSA software (Thorlund 2011). If the cumulative Z‐curve intersected the trial sequential monitoring boundary (TSMB), we inferred that a sufficient level of evidence had been achieved, potentially negating the necessity for further trials. However, when the cumulative Z‐curve did not intersect the TSMB or did not exceed the futility boundaries before the required information size was reached, we considered the evidence to be insufficient (Bjelakovic 2014). We also calculated the required information size, which is the minimum number of participants needed in a meta‐analysis to confirm or reject a specific intervention effect. We adjusted the required information size to account for statistical between‐trial heterogeneity using a diversity adjustment factor (Wetterslev 2009). In our meta‐analysis, the diversity‐adjusted required information size was based on the event proportion in the control group, the presumption of a plausible RR reduction of 20%, risk of type I error of 5%, risk of type II error of 20%, and the assumed diversity of the meta‐analysis (Wetterslev 2009).

Summary of findings and assessment of the certainty of the evidence

We prepared a summary of findings table using the GRADEpro Guideline Development tool (GRADEpro GDT), in which we presented the main findings of the review (Atkins 2004; Guyatt 2008). We presented the following outcomes in Table 1, given that they are essential for clinical decision‐making: any VTE, DVT, PE, SAEs, adverse effects such as myopathy, and mortality. Two review authors (LL and JHT) independently assessed the certainty of the evidence for these outcomes, and we resolved any disagreements through discussion with a third review author (KHY). We evaluated the certainty of the evidence using the GRADE approach (Guyatt 2011), assigning one of four levels of certainty (high, moderate, low, or very low) for the key outcomes based on the overall risk of bias, directness of the evidence, consistency of the results, precision of the effect estimates, and risk of publication bias (Schünemann 2022). We provided justifications for our decisions to downgrade the certainty of evidence in footnotes, and, where necessary, we provided comments to assist readers in understanding the results.

Results

Description of studies

Results of the search

The database searches identified a total of 5509 records, and we identified 126 records by reference tracking. After duplicates were removed, we assessed 4557 records by title and abstract. We assessed 4405 reports as ineligible. We assessed 152 potentially relevant reports by full text. Of these, we identified 27 studies (56 reports) that met our inclusion criteria (see Characteristics of included studies). Seven studies (8 reports) are ongoing (ChiCTR2100048407; NCT00259662; NCT01021488; NCT01063426; NCT01524653; NCT02285738; NCT03532139). See Figure 1 for details of the search results.

1.

PRISMA diagram of study selection process

Included studies

See Characteristics of included studies table and Table 5.

4. Summary of key characterstics of included studies.

Study	Age (mean ± SD) in years, statins/control group	Sex (% female), statins vs control group	Follow‐up	Setting	Funding
Ades 2018	61	‐	9 weeks	Hospital	Intramural grant from the University of Vermont Cancer Center
Amarenco 2006	63.0 ± 0.2/62.5 ± 0.2	39.7% vs 41.0%	4.9 years	Hospital	Pfizer
Asselbergs 2004	52.1 ± 11.9/50.5 ± 11	32.3% vs 37.8%	46 months	Hospital	Dutch Kidney Foundation, Netherlands Heart Foundation, Bristol Myers Squibb
Chan 2010	58.0 ± 12.9/57.9 ± 14.3	39.5% vs 37.0%	3.5 years	Hospital	The Canadian Institutes of Health Research, AstraZeneca Canada Inc
Colhoun 2004	61.5 ± 8.3/61.8 ± 8.0	31.9% vs 32.1%	3.9 years	Hospital	UK Department of Health, Diabetes UK, and Pfizer
Cowell 2005	68 ± 11/68 ± 10	32% vs 28%	25 months	Hospital	British Heart Foundation, Pfizer, Wellcome Trust Clinical Research Facility, Edinburgh
Crouse 2007	57 ± 6.2/57 ± 6.0	40% vs 41%	24 months	Primary care	AstraZeneca
Dinglas 2016	54±17/54±16	51% vs 51%	12 months	Hospital	National Heart, Lung and Blood Institute and the Investigator‐Sponsored Study Program of AstraZeneca
Downs 1998	58 ± 7/58 ± 7	15% vs 15%	5.2 years	Hospital	Merck
Fassett 2010	60.0 ± 15.0/60.3 ± 15.2	38% vs 32%	2.5 years	Hospital	Clifford Craig Medical Research Trust, Pfizer
Feldman 2010	74.0 ± 8.0/73.2 ± 8.7	53% vs 51%	1.5 years	Hospital	Pfizer
Fellström 2004	49.5 ± 10.9/50.0 ± 11.0	33.2% vs 34.8%	5.1 years	Hospital	Novartis Pharma, Basel, Switzerland
Fellström 2009	64.1±8.6/64.3±8.7	38.7% vs 37.0%	2.4 years	Hospital	AstraZeneca
Freeman 2011	75.4 ± 3.3/75.3 ± 3.4	74% vs 72.4%	3.2 years	Hospital	Chest, Heart, Stroke Scotland Project Grant R05A89, Chief Scientist Office NHS Clinical Research Grant 2006, and a GlaxoSmithKline Clinical Fellowship and Tenovus Scotland research grant
Glynn 2009	66.1 ± 7.64/66.1 ± 7.8	38.5% vs 37.9%	1.9 years	Community	AstraZeneca
HPS study 2002	64 ± 8.4	24.7% vs 24.7%	5.3 years	Hospital	UK Medical Research Council, British Heart Foundation, Merck, and Roche Vitamins
Kjekshus 2007	73 ± 7.1/73 ± 7.0	24% vs 24%	32.8 months	Hospital	AstraZeneca
Knopp 2006	61.1 ± 8.1/61.0 ± 8.2	34.3% vs 33.0%	4 years	Hospital	Pfizer
Koren 2004	61.1 ± 9.0/61.3 ± 8.6	17.8% vs 17.7%	51.5 months	Hospital	Parke‐Davis and Pfizer Pharmaceuticals
LIPID Study 1998	62/62	17% vs 17%	6.1 years	Hospital	Bristol‐Myers Squibb Pharmaceutical Research Institute and National Heart Foundation of Australia
Nakamura 2006	58.2 ± 7.3/58.4 ± 7.2	68% vs 69%	5.3 years	Hospital	Japanese Ministry of Health, Labor and Welfare and Sankyo
Sever 2003	63.1 ± 8.5/63.2 ± 8.6	18.9% vs 18.7%	3.3 years	Hospital	Pfizer, Servier Research Group, and Leo Laboratories
Sola 2006	53.3 ± 6.2/54.1 ± 6.9	36% vs 38%	1 year	Hospital	‐
Talasaz 2023	57/57	43% vs 44%	30 days	Hospital	The Rajaie Cardiovascular Medical and Research Center
Tavazzi 2008	68 ± 11/68 ± 11	23.8% vs 21.4%	3.9 years	Hospital	Società Prodotti Antibiotici, Pfizer, Sigma Tau, and AstraZeneca
Wanner 2005	65.7 ± 8.3/65.7 ± 8.3	46.2% vs 45.9%	4.08 years	Hospital	Pfizer
Yusuf 2016	65.8 ± 6.4/65.7 ± 6.3	46.4% vs 46.1%	5.6 years	Hospital	The Canadian Institutes of Health Research and AstraZeneca

Abbreviations: SD: standard deviation; vs: versus

This review includes 27 RCTs with 122,601 participants (Ades 2018; Amarenco 2006; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Crouse 2007; Dinglas 2016; Downs 1998; Fassett 2010; Feldman 2010; Fellström 2004; Fellström 2009; Freeman 2011; Glynn 2009; HPS study 2002; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Sola 2006; Talasaz 2023; Tavazzi 2008; Wanner 2005; Yusuf 2016).

Glynn 2009 was the only study that had VTE as a primary endpoint and collected VTE outcomes systematically. This study investigated the prevention effects of rosuvastatin on the incidence of VTE. This study used rosuvastatin 20 mg daily for healthy people aged 50 years and older without a history of cardiovascular or cerebrovascular events.

We reached out to the authors of the other 26 studies for VTE data; however, none responded. Subsequently, we contacted the authors of published meta‐analyses, namely Kunutsor 2017 and Rahimi 2012, to enquire how they obtained these data. Professor Kunustor explained that they obtained data from a meta‐analysis by Professor Rahimi and colleagues (Rahimi 2012), who, in turn, replied that they had received the data from the trialists and could not locate the original data since many years had passed. Given this, we extracted the VTE data from these existing meta‐analyses (Kunutsor 2017; Rahimi 2012); see Table 6.

5. VTE data source for included studies.

Study ID	VTE data published	VTE data requested	Response received	VTE data received	VTE data source
Ades 2018	Yes	‐	‐	‐	Study authors
Amarenco 2006	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Asselbergs 2004	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Chan 2010	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Colhoun 2004	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Cowell 2005	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Crouse 2007	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Dinglas 2016	Yes	‐	‐	‐	Study authors
Downs 1998	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Fassett 2010	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Feldman 2010	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Fellström 2004	Yes	‐	‐	‐	Study authors
Fellström 2009	Yes	‐	‐	‐	Study authors
Freeman 2011	Yes	‐	‐	‐	Study authors
Glynn 2009	Yes	‐	‐	‐	Study authors
HPS study 2002	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Kjekshus 2007	Yes	‐	‐	‐	Study authors
Knopp 2006	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Koren 2004	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
LIPID Study 1998	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Nakamura 2006	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Sever 2003	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Sola 2006	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Talasaz 2023	Yes	‐	‐	‐	Study authors
Tavazzi 2008	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Wanner 2005	‐	Yes	No	No	Published meta‐analysis (Kunutsor)
Yusuf 2016	Yes	‐	‐	‐	Study authors

Abbreviations: ID: identifier; VTE: venous thrombolism

Setting

All but two studies were hospital‐based RCTs. Glynn 2009 was a community‐based RCT, and Crouse 2007 included participants from primary care centers.

Participants

The health of the study participants was as follows.

• Four studies included people with chronic kidney disease (including renal transplant recipients, end‐stage renal disease, and persistent microalbuminuria) (Asselbergs 2004; Fassett 2010; Fellström 2004; Fellström 2009).

• Three studies included people with heart failure (Kjekshus 2007; Sola 2006; Tavazzi 2008).

• Three studies included people with type 2 diabetes mellitus (Colhoun 2004; Knopp 2006; Wanner 2005).

• Three studies included people with hypercholesterolemia (Crouse 2007; Downs 1998; Nakamura 2006).

• Two studies included people with a medical history of coronary heart disease (Koren 2004; LIPID Study 1998).

• Two studies included people with vascular disease (Freeman 2011; HPS study 2002).

• Two studies included people with aortic valve stenosis (Chan 2010; Cowell 2005).

• Two studies included people with cerebrovascular disease (Alzheimer's disease, stroke, or transient ischemic attack (TIA)) (Amarenco 2006; Feldman 2010).

• One study included a healthy population (Glynn 2009).

• One study included people with cancer (Ades 2018).

• One study included people with hypertension (Sever 2003).

• One included people who had cardiovascular risk factors (Yusuf 2016).

• One study included people with COVID‐19 (Talasaz 2023).

• One included mechanically ventilated patients with sepsis‐associated respiratory distress syndrome (Dinglas 2016).

Age

All studies focused on adult participants (> 18 years old); two focused on participants more than 60 years old (Freeman 2011; Kjekshus 2007) (Table 5).

Sex

All studies incorporated both male and female participants (Table 5). In most research, the proportion of female participants fell below 50%; it only exceeded this percentage in four studies (Dinglas 2016; Feldman 2010; Freeman 2011; Nakamura 2006).

Participant location

The participants in the studies were from several countries in Asia, America, and Europe. Fifteen studies were conducted in one country: USA (Ades 2018; Dinglas 2016; Downs 1998; Koren 2004; Sola 2006), UK (Cowell 2005; Freeman 2011; HPS study 2002), the Netherlands (Asselbergs 2004), Canada (Chan 2010), Australia (Fassett 2010), Germany (Wanner 2005), Japan (Nakamura 2006), Iran (Talasaz 2023), and Italy (Tavazzi 2008). Two studies were conducted in two countries (Colhoun 2004 in the UK and Ireland; LIPID Study 1998 in Australia and New Zealand). Nine studies were conducted in more than two countries (Amarenco 2006; Crouse 2007; Feldman 2010; Fellström 2009; Fellström 2004; Kjekshus 2007; Knopp 2006; Sever 2003; Yusuf 2016). The participant locations for other included studies are presented in the Characteristics of included studies table.

Interventions

The statin type, dose, and duration were different across studies.

• Eleven studies compared atorvastatin with a placebo or usual care.

  • Four studies compared 10 mg/day atorvastatin with a placebo (Colhoun 2004; Fassett 2010; Knopp 2006; Sever 2003).

  • Three studies compared 20 mg/day atorvastatin with a placebo (Sola 2006; Talasaz 2023; Wanner 2005).

  • Three studies compared 80 mg/day atorvastatin with a placebo (Amarenco 2006; Cowell 2005; Feldman 2010).

  • One study compared 80 mg/day atorvastatin with usual care (Koren 2004).

• Nine studies compared rosuvastatin with a placebo.

  • Four studies compared 10 mg/day rosuvastatin with a placebo (Fellström 2009; Kjekshus 2007; Tavazzi 2008; Yusuf 2016).

  • Three studies compared 20 mg/day rosuvastatin with a placebo (Ades 2018; Dinglas 2016; Glynn 2009).

  • Two studies compared 40 mg/day rosuvastatin with a placebo (Chan 2010; Crouse 2007).

• Four studies compared pravastatin with a placebo.

  • Three compared 40 mg/day pravastatin with a placebo (Asselbergs 2004; Freeman 2011; LIPID Study 1998).

  • One study compared 10 to 20 mg/day pravastatin plus diet with diet alone (Nakamura 2006).

• One study compared 20 to 40 mg/day lovastatin with a placebo (Downs 1998).

• One study compared 40 mg/day fluvastatin with a placebo (Fellström 2004).

• One study compared 40 mg/day simvastatin with a placebo (HPS study 2002).

Outcomes

Primary outcomes

• Any VTE: nine studies reported the incidence of VTE in the final publication (Ades 2018; Dinglas 2016; Fellström 2004; Fellström 2009; Freeman 2011; Glynn 2009; Kjekshus 2007; Talasaz 2023; Yusuf 2016) (Table 6). Of these, Glynn 2009 was the only study that had VTE as a primary endpoint and collected VTE outcomes systematically. The remaining 18 studies did not report the incidence of VTE (Amarenco 2006; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Crouse 2007; Downs 1998; Fassett 2010; Feldman 2010; HPS study 2002; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Sola 2006; Tavazzi 2008; Wanner 2005). We requested these data from the study authors but received no response. Consequently, we retrieved data on the incidence of VTE in these 18 studies from published meta‐analyses (Kunutsor 2017; Rahimi 2012) (Table 6).

• DVT: six studies reported the incidence of DVT (Dinglas 2016; Fellström 2009; Freeman 2011; Glynn 2009; Talasaz 2023; Yusuf 2016).

• PE: five studies reported the incidence of PE (Dinglas 2016; Fellström 2004; Fellström 2009; Glynn 2009; Kjekshus 2007).

Secondary outcomes

• Serious adverse events: 13 studies reported the incidence of SAEs (Amarenco 2006; Chan 2010; Colhoun 2004; Crouse 2007; Downs 1998; Dinglas 2016; Feldman 2010; Fellström 2009; Glynn 2009; HPS study 2002; Kjekshus 2007; Knopp 2006; Koren 2004).

• Adverse events

  • Ten studies reported the incidence of myopathy (Amarenco 2006; Chan 2010; Colhoun 2004; Crouse 2007; Fellström 2004; Glynn 2009; HPS study 2002; LIPID Study 1998; Tavazzi 2008; Yusuf 2016).

  • Seven studies reported the incidence of rhabdomyolysis (Amarenco 2006; Dinglas 2016; Downs 1998; Glynn 2009; HPS study 2002; Knopp 2006; Yusuf 2016).

  • Two studies reported the incidence of bleeding (Glynn 2009; Talasaz 2023).

  • Eleven studies reported the incidence of muscular weakness, stiffness, or pain (Amarenco 2006; Colhoun 2004; Crouse 2007; Downs 1998; Fellström 2004; Glynn 2009; Kjekshus 2007; Knopp 2006; Sever 2003; Tavazzi 2008; Yusuf 2016).

  • Ten studies reported the incidence of gastrointestinal disorders (Amarenco 2006; Chan 2010; Cowell 2005; Dinglas 2016; Fellström 2004; Fellström 2009; Glynn 2009; Kjekshus 2007; Sever 2003; Tavazzi 2008).

  • Twenty studies reported the incidence of hepatic disorders (Amarenco 2006; Chan 2010; Colhoun 2004; Crouse 2007; Dinglas 2016; Downs 1998; Feldman 2010; Fellström 2004; Fellström 2009; Glynn 2009; HPS study 2002; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Talasaz 2023; Tavazzi 2008; Wanner 2005). The definition for hepatic disorders was increased liver enzymes; the specific definition in each study is listed in Table 7.

  • Eight studies reported the incidence of renal disorders (Chan 2010; Crouse 2007; Dinglas 2016; Fellström 2009; Glynn 2009; Kjekshus 2007; Sever 2003; Tavazzi 2008).

• Mortality: 23 studies reported the incidence of mortality (Amarenco 2006; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Crouse 2007; Downs 1998; Fassett 2010; Feldman 2010; Fellström 2004; Fellström 2009; Glynn 2009; HPS study 2002; Talasaz 2023; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Tavazzi 2008; Wanner 2005; Yusuf 2016).

6. Adverse events.

Study	Serious adverse events	Adverse events	Definitions of hepatic disorders
Amarenco 2006	Unclear	Myopathy; rhabdomyolysis; muscular weakness, stiffness, or pain; gastrointestinal disorder; hepatic disorder	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limits of normal
Chan 2010	Cancer; diabetes mellitus; gastrointestinal; hepatic; pain; renal; angina	Myopathy; gastrointestinal disorder; hepatic disorder; renal disorder	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limits of normal
Colhoun 2004	Unclear	Myopathy; muscular weakness, stiffness, or pain; hepatic disorder	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limits of normal
Cowell 2005	Not reported	Gastrointestinal disorder	Not reported
Crouse 2007	Cardiovascular events	Myopathy; muscular weakness, stiffness, or pain; hepatic disorder; renal disorder	Aspartate or alanine aminotransferase increased
Dinglas 2016	Not reported	Gastrointestinal disorders; hepatic disorder; renal disorder	Alanine aminotransferase > 8 times upper limit of normal
Downs 1998	Unclear	Rhabdomyolysis; muscular weakness, stiffness, or pain	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limits of normal
Fassett 2010	Not reported	Not reported	Not reported
Feldman 2010	Falls; hepatitis; acute renal failure/rhabdomyolysis/pancreatitis; abdominal pain/nausea/chest discomfort; transaminases elevation; liver disorder; gastrointestinal hemorrhage; deaths	Hepatic disorder	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limits of normal
Fellström 2004	Not reported	Myopathy; gastrointestinal disorder; hepatic disorder	Alanine aminotransferase levels ≥ 3 times the upper limits of normal
Fellström 2009	Unclear	Gastrointestinal disorder; hepatic disorder; renal disorder	Cholelithiasis, cholecystitis, cholecystitis acute, biliary dilatation, cholangitis, bile duct stone, gallbladder perforation, hepatic cyst, hepatitis, hepatitis toxic, jaundice, bile duct obstruction, biliary colic, cholecystitis chronic, hepatic cirrhosis, hepatic function abnormal, hepatic necrosis, hyperbilirubinaemia, portal vein occlusion
Glynn 2009	Muscle weakness, stiffness, or pain; myopathy; rhabdomyolysis; newly diagnosed cancer; death from cancer; gastrointestinal disorder; renal disorder; bleeding event; hepatic disorder; newly diagnosed diabetes	Myopathy; rhabdomyolysis; bleeding; muscular weakness, stiffness, or pain; gastrointestinal disorder; hepatic disorder; renal disorder	Alanine aminotransferase levels ≥ 3 times the upper limits of normal
HPS study 2002	Liver‐related serious adverse event	Myopathy; rhabdomyolysis; hepatic disorder	Liver enzymes increased
Kjekshus 2007	Cardiac disorders; infection; general disorder; nervous system disorders; respiratory disorders; gastrointestinal disorders; vascular disorders; neoplasms; injury or procedural complications	Muscular weakness, stiffness, or pain; gastrointestinal disorder; hepatic disorder	Alanine aminotransferase levels ≥ 3 times the upper limits of normal
Knopp 2006	Headaches; kidney failure; gastrointestinal bleeding; transaminase elevation; cholestatic jaundice; duodenal ulcer; vertigo	Rhabdomyolysis; muscular weakness, stiffness, or pain; hepatic disorder	Aspartate or alanine aminotransferase levels ≥ 1.5 times the upper limit of normal
Koren 2004	Chest pain; atrial fibrillation; pneumonia; cellulitis; gastrointestinal hemorrhage	Not reported	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limit of normal
LIPID Study 1998	Not reported	Myopathy; hepatic disorder	Alanine aminotransferase levels ≥ 3 times the upper limits of normal
Nakamura 2006	Not reported	Not reported	Aspartate aminotransferase and alanine aminotransferase concentrations exceeded 100 IU/L
Sever 2003	Musculoskeletal and connective tissue disorders; neoplasms; renal and urinary disorders; respiratory, thoracic and mediastinal disorder	Muscular weakness, stiffness, or pain; gastrointestinal disorder; hepatic disorder; renal disorder	Not reported
Talasaz 2023	Not reported	Bleeding; hepatic disorder	Aspartate or alanine aminotransferase levels ≥ 3 times the upper limits of normal
Tavazzi 2008	Not reported	Myopathy; muscular weakness, stiffness, or pain; gastrointestinal disorder; hepatic disorder; renal disorder	Liver enzymes increased or hepatocellular jaundice
Wanner 2005	Not reported	Hepatic disorder	Alanine aminotransferase level > 4 times the upper limit of normal
Yusuf 2016	Not reported	Myopathy; rhabdomyolysis; muscular weakness, stiffness, or pain	Not reported

Follow‐up

The median follow‐up durations ranged from 9 weeks to 6.1 years. The timing of outcome assessment was one year or less in four studies (Ades 2018; Dinglas 2016; Sola 2006; Talasaz 2023), between one and five years in 17 studies (Amarenco 2006; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Crouse 2007; Fassett 2010; Feldman 2010; Fellström 2009; Freeman 2011; Glynn 2009; Kjekshus 2007; Knopp 2006; Koren 2004; Sever 2003; Tavazzi 2008; Wanner 2005), and more than five years in six studies (Downs 1998; Fellström 2004; HPS study 2002; LIPID Study 1998; Nakamura 2006; Yusuf 2016).

Funding

Twenty‐four studies were industry‐funded (Amarenco 2006; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Crouse 2007; Dinglas 2016; Downs 1998; Fassett 2010; Feldman 2010; Fellström 2004; Fellström 2009; Freeman 2011; Glynn 2009; HPS study 2002; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Tavazzi 2008; Wanner 2005; Yusuf 2016). Two studies were funded by research centers (Ades 2018; Talasaz 2023), and one study did not report the sources of funding (Sola 2006). See Table 5.

Excluded studies

Excluded studies

We excluded 93 studies, as they were clearly irrelevant. We excluded a further three studies for the following reasons: Delluc 2022 and Orsi 2019, as they focused on the secondary prevention of VTE by statins, and Barrett 2023 as it compared acetylsalicylic acid plus statin with placebos. See Characteristics of excluded studies.

Ongoing studies

We assessed seven studies as ongoing (ChiCTR2100048407; NCT00259662; NCT01021488; NCT01063426; NCT01524653; NCT02285738; NCT03532139). These RCTs focused on rosuvastatin (ChiCTR2100048407; NCT01021488; NCT01524653; NCT02285738; NCT03532139) and atorvastatin (NCT00259662; NCT01063426). The participants in these seven RCTs were adults requiring total knee replacement (ChiCTR2100048407; NCT01021488; NCT01063426), people with gynecologic tumors scheduled for resection (NCT00259662), and people with cancers (NCT01524653; NCT02285738; NCT03532139). Four RCTs assessed the incidence of DVT (ChiCTR2100048407; NCT00259662; NCT01021488; NCT01063426), and three assessed the incidence of VTE (NCT01524653; NCT02285738; NCT03532139). Two RCTs assessed the incidence of VTE for more than one year (NCT01524653; NCT02285738), and three RCTs assessed the incidence of VTE for less than one year (NCT01021488; NCT01063426; NCT03532139). The remaining two RCTs did not report the timing of assessments. Further details can be seen in the Characteristics of ongoing studies.

Risk of bias in included studies

We presented the risk of bias assessment for each included study in the risk of bias section of the Characteristics of included studies table, and we summarised the risk of bias in the studies in Figure 2 and Figure 3. We wrote to trial authors to ask for details about the risk of bias in their studies when information was ambiguous or lacking, but we did not receive any responses. We assessed four studies to be at low risk of bias overall (Glynn 2009; Sever 2003; Tavazzi 2008; Wanner 2005), and four studies to have an unclear risk of bias overall (Colhoun 2004; Cowell 2005; Kjekshus 2007; LIPID Study 1998). We assessed the remaining studies to be at high risk of bias overall (Ades 2018; Amarenco 2006; Asselbergs 2004; Chan 2010; Crouse 2007; Dinglas 2016; Downs 1998; Fassett 2010; Feldman 2010; Fellström 2004; Fellström 2009; Freeman 2011; HPS study 2002; Knopp 2006; Koren 2004; Nakamura 2006; Sola 2006; Talasaz 2023; Yusuf 2016).

2.

Risk of bias graph: review authors' judgements about each 'Risk of bias' item presented as percentages across all included studies

3.

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

Allocation

All included studies were reported to be RCTs. Of these, seven studies did not report specific details, so we assessed them as being at unclear risk of selection bias (Amarenco 2006; Dinglas 2016; Downs 1998; Fellström 2009; Knopp 2006; Koren 2004; Sola 2006). The remaining 20 studies reported the details of the randomization generation methods, and we assessed them as being at low risk. Of these 20 studies, 16 reported using a computer to generate the randomization code (Ades 2018; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Fassett 2010; Freeman 2011; Glynn 2009; HPS study 2002; Kjekshus 2007; Nakamura 2006; Sever 2003; Talasaz 2023; Tavazzi 2008; Wanner 2005; Yusuf 2016); three studies used fixed‐block randomization, but they did not report any further details (Crouse 2007; Fellström 2004: LIPID Study 1998); and one study simply stated that they used randomization code (Feldman 2010).

Twelve studies reported how they concealed allocation, and we judged them to be at low risk of bias (Ades 2018; Chan 2010; Cowell 2005; Fassett 2010; Freeman 2011; Glynn 2009; HPS study 2002; Sever 2003; Talasaz 2023; Tavazzi 2008; Wanner 2005; Yusuf 2016). Of these, five studies allocated participants using a computer system (Glynn 2009; HPS study 2002; Sever 2003; Talasaz 2023; Tavazzi 2008), four using the internet or telephone system (Chan 2010; Freeman 2011; Wanner 2005; Yusuf 2016), and one using numbered containers (Cowell 2005). In two studies, the Clinical Trial Pharmacist allocated a random number (Ades 2018; Fassett 2010). The remaining 15 studies were judged to be at unclear risk because they did not provide any details of concealed allocation (Amarenco 2006; Asselbergs 2004; Colhoun 2004; Crouse 2007; Dinglas 2016; Downs 1998; Feldman 2010; Fellström 2004; Fellström 2009; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sola 2006).

Incorporating both random sequence generation and allocation concealment, we judged 12 studies to be at low risk of bias in selection bias (Ades 2018; Chan 2010; Cowell 2005; Fassett 2010; Freeman 2011; Glynn 2009; HPS study 2002; Sever 2003; Talasaz 2023; Tavazzi 2008; Wanner 2005; Yusuf 2016), and the other studies as unclear.

Blinding

Performance bias

Seventeen studies reported and described that both the participants and personnel were blinded. Thus, they were at low risk of performance bias (Ades 2018; Chan 2010; Colhoun 2004; Crouse 2007; Dinglas 2016; Downs 1998; Feldman 2010; Fellström 2004; Fellström 2009; Glynn 2009; Kjekshus 2007; LIPID Study 1998; Sever 2003; Sola 2006; Talasaz 2023; Tavazzi 2008; Yusuf 2016). Four additional studies, although they used a single‐blind method, were assessed as low risk of performance bias. This assessment was made because two studies reported that participants were blinded (HPS study 2002; Wanner 2005), and the other two declared that personnel were blinded (Cowell 2005; Fassett 2010). Hence, we evaluated a total of 21 studies as having a low risk of performance bias. Conversely, two studies reported that neither the participants nor personnel were blinded, leading us to consider them at high risk of performance bias (Koren 2004; Nakamura 2006). The remaining studies did not provide any details about their double‐blinding methods and were therefore judged to be at an unclear risk of performance bias (Amarenco 2006; Asselbergs 2004; Freeman 2011; Knopp 2006).

Detection bias

Eighteen studies reported that the outcome assessors were blinded for the primary and/or secondary outcomes, including VTE, and so we judged them to be at low risk for detection bias (Ades 2018; Asselbergs 2004; Chan 2010; Colhoun 2004; Dinglas 2016; Downs 1998; Fellström 2004; Glynn 2009; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Sola 2006; Talasaz 2023; Tavazzi 2008; Wanner 2005). Nine studies did not report the details of blinding for outcome assessment, so we judged them to be at unclear risk of detection bias (Amarenco 2006; Cowell 2005; Crouse 2007; Fassett 2010; Feldman 2010; Fellström 2009; Freeman 2011; HPS study 2002; Yusuf 2016).

Incomplete outcome data

We deemed missing outcome data at 15% or more as potentially significant. We evaluated 11 studies as having a low risk of attrition bias, as the proportion of missing outcome data was sufficiently low that any resulting bias would be minimal to the point of being insignificant (Colhoun 2004; Cowell 2005; Dinglas 2016; Fellström 2009; Freeman 2011; Glynn 2009; Kjekshus 2007; LIPID Study 1998; Sever 2003; Tavazzi 2008; Wanner 2005). The remaining 16 studies we assessed as having a high risk of bias, as the proportions of missing data were high enough to possibly affect the study results (Adams 2013; Amarenco 2006; Asselbergs 2004; Chan 2010; Crouse 2007; Downs 1998; Fassett 2010; Feldman 2010; Fellström 2004; HPS study 2002; Talasaz 2023; Knopp 2006; Koren 2004; Nakamura 2006; Sola 2006; Yusuf 2016).

Selective reporting

Six studies reported the incidence of DVT (Dinglas 2016; Fellström 2009; Freeman 2011; Glynn 2009; Talasaz 2023; Yusuf 2016), and five studies reported the incidence of PE (Dinglas 2016; Fellström 2004; Fellström 2009; Glynn 2009; Kjekshus 2007). Two studies reported the incidence of VTE (Dinglas 2016; Fellström 2009). We considered these eight studies to be at low risk of bias. We did not find a protocol for one study (Ades 2018), so this study was considered to have an unclear risk of bias. The other studies did not report incidence of VTE; however, these could be retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012) (data in Kunutsor 2017 was retrieved from Rahimi 2012, and data in Rahimi 2012 was obtained from the trialists). It was difficult to judge whether these studies selectively reported VTE or not. After careful consideration and discussion amongst all review authors, we judged these studies to be at low risk of bias; although they failed to report unexpected outcomes, they did report all predefined outcomes (Amarenco 2006; Asselbergs 2004; Chan 2010; Colhoun 2004; Cowell 2005; Crouse 2007; Downs 1998; Fassett 2010; Feldman 2010; HPS study 2002; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Sola 2006; Tavazzi 2008; Wanner 2005).

Other potential sources of bias

The trial data in one study were collected by the sponsor (Fellström 2009), potentially introducing bias during the study. We deemed this study to have a high risk of bias. We did not find any potential conflicts of interest in the other studies.

Effects of interventions

See: Table 1

Any VTE (provoked or unprovoked, symptomatic or asymptomatic)

The time points for follow‐up ranged from 9 weeks to 6.1 years. The median incidence of VTE in the statin groups was 0.72% (ranging from 0% to 10.53%), while the median incidence of VTE in the control groups was 0.89% (ranging from 0% to 6.83%). Our pooled analysis of the 27 studies in the review indicated that, compared to control, statin usage may slightly reduce the overall incidence of VTE (statins: 518/61,494 (0.84%); control: 594/61,107 (0.97%); OR 0.86, 95% CI 0.76 to 0.98; P = 0.03, I² = 4%; 27 studies, 122,601 participants; low‐certainty evidence; Analysis 1.1).

1.1. Analysis.

Comparison 1: Statin versus placebo, Outcome 1: Any VTE

Subgroup analysis for any VTE

Subgroup analysis by provoked status

The test for subgroup differences revealed evidence of a difference in the incidence of provoked or unprovoked VTE (P < 0.01; Analysis 1.2). Statins may reduce the incidence of unprovoked VTE (statins: 61/23,132 (0.26%), control: 104/22,793 (0.46%); OR 0.58, 95% CI 0.42 to 0.79; P < 0.01, I² = 0%; five studies, 45,925 participants) compared to control. Statins had minimal or no effect on the incidence of provoked VTE (statins: 457/38,362 (1.19%), control: 490/38,314 (1.28%); OR 0.93, 95% CI 0.82 to 1.06; P = 0.29, I² = 0%; 22 studies, 76,676 participants), compared to control. The provoking factors encompass coronary heart disease, vascular disease, type II diabetes mellitus, chronic kidney disease, heart failure, aortic valve stenosis, cerebrovascular disease, cancer, hypertension, COVID‐19, and sepsis‐associated respiratory distress syndrome.

1.2. Analysis.

Comparison 1: Statin versus placebo, Outcome 2: Subgroup analysis for any VTE stratified by provoked status

Subgroup analysis by age

The test for subgroup differences found no clear evidence of a difference in the effect sizes between participants aged 50 to 69 years and those aged 70 to 97 years (P = 0.45; Analysis 1.3). Likewise, there was scant evidence that statins reduced the incidence of VTE in participants aged 50 to 69 years (statins: 29/10,879 (0.27%), control: 45/10,628 (0.42%); OR 0.63, 95% CI 0.39 to 1.01; P = 0.06, I² = 0%; four studies, 21,507 participants). This was also the case for participants aged 70 to 97 years (statins: 43/5,712 (0.75%); control: 45/5,682 (0.79%); OR 0.92, 95% CI 0.39 to 2.20; P = 0.85, I² = 75%; two studies, 11,394 participants) in comparison to control.

1.3. Analysis.

Comparison 1: Statin versus placebo, Outcome 3: Subgroup analysis for any VTE stratified by age

Subgroup analysis by sex

The test for subgroup differences found no evidence of a difference in the effect sizes between men and women (P = 0.41; Analysis 1.4). Compared with control group, statins may reduce the incidence of all VTE in men (statins: 22/5,475 (0.40%); control: 44/5,526 (0.80%); OR 0.50, 95% CI 0.30 to 0.84; P < 0.01; one study, 11,001 participants), but there was no clear evidence of a difference in VTE in women (statins: 12/3426 (0.35%); control: 16/3375 (0.47%); OR 0.74, 95% CI 0.35 to 1.56; P = 0.43; one study, 6,801 participants).

1.4. Analysis.

Comparison 1: Statin versus placebo, Outcome 4: Subgroup analysis for any VTE stratified by sex

Subgroup analysis by disease

The test for subgroup differences found no evidence of a difference in the effect sizes across different types of disease (P = 0.26; Analysis 1.5). Statins likely reduce the incidence of VTE in the healthy population by 43% compared to control (statins: 34/8901 (0.38%), control: 60/8901 (0.67%); OR 0.57, 95% CI 0.37 to 0.86; P < 0.01; one study, 17,802 participants.) However, compared to control, statins may have minimal or no effect on the incidence of VTE in participants with hypercholesterolaemia (statins: 13/7870 (0.17%), control: 13/7548 (0.17%); OR 0.91, 95% CI 0.42 to 2.00; P = 0.82, I² = 0%; three studies, 15,418 participants), heart failure (statins: 27/4853 (0.56%), control: 39/4840 (0.81%); OR 0.69, 95% CI 0.42 to 1.15; P = 0.16, I² = 2%; three studies, 9693 participants), chronic kidney disease (statins: 35/2936 (1.19%), control: 34/2929 (1.16%); OR 1.03, 95% CI 0.64 to 1.66; P = 0.90, I² = 0%; four studies, 5865 participants), diabetes mellitus (statins: 31/2995 (1.04%), control: 49/2962 (1.65%); OR 0.63, 95% CI 0.37 to 1.07; P = 0.09, I² = 20%; three studies, 5,957 participants), aortic valve stenosis (statins: 0/211 (0%), control: 2/213 (1.13%); OR 0.33, 95% CI 0.03 to 3.23; P = 0.34, I² = 0%; two studies, 424 participants), cerebrovascular disease (statins: 36/2,679 (1.34%), control: 30/2,692, (1.11%); OR 1.20, 95% CI 0.74 to 1.96; P = 0.46, I² = 0%; two studies, 5,371 participants), vascular disease (statins: 196/13,103 (1.50%), control: 198/13,132 (1.51%); OR 1.06, 95% CI 0.74 to 1.51; P = 0.77, I² = 41%; two studies, 26,235 participants), and other diseases (statins: 146/17,946 (0.81%), control: 169/17,890 (0.94%); OR 0.81, 95% CI 0.60 to 1.09; P = 0.16, I² = 26%; seven studies, 35,836 participants).

1.5. Analysis.

Comparison 1: Statin versus placebo, Outcome 5: Subgroup analysis for any VTE stratified by disease

Subgroup analysis by type of statin

The test for subgroup differences found evidence of a difference in the effect sizes across different types of statins (P = 0.05; Analysis 1.6). However, this should be interpreted with caution as, in some instances, only a limited number of studies contributed to individual subgroups. Our subgroup analysis indicated that rosuvastatin might reduce the incidence of VTE by 39% (statins: 106/22,682 (0.47%), control: 162/22,210 (0.73%); OR 0.64, 95% CI 0.50 to 0.82; P < 0.01, I² = 0%; nine studies, 44,892 participants) compared to control. However, we found no evidence of a difference in VTE incidence between those who received atorvastatin (statins: 99/12,544 (0.79%), control: 122/12,513 (0.97%), OR 0.82, 95% CI 0.62 to 1.07; P = 0.15, I² = 0%; 11 studies, 25,057 participants); fluvastatin (statins: 23/1050 (2.19%), control: 23/1052 (2.19%); OR 1.00, 95% CI 0.56 to 1.80; P = 0.99; one study, 2102 participants); lovastatin (statins: 9/3304 (0.27%), control: 12/3301 (0.36%); OR 0.75, 95% CI 0.32 to 1.78; P = 0.51; one study, 6605 participants); pravastatin (statins: 113/11,645 (0.97%), control: 97/11,764 (0.82%); OR 1.16, 95% CI 0.89 to 1.53; P = 0.28, I² = 0%; four studies, 23,409 participants); and simvastatin (statins: 168/10,269 (1.64%), control: 178/10,267 (1.73%); OR 0.94, 95% CI 0.76 to 1.17; P = 0.59; one study, 20,536 participants) and control group.

1.6. Analysis.

Comparison 1: Statin versus placebo, Outcome 6: Subgroup analysis for any VTE stratified by type of statin

Subgroup analysis by dose

Rosuvastatin doses

The test for subgroup differences found no evidence of a difference in the effect sizes across different doses of rosuvastatin (P = 0.99; Analysis 1.7). We should interpret this with caution because a small number of studies contributed to the individual subgroups. Low‐dose rosuvastatin (10 to 20 mg daily) might reduce the incidence of VTE (statins: 105/21,848 (0.48%), control: 161/21,794 (0.74%); OR 0.65, 95% CI 0.50 to 0.85; P < 0.01, I² = 9%; seven studies, 43,642 participants) (Analysis 1.7). We found no clear evidence of a difference in VTE incidence in the groups receiving rosuvastatin 40 mg daily (statins: 1/834 (0.12%), control: 1/416 (0.24%), OR 0.64, 95% CI 0.07 to 6.13; P = 0.69, I² = 0%; two studies, 1,250 participants), compared to control.

1.7. Analysis.

Comparison 1: Statin versus placebo, Outcome 7: Subgroup analysis for any VTE statified by different doses of rosuvstatin

Atorvastatin doses

The test for subgroup differences found no evidence of a difference in the effect sizes across different doses of atorvastatin (P = 0.08; Analysis 1.8). Daily intake of 10 mg atorvastatin might reduce the incidence of VTE (statins: 32/7,608 (0.42%), control: 57/7,531 (0.76%); OR 0.55, 95% CI 0.36 to 0.86; P < 0.01, I² = 0%; four studies, 15,139 participants). We found no clear evidence of a difference in VTE incidence in the groups receiving 20 mg of atorvastatin daily (statins: 22/963 (2.28%), control: 24/987 (2.43%), OR 0.94, 95% CI 0.52 to 1.69; P = 0.83, I² = 0%; three studies, 1950 participants), or 80 mg of atorvastatin daily (statins: 45/3973 (1.13%), control: 41/3995 (1.03%), OR 1.10, 95% CI 0.72 to 1.69; P = 0.65, I² = 0%; four studies, 7968 participants), compared to control.

1.8. Analysis.

Comparison 1: Statin versus placebo, Outcome 8: Subgroup analysis for any VTE stratified by different doses of atorvastatin

Subgroup analysis by treatment duration

The test for subgroup differences revealed no evidence of statistical variance in effect sizes between participants with varying durations of statin treatment (P = 0.67; Analysis 1.9). A statin treatment duration longer than one year but shorter than five years likely reduces the incidence of VTE compared to control (statins: 188/31,390 (0.60%), control: 239/30,939 (0.77%); OR 0.79, 95% CI 0.65 to 0.97; P = 0.02, I² = 3%; 17 studies; 62,329 participants). However, a statin treatment duration exceeding five years (statins: 296/29,362 (1.00%), control: 319/29,432 (1.08%); OR 0.90, 95% CI 0.71 to 1.14; P = 0.39, I² = 30%; six studies, 58,794 participants) or less than one year (statins: 34/742 (2.02%), control: 36/736 (1.90%); OR 0.91, 95% CI 0.56 to 1.47; P = 0.69; four studies, 1,478 participants) likely do not reduce the incidence of VTE compared to control.

1.9. Analysis.

Comparison 1: Statin versus placebo, Outcome 9: Subgroup analysis for any VTE stratified by treatment duration

Sensitivity analysis for any VTE

Sensitivity analysis excluding studies that did not report the incidence of VTE in their original publications

We carried out sensitivity analysis excluding those studies encompassing unpublished data to evaluate the impact on the results. After omitting studies that did not disclose the incidence of VTE, a sensitivity analysis indicated that statins may have little to no effect on the incidence of VTE (statins: 153/23,737 (0.64%), control: 204/23,719 (0.86%); OR 0.77, 95% CI 0.57 to 1.03; P = 0.08, I² = 40%; nine studies, 47,456 participants; Analysis 1.10), compared to control.

1.10. Analysis.

Comparison 1: Statin versus placebo, Outcome 10: Sensitivity analysis for any VTE excluding unpublished data

Sensitivity analysis excluding studies that have a high risk of bias

After excluding studies with a high risk of bias, a sensitivity analysis revealed that statins likely reduce the incidence of VTE in studies with a low risk of bias (statins: 177/25,504 (0.69%); control: 231/25,450 (0.91%); OR 0.73, 95% CI 0.56 to 0.95; P = 0.02, I² = 30%; eight studies; 50,954 participants; Analysis 1.11), when compared to control.

1.11. Analysis.

Comparison 1: Statin versus placebo, Outcome 11: Sensitivity analysis for any VTE excluding studies with high risk of bias

Sensitivity analysis excluding studies that have a high proportion of missing outcome data

We deemed missing outcome data at 15% or more as potentially significant. After excluding studies with a high proportion of missing outcome data, a sensitivity analysis showed that statins may not reduce the incidence of VTE in those studies with less than 15% missing outcome data (statins: 236/30,106 (0.78%), control: 284/30,059 (0.94%); OR 0.81, 95% CI 0.65 to 1.02; P = 0.10, I² = 29%; 11 studies, 60,165 participants; Analysis 1.12), compared to control.

1.12. Analysis.

Comparison 1: Statin versus placebo, Outcome 12: Sensitivity analysis for any VTE excluding data from > 15% missing outcome data

DVT (provoked or unprovoked)

Six studies reported the incidence of DVT (Dinglas 2016; Fellström 2009; Freeman 2011; Glynn 2009; Talasaz 2023; Yusuf 2016). The time points for follow‐up ranged from one year to 5.6 years. The meta‐analysis demonstrated no significant difference in DVT incidence between the statin and control groups (statins: 69/20,154 (0.34%), control: 102/20,151 (0.51%); OR 0.70, 95% CI 0.41 to 1.18; P = 0.18, I² = 53%; six studies, 40,305 participants; low‐certainty evidence; Analysis 1.13).

1.13. Analysis.

Comparison 1: Statin versus placebo, Outcome 13: Subgroup analysis for DVT stratified by type of statin

Subgroup analysis for DVT

Subgroup analysis by type of statin

The test for subgroup differences revealed evidence of variance in effect sizes across different types of statins (P = 0.01; Analysis 1.13). Our subgroup analysis suggested that rosuvastatin might reduce the incidence of DVT by 52% (statins: 35/17,030 (0.21%), control: 73/16,989 (0.43%); OR 0.48, 95% CI 0.32 to 0.72; P < 0.01, I² = 0%; four studies, 34,019 participants), compared to control. We did not detect clear evidence of a difference in VTE incidence in the groups receiving atorvastatin (statins: 6/290 (2.07%), control: 9/297 (3.03%); OR 0.68, 95% CI 0.24 to 1.92; P = 0.46; one study, 587 participants), and pravastatin (statins: 28/2834 (0.99%), control: 20/2865 (0.70%); OR 1.42, 95% CI 0.80 to 2.53; P = 0.23; one study, 5,699 participants), compared to control.

Sensitivity analysis for DVT

Sensitivity analyses excluding each study

The statistical heterogeneity was high (I² = 53%, P = 0.06; Analysis 1.13). We conducted sensitivity analyses by excluding one study at a time. After excluding Freeman 2011, there was no statistical heterogeneity (I² = 0%, P = 0.76; Analysis 1.14), and the results of the meta‐analysis shifted in favor of statins (statins: 41/17,320 (0.24%), control: 82/17,286 (0.47%); OR 0.50, 95% CI 0.34 to 0.73; P < 0.01; five studies; 34,606 participants).

1.14. Analysis.

Comparison 1: Statin versus placebo, Outcome 14: DVT: sensitivity analysis

PE (provoked or unprovoked)

Five studies reported the incidence of PE (Dinglas 2016; Fellström 2004; Fellström 2009; Glynn 2009; Kjekshus 2007). The time points for follow‐up ranged from one year to 5.1 years. There was no clear evidence of a difference in the incidence of PE between the groups (statins: 31/14,233 (0.22%), control: 38/14,194 (0.27%); OR 0.83, 95% CI 0.46 to 1.52; P = 0.55, I² = 16%; five studies; 28,427 participants; low‐certainty evidence; Analysis 1.15).

1.15. Analysis.

Comparison 1: Statin versus placebo, Outcome 15: PE

Severe adverse events (SAEs) related to statin use

Thirteen studies reported the incidence of SAEs (Amarenco 2006; Chan 2010; Colhoun 2004; Crouse 2007; Dinglas 2016; Downs 1998; Feldman 2010; Fellström 2009; Glynn 2009; HPS study 2002; Kjekshus 2007; Knopp 2006; Koren 2004). The time points for follow‐up ranged from 1.5 to 5.2 years. SAEs are defined as life‐threatening events that result in death or permanent disability, lead to or prolong hospitalization, or involve a diagnosis of any form of cancer. The most commonly reported SAEs varied across the studies and are listed in Table 7. Statin usage may slightly reduce the incidence of any SAE compared to control (statins: 7,538/33,862 (22.26%), control: 7,617/33,158 (22.97%); OR 0.95, 95% CI 0.91 to 0.99; P = 0.02, I² = 0%; 13 studies, 67,020 participants; low‐certainty evidence; Analysis 1.16).

1.16. Analysis.

Comparison 1: Statin versus placebo, Outcome 16: Subgroup analysis for any serious adverse event stratified by type of statin

Subgroup analyses by types of statins

The subgroup differences test found no evidence of a variation in effect sizes across different types of statins (P = 0.61; Analysis 1.16). Atorvastatin usage may slightly reduce the incidence of any SAE compared to control (statins: 2011/6272 (32.06%), control: 2003/6032 (33.21%); OR 0.92, 95% CI 0.85 to 1.00; P = 0.05, I² = 0%; five studies, 12,304 participants). However, no clear evidence of a difference in the incidence of SAEs existed between other statin types and the control group: lovastatin (statins: 1131/3304 (34.23%), control: 1126/3301 (34.11%); OR 1.01, 95%CI 0.91 to 1.11; P = 0.92; one study, 6605 participants); rosuvastatin (statins: 4199/14,017 (29.96%), control: 4288/13,558 (31.63%); OR 0.94, 95% CI 0.89 to 1.00; P = 0.07, I² = 0%; six studies, 27,575 participants); and simvastatin (statins: 197/10,269 (1.92%), control: 200/10,267 (1.95%); OR 0.98, 95% CI 0.81 to 1.20; P = 0.88; one study, 20,536 participants).

Adverse events related to statin use

Myopathy

Ten studies reported the incidence of myopathy (Amarenco 2006; Chan 2010; Colhoun 2004; Crouse 2007; Fellström 2004; Glynn 2009; HPS study 2002; LIPID Study 1998; Tavazzi 2008; Yusuf 2016). The time points for follow‐up ranged from 1.9 to 6.1 years. There was no clear evidence of a difference in the incidence of myopathy between the groups (statins: 153/38,005 (0.40%), control: 90/37,546 (0.24%); OR 1.10, 95% CI 0.83 to 1.45; P = 0.51, I² = 0%; 10 studies, 75,551 participants; low‐certainty evidence; Analysis 1.17).

1.17. Analysis.

Comparison 1: Statin versus placebo, Outcome 17: Subgroup analysis for myopathy stratified by type of statin

Subgroup analysis by type of statin

The test for subgroup differences found no evidence of a difference in the effect sizes across different types of statins (P = 0.51; Analysis 1.17). Similar results were observed between each type of statin and the control group: atorvastatin (statins: 8/3,793 (0.21%), control: 8/3,775 (0.21%); OR 1.00, 95% CI 0.37 to 2.66; P = 1.00, I² = 0%; two studies, 7568 participants); rosuvastatin (statins: 124/18,381 (0.67%), control: 67/17,950 (0.37%); OR 1.06, 95% CI 0.77 to 1.46; P = 0.71, I² = 0%; five studies, 36,331 participants); pravastatin (statins: 8/4512 (0.18%); control: 10/4502 (0.22%); OR 0.80, 95% CI 0.31 to 2.02; P = 0.63; one study, 9014 participants); fluvastatin (statins: 3/1050 (0.29%), control: 1/1052 (0.10%); OR 3.01, 95% CI 0.31 to 29.00; P = 0.34; one study, 2102 participants); and simvastatin (statins: 10/10,269 (0.10%), control: 4/10,267 (0.04%); OR 2.50, 95% CI 0.78 to 7.98; P = 0.12; one study, 20,536 participants).

Rhabdomyolysis

Seven studies reported the incidence of rhabdomyolysis (Amarenco 2006; Downs 1998; Dinglas 2016; Glynn 2009; HPS study 2002; Knopp 2006; Yusuf 2016). The time points for follow‐up ranged from one year to 5.6 years. There was no evidence of a difference in the incidence of rhabdomyolysis between the groups (statins: 12/32,527 (0.04%), control: 9/32,461 (0.03%); OR 1.25, 95% CI 0.54 to 2.92; P = 0.60, I² = 0%; seven studies, 64,988 participants; low‐certainty evidence; Analysis 1.18).

1.18. Analysis.

Comparison 1: Statin versus placebo, Outcome 18: Subgroup analysis for rhabdomyolysis stratified by type of statin

Subgroup analysis by type of statin

The test for subgroup differences found no evidence of a difference in the effect sizes across different types of statins (P = 0.57; Analysis 1.18). Similar results were observed between each statin type and the control group: atorvastatin (statins: 3/3313 (0.09%), control: 4/3282 (0.12%); OR 0.74, 95% CI 0.17 to 3.35; P = 0.70; two studies, 6595 participants); rosuvastatin (statins: 3/15,641 (0.02%), control: 0/15,611 (0%); OR 2.97, 95% CI 0.47 to 18.83; P = 0.25, I² = 0%; three studies, 31,252 participants); lovastatin (statins: 1/3304 (0.03%), control: 2/3301 (0.06%); OR 0.50, 95% CI 0.05 to 5.51; P = 0.57; one study, 6,605 participants); and simvastatin (statins: 5/10,269 (0.05%), control: 3/10,267 (0.03%); OR 1.67, 95% CI 0.40 to 6.98; P = 0.48; one study, 20,536 participants).

Bleeding

Two studies reported the incidence of bleeding (Glynn 2009; Talasaz 2023). The time points for follow‐up ranged from 90 days to 1.9 years. There was no clear evidence of a difference in bleeding between the groups (statins: 275/9191 (2.99%), control: 288/9198 (3.13%); OR 0.95, 95% CI 0.81 to 1.13; P = 0.59, I² = 0%; two studies, 18,389 participants; moderate‐certainty evidence; Analysis 1.19).

1.19. Analysis.

Comparison 1: Statin versus placebo, Outcome 19: Bleeding

Muscular weakness, stiffness, or pain

Eleven studies reported the incidence of muscular weakness, stiffness, or pain (Amarenco 2006; Colhoun 2004; Crouse 2007; Downs 1998; Fellström 2004; Glynn 2009; Kjekshus 2007; Knopp 2006; Sever 2003; Tavazzi 2008; Yusuf 2016). The time points for follow‐up ranged from 1.9 to 5.6 years. There was no clear evidence supporting a difference in muscular weakness, stiffness, or pain between the statins and control groups (statins: 3276/34,952 (9.37%), control: 3080/34,433 (8.94%); OR 1.05, 95% CI 0.97 to 1.13; P = 0.22, I² = 28%; 11 studies, 69,385 participants; low‐certainty evidence; Analysis 1.20).

1.20. Analysis.

Comparison 1: Statin versus placebo, Outcome 20: Subgroup analysis for muscular weakness, stiffness, or pain, stratified by type of statin

Subgroup analysis by type of statin

The test for subgroup differences found no evidence of a difference in the effect sizes across different types of statins (P = 0.80; Analysis 1.20). Similar results were observed between each statin type and the control group: atorvastatin (statins: 602/9842 (6.12%), control: 593/9771 (6.07%); OR 1.02, 95% CI 0.82 to 1.25; P = 0.88, I² = 55%; four studies, 19,613 participants); rosuvastatin (statins: 2138/20,761 (10.30%), control: 1946/20,312 (9.58%); OR 1.09, 95% CI 0.98 to 1.20; P = 0.11, I² = 30%; five studies, 41,073 participants); lovastatin (statins: 10/3304 (0.30%), control: 10/3301 (0.30%); OR 1.00, 95% CI 0.42 to 2.40; P = 1.00; one study, 6,605 participants); and fluvastatin (statins: 526/1045 (50.33%), control: 531/1049 (50.62%); OR 0.99, 95% CI 0.83 to 1.17; P = 0.90; one study, 2094 participants).

Gastrointestinal disorders

Ten studies reported the incidence of gastrointestinal disorders (Amarenco 2006; Chan 2010; Cowell 2005; Dinglas 2016; Fellström 2004; Fellström 2009; Glynn 2009; Kjekshus 2007; Sever 2003; Tavazzi 2008). The time points for follow‐up ranged from one year to 5.1 years. There was no clear evidence of a difference detected in the incidence of gastrointestinal disorders between the groups (statins: 3262/24,195 (13.48%), control: 3200/24,141 (13.26%); OR 1.01, 95% CI 0.91 to 1.13; P = 0.83, I² = 53%; 10 studies, 48,336 participants; moderate‐certainty evidence; Analysis 1.21).

1.21. Analysis.

Comparison 1: Statin versus placebo, Outcome 21: Subgroup analysis for gastrointestinal disorders stratified by type of statin

Subgroup analysis by type of statin

The test for subgroup differences found no evidence of a difference in effect size across different types of statins (P = 0.21; Analysis 1.21). Similar results were observed between each statin type and the control group. This includes rosuvastatin (statins: 2163/15,602 (13.86%), control: 2178/15,566 (13.99%); OR 0.92, 95% CI 0.77 to 1.09; P = 0.32, I² = 55%; six studies, 31,168 participants); atorvastatin (statins: 537/7543, (7.12%), control: 481/7523, (6.39%); OR 1.15, 95% CI 0.91 to 1.46; P = 0.24, I² = 55%; three studies, 15,066 participants); and fluvastatin (statins: 562/1050 (53.52%), control: 541/1052 (51.43%); OR 1.09, 95% CI 0.92 to 1.29; P = 0.34; one study, 2102 participants).

Sensitivity analysis for gastrointestinal disorders

We conducted sensitivity analyses by excluding one study at a time. For gastrointestinal disorders, we identified Kjekshus 2007 as a possible source of heterogeneity; after we excluded this study, heterogeneity decreased (I² = 13%). However, the meta‐analysis results remained unchanged (statins: 3091/21,681 (14.26%), control: 2977/21,644 (13.75%); OR 1.06, 95% CI 0.98 to 1.14; P = 0.15, I² = 13%; nine studies, 43,325 participants; Analysis 1.22).

1.22. Analysis.

Comparison 1: Statin versus placebo, Outcome 22: Sensitivity analysis for gastrointestinal disorders

Hepatic disorders

Twenty studies reported the incidence of hepatic disorders (i.e. increased liver enzymes) (Amarenco 2006; Chan 2010; Colhoun 2004; Crouse 2007; Dinglas 2016; Downs 1998; Feldman 2010; Fellström 2004; Fellström 2009; Glynn 2009; HPS study 2002; Kjekshus 2007; Knopp 2006; Koren 2004; LIPID Study 1998; Nakamura 2006; Sever 2003; Talasaz 2023; Tavazzi 2008; Wanner 2005). The time points for follow‐up ranged from 90 days to 5.1 years. Statins may increase the incidence of hepatic disorder detection compared to control (statins: 929/51,851 (1.79%), control: 791/51,235 (1.54%); OR 1.27, 95% CI 1.03 to 1.56; P = 0.02, I² = 62%; 20 studies, 103,086 participants; very low‐certainty evidence; Analysis 1.23).

1.23. Analysis.

Comparison 1: Statin versus placebo, Outcome 23: Subgroup analysis for hepatic disorders stratified by type of statin

Subgroup analysis by types of statins

The test for subgroup differences revealed evidence of variance in effect sizes across different types of statins (P = 0.03; Analysis 1.23). However, this should be interpreted with caution as, in four of the six statin types, only one or two studies contributed to individual subgroups. Rosuvastatin may increase the incidence of hepatic disorder detection compared to control (statins: 330/16,302 (2.02%), control: 252/15,847 (1.59%); OR 1.48, 95% CI 1.08 to 2.03; P = 0.02, I² = 40%; seven studies, 32,149 participants); however, no clear evidence of a difference in hepatic disorders existed between other statin types and the control group: atorvastatin (statins: 109/12,545 (0.87%), control: 63/12,537 (0.50%); OR 1.57, 95% CI 0.75 to 3.31; P = 0.23, I² = 73%; eight studies, 25,082 participants); pravastatin (statins: 252/8381 (3.01%), control: 245/8231 (2.98%); OR 1.00, 95% CI 0.84 to 1.20; P = 0.96, I² = 0%; two studies, 16,612 participants); simvastatin (statins: 182/10,269 (1.77%), control: 163/10,267 (1.59%); OR 1.12, 95% CI 0.90 to 1.38; P = 0.30; one study, 20,536 participants); and lovastatin (statins: 18/3304 (0.54%), control: 11/3301 (0.33%); OR 1.64, 95% CI 0.77 to 3.47; P = 0.20; one study, 6605 participants). On the other hand, fluvastatin might have the potential to decrease the incidence of hepatic disorders (statins: 38/1050 (3.62%), control: 57/1052 (5.42%); OR 0.66, 95% CI 0.43 to 1.00; P = 0.05; one study, 2102 participants).

Sensitivity analysis

We conducted sensitivity analyses by excluding one study at a time. For hepatic disorders, one study may have contributed to the high heterogeneity (Amarenco 2006). After we excluded this study, heterogeneity was reduced (I² = 42%), and the meta‐analysis results changed, indicating that statin use may not be associated with increased risk of hepatic disorders (statins: 878/49,486 (1.77%), control: 780/48,869 (1.60%); OR 1.14, 95% CI 0.97 to 1.35; P = 0.12, I² = 42%; 19 studies, 98,355 participants; Analysis 1.24).

1.24. Analysis.

Comparison 1: Statin versus placebo, Outcome 24: Sensitivity analysis for hepatic disorders

Renal disorders

Eight studies reported the incidence of renal disorders (Chan 2010; Crouse 2007; Dinglas 2016; Fellström 2009; Glynn 2009; Kjekshus 2007; Sever 2003; Tavazzi 2008). The time points for follow‐up ranged from one year to 3.9 years. There was no clear evidence of a difference detected in the incidence of renal disorders between the groups (statins: 717/21,470 (3.34%), control: 640/20,984 (3.05%); OR 1.11, 95% CI 0.99 to 1.23; P = 0.07, I² = 0%; eight studies, 42,454 participants; moderate‐certainty evidence; Analysis 1.25).

1.25. Analysis.

Comparison 1: Statin versus placebo, Outcome 25: Subgroup analysis for renal disorders stratified by types of statins

Subgroup analysis by type of statin

The test for subgroup differences revealed no evidence of a difference in the effect size across various types of statins (P = 0.42; Analysis 1.25). Similar results were observed between each statin type and the control group: atorvastatin (statins: 97/5168 (1.88%), control: 78/5137 (1.52%); OR 1.24, 95% CI 0.92 to 1.68; P = 0.16; one study, 10,305 participants), and rosuvastatin (statins: 620/16,302 (3.80%), control: 562/15,847 (3.55%); OR 1.09, 95% CI 0.97 to 1.22; P = 0.16, I² = 0%; seven studies, 32,149 participants).

Mortality

All‐cause mortality

All but three studies reported mortality (Ades 2018; Freeman 2011; Sola 2006). The time points for follow‐up ranged from 90 days to 6.1 years. Pooling the data from 24 studies demonstrated that statin use may decrease the incidence of any death compared to control (statins: 5831/58,587 (9.95%), control: 6345/58,174 (10.91%); OR 0.90, 95% CI 0.86 to 0.95; P < 0.01, I² = 21%; 24 studies, 116,761 participants; low‐certainty evidence; Analysis 1.26).

1.26. Analysis.

Comparison 1: Statin versus placebo, Outcome 26: Subgroup analysis for mortality stratified by types of statins

Subgroup analysis by type of statin

The test for subgroup differences found evidence of a disparity in the effect size across different types of statins (P = 0.02; Analysis 1.26), but this result warrants cautious interpretation, as a small number of studies contributed to the individual subgroups. The use of atorvastatin did not show a difference in the incidence of any death compared to control (statins: 1057/12,490 (8.46%), control: 1144/12,458 (9.18%); OR 0.91, 95% CI 0.83 to 1.00; P = 0.06, I² = 0%; 10 studies, 24,948 participants). There was also no clear evidence of a difference with lovastatin (statins: 80/3304 (2.42%), control: 77/3301 (2.33%); OR 1.04, 95% CI 0.76 to 1.43; P = 0.81; one study, 6605 participants), rosuvastatin (statins: 2664/22,663 (11.74%), control: 2763/22,197 (12.45%); OR 0.94, 95% CI 0.88 to 1.01; P = 0.12, I² = 15%; eight studies, 44,860 participants), and fluvastatin (statins: 143/1050 (13.62%), control: 138/1052 (13.12%); OR 1.04, 95% CI 0.81 to 1.34; P = 0.74; one study, 2102 participants), compared to control. Pravastatin use probably lowered the incidence of any death (statins: 559/8811 (6.34%), control: 716/8899 (8.05%); OR 0.76, 95% CI 0.67 to 0.85; P < 0.01, I² = 0%; three studies, 17,710 participants), as compared to control. Simvastatin use likely reduced the incidence of any death (statins: 1328/10,269 (12.93%), control: 1507/10,267 (14.68%); OR 0.86, 95% CI 0.80 to 0.93; P < 0.01; one study, 20,536 participants), compared to control.

Mortality after PE

Three studies reported the incidence of mortality after PE (Fellström 2004; Glynn 2009; Kjekshus 2007). The time points for follow‐up ranged from 1.9 to 5.1 years. There was no discernible difference between the groups in the incidence of mortality following PE (statins: 7/12,465 (0.06%), control: 11/12,450 (0.09%); OR 0.71, 95% CI 0.13 to 3.82; P = 0.69, I² = 52%; three studies, 24,915 participants; Analysis 1.27).

1.27. Analysis.

Comparison 1: Statin versus placebo, Outcome 27: Mortality after PE

Sensitivity analysis for mortality after PE

We conducted sensitivity analyses by excluding one study at a time. High heterogeneity in the data concerning mortality after PE may have resulted from one particular study (Fellström 2004). Once we excluded this study, the heterogeneity decreased significantly (I² = 0%), leading to a change in the meta‐analysis results. These results suggested that the use of statins may not be associated with an increase in the incidence of mortality following PE (statins: 3/11,415 (0.03%), control: 10/11,398 (0.09%); OR 0.30, 95% CI 0.08 to 1.12; P = 0.07, I² = 0%; two studies, 22,813 participants; low‐certainty evidence; Analysis 1.28).

1.28. Analysis.

Comparison 1: Statin versus placebo, Outcome 28: Sensitivity analysis for mortality after PE

Publication bias

Our funnel plot for VTE, which comprised more than nine studies, exhibited notable asymmetry. This may indicate publication bias (Figure 4). We factored this into our GRADE assessment.

4.

Funnel plot for all cases of VTE

Abbreviations: log: logarithm; OR: odds ratio; SE: standard error; VTE: venous thromboembolism

Trial sequential analysis (TSA)

The results of the TSA for the incidence of VTE, are shown in Figure 5. The cumulative Z‐curve did not cross the trial sequential monitoring boundaries, even when the required information size was achieved. This indicates that despite a sufficiently large sample size, the evidence did not prove a statistically significant treatment effect.

5.

Trial sequential analysis results for the incidence of venous thromboembolism

Abbreviations: HPS: Heart Protection Study; LIPIDS: Long‐term Intervention with Pravastatin in Ischaemic Disease

Discussion

We present a summary of the key results in Table 1.

Summary of main results

We identified 27 RCTs (122,601 participants) that compared statins with control for the primary prevention of VTE. The median follow‐up durations ranged from 90 days to 6.1 years. Our analysis presented low‐certainty evidence that, compared to control, the use of statins for the primary prevention of VTE may slightly reduce its overall incidence (low‐certainty evidence); see Table 1.

Subgroup analysis suggested there may be a difference in the effect of statins on unprovoked and provoked VTE, with the effect clearly seen in studies of unprovoked VTE. Subgroup analysis suggested that the benefits of statins on the incidence of VTE depend on the specific statins prescribed. Notably, rosuvastatin might decrease the incidence of VTE compared to control, particularly for low‐dose rosuvastatin (10 to 20 mg daily). No clear evidence of a difference in VTE incidence was found between individuals who received fluvastatin, lovastatin, simvastatin, and pravastatin compared to control groups. There was no evidence of subgroup differences for age, sex, disease, dose, or treatment duration. However, all these findings from subgroup analysis should be interpreted with caution as they are exploratory.

No clear difference was detected in the incidence of DVT and PE between the statin and control groups, or in the incidence of the adverse effect of myopathy (all low‐certainty evidence). Further, the results were similar between groups for other adverse effects: rhabdomyolysis, bleeding, muscular weakness, stiffness or pain, gastrointestinal disorders, and renal disorders. Unfortunately, the meta‐analysis for hepatic disorders was unstable as it varied based on sensitivity analysis. Statin use might lower the incidence of any SAEs and death from any cause (both low‐certainty evidence) compared to control. However, there was no difference between statins and control in the incidence of death following PE.

Overall completeness and applicability of evidence

This review included a variety of participants, but the preventive effects of statins on VTE were confirmed by only three studies (Glynn 2009; Sola 2006; Tavazzi 2008). Glynn 2009 demonstrated that statins might reduce the incidence of VTE in initially healthy men and women. The median follow‐up was 1.9 years, and symptomatic VTE occurred in 94 participants, 34 in the rosuvastatin group and 60 in the control group. The VTE rates were 0.18 and 0.32 per 100 person‐years of follow‐up in the rosuvastatin and control groups, respectively. This suggests that statins reduced one or two VTE events amongst 1000 participants who took statins. Twenty‐six studies included participants with a medical history of coronary heart disease (Koren 2004; LIPID Study 1998) or cardiovascular risk factors (Yusuf 2016), vascular disease (Freeman 2011; HPS study 2002), diabetes mellitus (Colhoun 2004; Knopp 2006; Wanner 2005), chronic kidney disease (Asselbergs 2004; Fassett 2010; Fellström 2004; Fellström 2009), heart failure (Kjekshus 2007; Sola 2006; Tavazzi 2008), hypercholesterolemia (Crouse 2007; Downs 1998; Nakamura 2006), aortic valve stenosis (Chan 2010; Cowell 2005), cerebrovascular disease (Amarenco 2006; Feldman 2010), cancer (Adams 2013), hypertension (Sever 2003), COVID‐19 (Talasaz 2023), and sepsis‐associated acute respiratory distress syndrome (Dinglas 2016). Twenty‐four of these 26 studies did not confirm the preventive effects of statins on the incidence of VTE; only two studies, Kjekshus 2007 and Yusuf 2016, showed reductions in the incidence of VTE. Yusuf 2016 addressed intermediate‐risk participants without CVD, with risk factors including an elevated waist‐to‐hip ratio, recent or current smoking, low HDL cholesterol level, impaired fasting glucose or impaired glucose tolerance, early diabetes mellitus, family history of premature coronary heart disease, early renal dysfunction, and hypertension. Kjekshus 2007 included older participants with systolic heart failure and also demonstrated preventive effects for VTE. These participants had chronic New York Heart Association (NYHA) class II, III, or IV ischemic heart failure. The prevention effect of statins in participants with heart failure was not confirmed by another two studies (Sola 2006; Tavazzi 2008). Tavazzi 2008 included participants with chronic heart failure, which included ischemic, dilatative, and hypertensive causes. Sola 2006 included participants with heart failure due to nonischemic etiologies. Neither of these studies showed any reduction in VTE incidence (Sola 2006; Tavazzi 2008). This might be explained by the heterogeneity in the included participants.

Our review aimed to compare all types of statins with control. The statins included were atorvastatin, rosuvastatin, pravastatin, lovastatin, fluvastatin, pitavastatin, and simvastatin. Most studies focused on atorvastatin and rosuvastatin; we did not find any studies about pitavastatin. Of the statins we evaluated, only rosuvastatin seemed to be associated with a reduced incidence of VTE, albeit the reduction in incidence was very small. We also evaluated the effects of different doses of rosuvastatin (10 to 20 mg/day, and 40 mg/day) and atorvastatin (10 mg/day, 20 mg/day, and 80 mg/day). Rosuvastatin 40 mg daily did not show any preventive effects for VTE, but the pooled results of studies of low‐dose rosuvastatin showed VTE risk reductions. For atorvastatin, 10 mg daily reduced the incidence of VTE, while the other two doses of atorvastatin (20 mg/day and 80 mg/day) did not. Even though we included most statins, there was only one study each for lovastatin, fluvastatin, and simvastatin. Hence, the data on the preventive effects of the incidence of VTE for lovastatin, fluvastatin, and simvastatin were insufficient for a comparative effectiveness analysis.

This meta‐analysis incorporated both published and unpublished data on the preventive effect of statins on VTE. Of the 27 studies included, 18 failed to report the incidence of VTE. The under‐reporting of the incidence of VTE may lead to biased evaluation of the preventive effects of statins. We attempted to obtain data from the respective authors of the studies but received no responses. Therefore, data relating to the incidence of VTE in these 18 studies were retrieved from existing meta‐analyses (Kunutsor 2017; Rahimi 2012). A meta‐analysis by Stuijver 2015 found that incidents of VTE in RCTs are significantly under‐reported, not just in the general population but also amongst cancer patients and pregnant populations.

The duration of follow‐up varied; most studies tracked participants for more than one year but less than five years. The combined results indicated that participants who took statins for less than one year or more than five years did not experience a lower risk of VTE. However, for those who took statins for a duration exceeding one year but less than five years, there appeared to be an association with a reduced risk of VTE. It remains unclear why extended use of statins was not associated with a decreased incidence of VTE. Further evidence is needed to clarify these findings and explore the long‐term effects of statins use on VTE risk.

The populations in the included studies were from several countries across Asia, America, and Europe. For instance, Glynn 2009 investigated participants from 1315 different sites in 27 countries, such as Argentina, Belgium, Brazil, Bulgaria, and Canada. Knopp 2006 looked at participants from 70 centers across 14 countries, including Australia, Austria, Canada, Finland, France, Germany, Italy, the Netherlands, New Zealand, Norway, and a few others. Kjekshus 2007 took into account participants from 371 sites in 19 European countries, Russia, and South Africa. Feldman 2010 evaluated participants from 10 countries (USA, Germany, Canada, United Kingdom, Australia, Spain, South Africa, Sweden, Austria, and Denmark). Given the global representation, the meta‐analysis results are likely to be applicable to participants worldwide.

This review focused on the use of statins for the primary prevention of VTE only; studies assessing secondary prevention of VTE were excluded. The SAVER trial evaluated the use of rosuvastatin in secondary VTE prevention (Delluc 2022), finding that its adjuvant use did not reduce the risk of recurrent VTE in patients with symptomatic major VTE.

Quality of the evidence

We downgraded the certainty of the evidence due to concerns about the risk of bias, possible publication bias, inconsistency, and imprecision. All included studies were reported to be RCTs. However, only 20 studies reported details on randomization, while 12 studies reported details on concealed allocation. Twenty‐one studies provided details on blinding, and it was reported that the outcome assessors were blinded in 18 studies. In 16 studies, the proportion of missing data was deemed sufficiently high to possibly affect the study results. Although 18 studies evaluated the preventive effects of statins, they did not report the incidence of VTE despite having collected the data. Since all predefined outcomes in the protocols were reported, we considered the risk of bias for selective reporting to be low.

For VTE, any SAE, and all‐cause mortality, we downgraded the certainty of evidence to low due to the high risk of bias overall in some of the included studies. We observed that there might be publication bias for these outcomes. For DVT, the total number of events was less than 300, indicating imprecision in estimating the effect due to the small number of events, and high heterogeneity was observed across the studies. Hence, we downgraded the certainty from high to low. The studies that provided evidence for PE and myopathy were at high risk of bias and had fewer than 300 total events, leading us to downgrade the certainty from high to low. In summary, the overall certainty of the evidence is low (Table 1).

Potential biases in the review process

This review possesses several limitations that may potentially skew the results.

First, the most recent search was undertaken on 13 March 2023. While this might be considered a limitation, we have checked the status of the ongoing studies and all remain ongoing. Thus, we consider the review to be up to date. As mentioned, 18 studies did not report the incidence of VTE, and, despite reaching out to the authors for data, we received no feedback. Consequently, we sourced the necessary information from previously published meta‐analyses (Kunutsor 2017; Rahimi 2012). There may be a degree of bias due to the inclusion of these unpublished data.

Secondly, some studies may have been missed as not all findings were published. Indeed, our funnel plot indicated potential reporting bias.

Thirdly, we decided to pool data despite distinct clinical and methodological heterogeneity across the studies we included. For instance, we combined data from healthy participants with that of participants suffering from a variety of conditions, such as hypercholesterolemia, coronary heart disease, cardiovascular risk factors, vascular disease, diabetes mellitus, chronic kidney disease, heart failure, aortic valve stenosis, cerebrovascular disease, cancer, hypertension, COVID‐19, and sepsis‐associated respiratory distress syndrome. We are aware that patients in the intensive care unit (ICU) with COVID‐19 who have gone through extracorporeal membrane oxygenation (Talasaz 2023) or have sepsis‐associated acute respiratory distress syndrome (Dinglas 2016) are more susceptible to VTE. We also merged studies with different clinical and methodological variations, including different follow‐up durations and levels of risk of bias. Such clinical and methodological variations could result in disparities in the observed intervention effects.

Thirdly, we pooled studies that evaluated the effects of six different types of statins, given at different dosages and for different treatment durations. All these factors could exert various effects, leading to inconsistent results (Miksza 2019).

Finally, we conducted several subgroup analyses; some were pre‐planned, while others were added post‐hoc. This approach increases the potential for a type I error.

Agreements and disagreements with other studies or reviews

Our results are in line with other meta‐analyses that suggest statin use can lower the incidence of VTE (Agarwal 2010; Kunutsor 2017; Pai 2011; Ray 2003; Squizzato 2010; Zaccardi 2018). The meta‐analysis by Kunutsor 2017 incorporated 36 studies (13 cohort studies with 3,148,259 participants and 23 RCTs of statins versus placebo or no treatment with 118,464 participants). Their analysis concluded that observational studies (RR 0.75, 95% CI 0.65 to 0.87, P < 0.01) and interventional RCTs (RR 0.85, 95% CI 0.73 to 0.99, P = 0.04) indicate a beneficial effect of statin use on VTE when compared with placebo or no treatment. Our results are also in line with case‐control studies and cohort studies that indicate diminished risk of VTE with statins (Herrington 2002; Huerta 2007; Lacut 2004; Lacut 2008; Ramcharan 2009; Ray 2001b; Sørensen 2009). However, there are studies and systematic reviews that conflict with our results. Simvastatin was linked with a lowered risk of PE (OR 0.51, 95% CI 0.29 to 0.91), whereas pravastatin was not (OR 1.85, 95% CI 0.65 to 5.26) in Doggen 2004. According to a retrospective study (Yang 2002), statin use was not associated with a lower risk of idiopathic VTE (RR 0.8, 95% CI 0.3 to 2.7). Another meta‐analysis of published and unpublished data from RCTs indicated that assignment to statin therapy did not notably reduce the risk of VTE events (Rahimi 2012), revealing no heterogeneity between the effects on DVT and those on PE.

Class effect refers to the similar outcomes, therapeutic effects, and similar adverse effects of two or more medicines. The definition of drug class effect is based on three concepts: a similar chemical structure, a similar mechanism of action, or similar pharmacological effects (Soares 2002). Undas 2022 proposed that there might be a class effect for statins, making them capable of decreasing VTE risk. However, in our meta‐analysis, we did not observe the class effect on VTE. Subgroup analysis by type of statin in our study demonstrated that rosuvastatin might lower the incidence of VTE, unlike atorvastatin, fluvastatin, lovastatin, pravastatin, and simvastatin. We verified that rosuvastatin lowered the occurrence of VTE in a healthy population, consistent with a recent network meta‐analysis (Zaccardi 2018). This network meta‐analysis also revealed a greater benefit for rosuvastatin at any dosage (10, 20, and 40 mg) than other statins in the prevention of VTE. Joseph 2022 conducted an individual participant data meta‐analysis of Yusuf 2016 and Glynn 2009, which suggested that rosuvastatin use could lessen the incidence of VTE, and the relative effect of reducing VTE risk with rosuvastatin was fairly consistent across risk factor subgroups (Joseph 2022).

Authors' conclusions

Implications for practice.

This review indicated that statin use for the primary prevention of venous thromboembolism (VTE) might slightly lower VTE incidence. Statins may have little to no effect on the occurrence of deep vein thrombosis, pulmonary embolism, or myopathy when compared to control. Statin use may slightly decrease the incidence of death or serious adverse events. The current evidence base is of low certainty due to risk of bias, imprecision, and potential publication bias.

The incidence of VTE ranged from 0% to 10.53% (median 0.72%) amongst the study participants who received statins, and from 0% to 6.83% (median 0.89%) amongst those who did not. This indicates that 99.28% (89.47% to 100%) of participants were not diagnosed with VTE during the placebo treatment or usual care periods, and 99.11% (93.17% to 100%) of participants were not diagnosed with VTE during the statin treatment periods. The absolute difference in VTE incidence was 0.13%, which suggests that statins might prevent one case of VTE amongst 600 participants treated with statins. Therefore, for primary prevention of VTE, the effect is arguably too weak to be significant, which implies statins may not represent the optimal medicinal preventive measure for VTE, based on the limited and low‐certainty evidence available. The incidence of death was 9.83% (5723/58,208) in the statin groups and 10.82% (6254/57,808) in the control groups. Although statin use may reduce the incidence of death, the absolute difference in the incidence of death was 0.96%. This indicates that statins might save one life amongst 100 individuals.

Implications for research.

Most studies in our review did not report the incidence of VTE in their publications. In future, if possible, randomised controlled trials (RCTs) that investigate statins for the primary prevention of any diseases should evaluate and report the incidence of VTE in their results. Studies that are already completed should re‐evaluate and report any incidences of VTE. No RCTs have evaluated the preventive effects of pitavastatin on VTE. Hence, determining whether pitavastatin can reduce the incidence of VTE should be considered. Trials that evaluate VTE as a primary endpoint should be large‐scale and should be conducted over a reasonable duration (at least a year, based on our meta‐analysis). Furthermore, we recommend the conduct of prospective studies that scrupulously investigate the underlying mechanisms of the effects of statins, especially rosuvastatin and atorvastatin, on the prevention of VTE. Some studies in this review did not report details on random sequence generation, allocation concealment methods, blinding of participants, blinding of outcome assessment, and outcome data; future studies should provide thorough reporting on these methodological aspects.

History

Protocol first published: Issue 6, 2021

Notes

Parts of the Methods section are based on a standard template established by Cochrane Vascular.

Appendices

Appendix 1. Sources searched and search strategies

Source	Search strategy	Hits retrieved
1. Cochrane Vascular Specialised Register via Cochrane Register of Studies (CRS) (Date of most recent search: 13 March 2023)	#1 Simvastatin OR Hydroxymethylglutaryl‐CoA Reductase Inhibitors OR Atorvastatin OR Fluvastatin OR Pravastatin OR 3‐hydroxy‐3‐methylglutaryl HMG coenzyme A CoA reductase inhibitors OR HMG‐CoA reductase inhibitors OR lovastatin OR pitavastatin OR rosuvastatin OR Statin* AND INREGISTER	Aug 2021: 320 June 2022: 20 March 2023: 15
2. Cochrane Central Register of Controlled Trials (CENTRAL) via CRSO (Date of most recent search: 13 March 2023)	#1 MESH DESCRIPTOR Pulmonary Embolism EXPLODE ALL TREES 1033 #2 MESH DESCRIPTOR Thromboembolism EXPLODE ALL TREES 2109 #3 MESH DESCRIPTOR Thrombosis EXPLODE ALL TREES 4852 #4 MESH DESCRIPTOR Venous Thromboembolism EXPLODE ALL TREES 675 #5 MESH DESCRIPTOR Venous Thrombosis EXPLODE ALL TREES 2708 #6 ((vein* or ven*) adj thromb*):TI,AB,KY 11559 #7 (blood adj3 clot*):TI,AB,KY 5480 #8 (deep vein thrombosis):TI,AB,KY 4925 #9 (lung adj3 clot*):TI,AB,KY 12 #10 (DVT or VTE):TI,AB,KY 3802 #11 (peripheral vascular thrombosis):TI,AB,KY 0 #12 (post‐thrombotic syndrome):TI,AB,KY 231 #13 (pulmonary embolism):TI,AB,KY 3374 #14 (pulmonary adj3 clot*):TI,AB,KY 17 #15 (thrombus* or thrombopro* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or embol* or microembol*):TI,AB,KY 33472 #16 (venous thromboembolism):TI,AB,KY 3830 #17 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 38489 #18 MESH DESCRIPTOR Simvastatin EXPLODE ALL TREES 1805 #19 MESH DESCRIPTOR Hydroxymethylglutaryl‐CoA Reductase Inhibitors EXPLODE ALL TREES 6524 #20 MESH DESCRIPTOR Atorvastatin EXPLODE ALL TREES 1762 #21 MESH DESCRIPTOR Fluvastatin EXPLODE ALL TREES 325 #22 MESH DESCRIPTOR Pravastatin EXPLODE ALL TREES 1014 #23 MESH DESCRIPTOR Rosuvastatin Calcium EXPLODE ALL TREES 1128 #24 (3‐hydroxy‐3‐methylglutaryl HMG coenzyme A CoA reductase inhibitors):TI,AB,KY 0 #25 atorvastatin:TI,AB,KY 5431 #26 fluvastatin:TI,AB,KY 689 #27 (HMG‐CoA reductase inhibitors):TI,AB,KY 413 #28 (Hydroxymethylglutaryl‐CoA Reductase Inhibitors):TI,AB,KY 3643 #29 lovastatin:TI,AB,KY 915 #30 pitavastatin:TI,AB,KY 512 #31 pravastatin:TI,AB,KY 1924 #32 rosuvastatin:TI,AB,KY 2535 #33 simvastatin:TI,AB,KY 3771 #34 Statin*:TI,AB,KY 9673 #35 #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 17577 #36 #17 AND #35 487	Aug 2021: 487 June 2022: 34 March 2023: 35
3. MEDLINE (Ovid MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE) 1946 to present (Date of most recent search: 13 March 2023)	1 Pulmonary Embolism/ 2 Thromboembolism / 3 Thrombosis/ 4 exp Venous Thromboembolism/ 5 exp Venous Thrombosis/ 6 ((vein* or ven*) adj thromb*).ti,ab. 7 (blood adj3 clot*).ti,ab. 8 deep vein thrombosis.ti,ab. 9 (lung adj3 clot*).ti,ab. 10 (DVT or VTE).ti,ab. 11 peripheral vascular thrombosis.ti,ab. 12 post‐thrombotic syndrome.ti,ab. 13 pulmonary embolism.ti,ab. 14 (pulmonary adj3 clot*).ti,ab. 15 (thrombus* or thrombopro* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or embol* or microembol*).ti,ab. 16 venous thromboembolism.ti,ab. 17 or/1‐16 18 Simvastatin/ 19 exp Hydroxymethylglutaryl‐CoA Reductase Inhibitors/ 20 Atorvastatin/ 21 Fluvastatin/ 22 Pravastatin/ 23 Rosuvastatin Calcium/ 24 "3‐hydroxy‐3‐methylglutaryl (HMG)‐coenzyme A (CoA) reductase inhibitors".ti,ab. 25 atorvastatin.ti,ab. 26 fluvastatin.ti,ab. 27 "HMG‐CoA reductase inhibitors".ti,ab. 28 "Hydroxymethylglutaryl‐CoA Reductase Inhibitors".ti,ab. 29 lovastatin.ti,ab. 30 pitavastatin.ti,ab. 31 pravastatin.ti,ab. 32 rosuvastatin.ti,ab. 33 simvastatin.ti,ab. 34 Statin*.ti,ab. 35 or/18‐34 36 17 and 35 37 randomized controlled trial.pt. 38 controlled clinical trial.pt. 39 randomized.ab. 40 placebo.ab. 41 drug therapy.fs. 42 randomly.ab. 43 trial.ab. 44 groups.ab. 45 or/37‐44 46 exp animals/ not humans.sh. 47 45 not 46 48 36 and 47	Aug 2021: 1032 June 2022: 98 March 2023: 67
4. Embase via OVID (Date of most recent search: 13 March 2023)	1 lung embolism/ 2 thromboembolism/ 3 thrombosis/ 4 exp venous thromboembolism/ 5 exp vein thrombosis/ 6 ((vein* or ven*) adj thromb*).ti,ab. 7 (blood adj3 clot*).ti,ab. 8 deep vein thrombosis.ti,ab. 9 (lung adj3 clot*).ti,ab. 10 (DVT or VTE).ti,ab. 11 peripheral vascular thrombosis.ti,ab. 12 post‐thrombotic syndrome.ti,ab. 13 pulmonary embolism.ti,ab. 14 (pulmonary adj3 clot*).ti,ab. 15 (thrombus* or thrombopro* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or embol* or microembol*).ti,ab. 16 venous thromboembolism.ti,ab. 17 or/1‐16 18 simvastatin/ 19 exp Hydroxymethylglutaryl‐CoA Reductase Inhibitors/ 20 atorvastatin/ 21 Fluvastatin/ 22 Pravastatin/ 23 Rosuvastatin Calcium/ 24 "3‐hydroxy‐3‐methylglutaryl (HMG)‐coenzyme A (CoA) reductase inhibitors".ti,ab. 25 atorvastatin.ti,ab. 26 fluvastatin.ti,ab. 27 "HMG‐CoA reductase inhibitors".ti,ab. 28 "Hydroxymethylglutaryl‐CoA Reductase Inhibitors".ti,ab. 29 lovastatin.ti,ab. 30 pitavastatin.ti,ab. 31 pravastatin.ti,ab. 32 rosuvastatin.ti,ab. 33 simvastatin.ti,ab. 34 Statin*.ti,ab. 35 or/18‐34 36 17 and 35 37 randomized controlled trial/ 38 controlled clinical trial/ 39 random$.ti,ab. 40 randomisation/ 41 intermethod comparison/ 42 placebo.ti,ab. 43 (compare or compared or comparison).ti. 44 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 45 (open adj label).ti,ab. 46 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 47 double blind procedure/ 48 parallel group$1.ti,ab. 49 (crossover or cross over).ti,ab. 50 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 51 (assigned or allocated).ti,ab. 52 (controlled adj7 (study or design or trial)).ti,ab. 53 (volunteer or volunteers).ti,ab. 54 trial.ti. 55 or/37‐54 56 36 and 55	Aug 2021: 2462 June 2022: 362 March 2023: 333
5. CINAHL (Cumulative Index to Nursing and Allied Health Literature) via EBSCO (Date of most recent search: 13 March 2023)	S52 S36 AND S51 S51 S37 OR S38 OR S39 OR S40 OR S41 OR S42 OR S43 OR S44 OR S45 OR S46 OR S47 OR S48 OR S49 OR S50 S50 MH "Random Assignment" S49 MH "Triple‐Blind Studies" S48 MH "Double‐Blind Studies" S47 MH "Single‐Blind Studies" S46 MH "Crossover Design" S45 MH "Factorial Design" S44 MH "Placebos" S43 MH "Clinical Trials" S42 TX "multi‐centre study" OR "multi‐center study" OR "multicentre study" OR "multicenter study" OR "multi‐site study" S41 TX crossover OR "cross‐over" S40 AB placebo* S39 TX random* S38 TX trial* S37 TX "latin square" S36 S17 AND S35 S35 S18 OR S19 OR S20 OR S21 OR S22 OR S23 OR S24 OR S25 OR S26 OR S27 OR S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 S34 TX Statin* S33 TX simvastatin S32 TX rosuvastatin S31 TX pravastatin S30 TX pitavastatin S29 TX lovastatin S28 TX "Hydroxymethylglutaryl‐CoA Reductase Inhibitors" S27 TX "HMG‐CoA reductase inhibitors" S26 TX fluvastatin S25 TX atorvastatin S24 TX "3‐hydroxy‐3‐methylglutaryl (HMG)‐coenzyme A (CoA) reductase inhibitors" S23 (MH "Rosuvastatin") S22 (MH "Pravastatin") S21 (MH "Fluvastatin") S20 (MH "Atorvastatin+") S19 TX "Hydroxymethylglutaryl‐CoA Reductase Inhibitors" S18 (MH "Simvastatin+") S17 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 S16 TX "venous thromboembolism" S15 TX thrombus* or thrombopro* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or embol* or microembol* S14 TX pulmonary N3 clot* S13 TX "pulmonary embolism" S12 TX "post‐thrombotic syndrome" S11 TX peripheral vascular thrombosis S10 TX DVT or VTE S9 TX lung N3 clot* S8 TX "deep vein thrombosis" S7 TX blood N3 clot* S6 TX ((vein* or ven*) adj thromb*) S5 (MH "Venous Thrombosis+") S4 (MH "Venous Thromboembolism") S3 (MH "Thrombosis") S2 (MH "Thromboembolism") S1 (MH "Pulmonary Embolism")	Aug 2021: 194 June 2022: 7 March 2023: 5
6. Clinicaltrials.gov (clinicaltrials.gov) (Date of most recent search: 13 March 2023)	venous thromboembolism OR Pulmonary Embolism OR Thromboembolism OR Thrombosis OR Thromboembolism OR DVT or VTE | Simvastatin OR Hydroxymethylglutaryl‐CoA Reductase Inhibitors OR Atorvastatin OR Fluvastatin OR Pravastatin OR HMG‐CoA reductase inhibitors OR lovastatin OR pitavastatin OR rosuvastatin OR Statin*	Aug 2021: 24 June 2022: 3 March 2023: 0
7. World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (who.int/trialsearch) (Date of most recent search: 10 August 2021. Database not available 13 March 2023)	venous thromboembolism AND Statin*	Aug 2021: 11 June 2022: N/A March 2023: N/A
TOTAL before de‐duplication		Aug 2021: 4530 June 2022: 524 March 2023: 455
TOTAL after de‐duplication		Aug 2021: 3672 June 2022: 461 March 2023: 396

Data and analyses

Comparison 1. Statin versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Any VTE	27	122601	Odds Ratio (M‐H, Random, 95% CI)	0.86 [0.76, 0.98]
1.2 Subgroup analysis for any VTE stratified by provoked status	27	122601	Odds Ratio (M‐H, Random, 95% CI)	0.86 [0.76, 0.98]
1.2.1 Unprovoked VTE	5	45925	Odds Ratio (M‐H, Random, 95% CI)	0.58 [0.42, 0.79]
1.2.2 Provoked VTE	22	76676	Odds Ratio (M‐H, Random, 95% CI)	0.93 [0.82, 1.06]
1.3 Subgroup analysis for any VTE stratified by age	5	32901	Odds Ratio (M‐H, Random, 95% CI)	0.81 [0.51, 1.27]
1.3.1 Age 50 to 69 years	4	21507	Odds Ratio (M‐H, Random, 95% CI)	0.63 [0.39, 1.01]
1.3.2 Age 70 to 97 years	2	11394	Odds Ratio (M‐H, Random, 95% CI)	0.92 [0.39, 2.20]
1.4 Subgroup analysis for any VTE stratified by sex	1	17802	Odds Ratio (M‐H, Random, 95% CI)	0.57 [0.37, 0.87]
1.4.1 Men	1	11001	Odds Ratio (M‐H, Random, 95% CI)	0.50 [0.30, 0.84]
1.4.2 Women	1	6801	Odds Ratio (M‐H, Random, 95% CI)	0.74 [0.35, 1.56]
1.5 Subgroup analysis for any VTE stratified by disease	27	122601	Odds Ratio (M‐H, Random, 95% CI)	0.86 [0.76, 0.98]
1.5.1 Healthy	1	17802	Odds Ratio (M‐H, Random, 95% CI)	0.57 [0.37, 0.86]
1.5.2 Hypercholesterolaemia	3	15418	Odds Ratio (M‐H, Random, 95% CI)	0.91 [0.42, 2.00]
1.5.3 Heart failure	3	9693	Odds Ratio (M‐H, Random, 95% CI)	0.69 [0.42, 1.15]
1.5.4 Chronic kidney disease	4	5865	Odds Ratio (M‐H, Random, 95% CI)	1.03 [0.64, 1.66]
1.5.5 Diabetes mellitus	3	5957	Odds Ratio (M‐H, Random, 95% CI)	0.63 [0.37, 1.07]
1.5.6 Aortic valve stenosis	2	424	Odds Ratio (M‐H, Random, 95% CI)	0.33 [0.03, 3.23]
1.5.7 Cerebrovascular disease	2	5371	Odds Ratio (M‐H, Random, 95% CI)	1.20 [0.74, 1.96]
1.5.8 Vascular disease	2	26235	Odds Ratio (M‐H, Random, 95% CI)	1.06 [0.74, 1.51]
1.5.9 Other	7	35836	Odds Ratio (M‐H, Random, 95% CI)	0.81 [0.60, 1.09]
1.6 Subgroup analysis for any VTE stratified by type of statin	27	122601	Odds Ratio (M‐H, Random, 95% CI)	0.86 [0.76, 0.98]
1.6.1 Atorvastatin	11	25057	Odds Ratio (M‐H, Random, 95% CI)	0.82 [0.62, 1.07]
1.6.2 Fluvastatin	1	2102	Odds Ratio (M‐H, Random, 95% CI)	1.00 [0.56, 1.80]
1.6.3 Lovastatin	1	6605	Odds Ratio (M‐H, Random, 95% CI)	0.75 [0.32, 1.78]
1.6.4 Pravastatin	4	23409	Odds Ratio (M‐H, Random, 95% CI)	1.16 [0.89, 1.53]
1.6.5 Rosuvastatin	9	44892	Odds Ratio (M‐H, Random, 95% CI)	0.64 [0.50, 0.82]
1.6.6 Simvastation	1	20536	Odds Ratio (M‐H, Random, 95% CI)	0.94 [0.76, 1.17]
1.7 Subgroup analysis for any VTE statified by different doses of rosuvstatin	9	44892	Odds Ratio (M‐H, Random, 95% CI)	0.64 [0.50, 0.82]
1.7.1 10‐20 mg/day	7	43642	Odds Ratio (M‐H, Random, 95% CI)	0.65 [0.50, 0.85]
1.7.2 40 mg/day	2	1250	Odds Ratio (M‐H, Random, 95% CI)	0.64 [0.07, 6.13]
1.8 Subgroup analysis for any VTE stratified by different doses of atorvastatin	11	25057	Odds Ratio (M‐H, Random, 95% CI)	0.82 [0.62, 1.07]
1.8.1 10 mg/day	4	15139	Odds Ratio (M‐H, Random, 95% CI)	0.55 [0.36, 0.86]
1.8.2 20 mg/day	3	1950	Odds Ratio (M‐H, Random, 95% CI)	0.94 [0.52, 1.69]
1.8.3 80 mg/day	4	7968	Odds Ratio (M‐H, Random, 95% CI)	1.10 [0.72, 1.69]
1.9 Subgroup analysis for any VTE stratified by treatment duration	27	122601	Odds Ratio (M‐H, Random, 95% CI)	0.86 [0.76, 0.98]
1.9.1 Short term	4	1478	Odds Ratio (M‐H, Random, 95% CI)	0.91 [0.56, 1.47]
1.9.2 Medium‐term	17	62329	Odds Ratio (M‐H, Random, 95% CI)	0.79 [0.65, 0.97]
1.9.3 Long‐term	6	58794	Odds Ratio (M‐H, Random, 95% CI)	0.90 [0.71, 1.14]
1.10 Sensitivity analysis for any VTE excluding unpublished data	9	47456	Odds Ratio (M‐H, Random, 95% CI)	0.77 [0.57, 1.03]
1.11 Sensitivity analysis for any VTE excluding studies with high risk of bias	8	50954	Odds Ratio (M‐H, Random, 95% CI)	0.73 [0.56, 0.95]
1.12 Sensitivity analysis for any VTE excluding data from > 15% missing outcome data	11	60165	Odds Ratio (M‐H, Random, 95% CI)	0.81 [0.65, 1.02]
1.13 Subgroup analysis for DVT stratified by type of statin	6	40305	Odds Ratio (M‐H, Random, 95% CI)	0.70 [0.41, 1.18]
1.13.1 Rosuvastatin	4	34019	Odds Ratio (M‐H, Random, 95% CI)	0.48 [0.32, 0.72]
1.13.2 Atorvastatin	1	587	Odds Ratio (M‐H, Random, 95% CI)	0.68 [0.24, 1.92]
1.13.3 Pravastatin	1	5699	Odds Ratio (M‐H, Random, 95% CI)	1.42 [0.80, 2.53]
1.14 DVT: sensitivity analysis	5	34606	Odds Ratio (M‐H, Random, 95% CI)	0.50 [0.34, 0.73]
1.15 PE	5	28427	Odds Ratio (M‐H, Random, 95% CI)	0.83 [0.46, 1.52]
1.16 Subgroup analysis for any serious adverse event stratified by type of statin	13	67020	Odds Ratio (M‐H, Random, 95% CI)	0.95 [0.91, 0.99]
1.16.1 Atorvastatin	5	12304	Odds Ratio (M‐H, Random, 95% CI)	0.92 [0.85, 1.00]
1.16.2 Lovastatin	1	6605	Odds Ratio (M‐H, Random, 95% CI)	1.01 [0.91, 1.11]
1.16.3 Rosuvastatin	6	27575	Odds Ratio (M‐H, Random, 95% CI)	0.94 [0.89, 1.00]
1.16.4 Simvastatin	1	20536	Odds Ratio (M‐H, Random, 95% CI)	0.98 [0.81, 1.20]
1.17 Subgroup analysis for myopathy stratified by type of statin	10	75551	Odds Ratio (M‐H, Random, 95% CI)	1.10 [0.83, 1.45]
1.17.1 Atorvastatin	2	7568	Odds Ratio (M‐H, Random, 95% CI)	1.00 [0.37, 2.66]
1.17.2 Rosuvastatin	5	36331	Odds Ratio (M‐H, Random, 95% CI)	1.06 [0.77, 1.46]
1.17.3 Pravastatin	1	9014	Odds Ratio (M‐H, Random, 95% CI)	0.80 [0.31, 2.02]
1.17.4 Fluvastatin	1	2102	Odds Ratio (M‐H, Random, 95% CI)	3.01 [0.31, 29.00]
1.17.5 Simvastatin	1	20536	Odds Ratio (M‐H, Random, 95% CI)	2.50 [0.78, 7.98]
1.18 Subgroup analysis for rhabdomyolysis stratified by type of statin	7	64988	Odds Ratio (M‐H, Random, 95% CI)	1.25 [0.54, 2.92]
1.18.1 Atorvastatin	2	6595	Odds Ratio (M‐H, Random, 95% CI)	0.74 [0.17, 3.35]
1.18.2 Rosuvastatin	3	31252	Odds Ratio (M‐H, Random, 95% CI)	2.97 [0.47, 18.83]
1.18.3 Lovastatin	1	6605	Odds Ratio (M‐H, Random, 95% CI)	0.50 [0.05, 5.51]
1.18.4 Simvastatin	1	20536	Odds Ratio (M‐H, Random, 95% CI)	1.67 [0.40, 6.98]
1.19 Bleeding	2	18389	Odds Ratio (M‐H, Random, 95% CI)	0.95 [0.81, 1.13]
1.20 Subgroup analysis for muscular weakness, stiffness, or pain, stratified by type of statin	11	69385	Odds Ratio (M‐H, Random, 95% CI)	1.05 [0.97, 1.13]
1.20.1 Atorvastatin	4	19613	Odds Ratio (M‐H, Random, 95% CI)	1.02 [0.82, 1.25]
1.20.2 Rosuvastatin	5	41073	Odds Ratio (M‐H, Random, 95% CI)	1.09 [0.98, 1.20]
1.20.3 Lovastatin	1	6605	Odds Ratio (M‐H, Random, 95% CI)	1.00 [0.42, 2.40]
1.20.4 Fluvastatin	1	2094	Odds Ratio (M‐H, Random, 95% CI)	0.99 [0.83, 1.17]
1.21 Subgroup analysis for gastrointestinal disorders stratified by type of statin	10	48336	Odds Ratio (M‐H, Random, 95% CI)	1.01 [0.91, 1.13]
1.21.1 Rosuvastatin	6	31168	Odds Ratio (M‐H, Random, 95% CI)	0.92 [0.77, 1.09]
1.21.2 Atorvastatin	3	15066	Odds Ratio (M‐H, Random, 95% CI)	1.15 [0.91, 1.46]
1.21.3 Fluvastatin	1	2102	Odds Ratio (M‐H, Random, 95% CI)	1.09 [0.92, 1.29]
1.22 Sensitivity analysis for gastrointestinal disorders	9	43325	Odds Ratio (M‐H, Random, 95% CI)	1.06 [0.98, 1.14]
1.23 Subgroup analysis for hepatic disorders stratified by type of statin	20	103086	Odds Ratio (M‐H, Random, 95% CI)	1.27 [1.03, 1.56]
1.23.1 Fluvastatin	1	2102	Odds Ratio (M‐H, Random, 95% CI)	0.66 [0.43, 1.00]
1.23.2 Atorvastatin	8	25082	Odds Ratio (M‐H, Random, 95% CI)	1.57 [0.75, 3.31]
1.23.3 Rosuvastatin	7	32149	Odds Ratio (M‐H, Random, 95% CI)	1.48 [1.08, 2.03]
1.23.4 Pravastatin	2	16612	Odds Ratio (M‐H, Random, 95% CI)	1.00 [0.84, 1.20]
1.23.5 Simvastatin	1	20536	Odds Ratio (M‐H, Random, 95% CI)	1.12 [0.90, 1.38]
1.23.6 Lovastatin	1	6605	Odds Ratio (M‐H, Random, 95% CI)	1.64 [0.77, 3.47]
1.24 Sensitivity analysis for hepatic disorders	19	98355	Odds Ratio (M‐H, Random, 95% CI)	1.14 [0.97, 1.35]
1.25 Subgroup analysis for renal disorders stratified by types of statins	8	42454	Odds Ratio (M‐H, Random, 95% CI)	1.11 [0.99, 1.23]
1.25.1 Atorvastatin	1	10305	Odds Ratio (M‐H, Random, 95% CI)	1.24 [0.92, 1.68]
1.25.2 Rosuvastatin	7	32149	Odds Ratio (M‐H, Random, 95% CI)	1.09 [0.97, 1.22]
1.26 Subgroup analysis for mortality stratified by types of statins	24	116761	Odds Ratio (M‐H, Random, 95% CI)	0.90 [0.86, 0.95]
1.26.1 Atorvastatin	10	24948	Odds Ratio (M‐H, Random, 95% CI)	0.91 [0.83, 1.00]
1.26.2 Lovastatin	1	6605	Odds Ratio (M‐H, Random, 95% CI)	1.04 [0.76, 1.43]
1.26.3 Rosuvastatin	8	44860	Odds Ratio (M‐H, Random, 95% CI)	0.94 [0.88, 1.01]
1.26.4 Fluvastatin	1	2102	Odds Ratio (M‐H, Random, 95% CI)	1.04 [0.81, 1.34]
1.26.5 Pravastatin	3	17710	Odds Ratio (M‐H, Random, 95% CI)	0.76 [0.67, 0.85]
1.26.6 Simvastation	1	20536	Odds Ratio (M‐H, Random, 95% CI)	0.86 [0.80, 0.93]
1.27 Mortality after PE	3	24915	Odds Ratio (M‐H, Random, 95% CI)	0.71 [0.13, 3.82]
1.28 Sensitivity analysis for mortality after PE	2	22813	Odds Ratio (M‐H, Random, 95% CI)	0.30 [0.08, 1.12]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Ades 2018.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo Endpoint classification: safety and efficacy study Intervention model: cross‐over assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 9 weeks Setting: hospital
Participants	Ages eligible for study: > 18 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 38 Location: USA Inclusion criteria: "adult patients at the University of Vermont Cancer Center with locally advanced or metastatic cancer prior to or during any systemic chemotherapy or targeted therapy. Individuals who had an estimated overall survival of greater than 6 months and an anticipated duration of therapy of at least 3 months were enrolled." Exclusion criteria: "current antithrombotic or statin therapy, systemic treatment with hormonal therapy alone, potential for concurrent medication interaction with rosuvastatin, thalidomide or lenalidomide‐containing regimens, adjuvant therapy, known statin intolerance, transaminase levels above three times the upper limit of normal, GFR < 40 mL min‐1 and Asian descent, because of slower rosuvastatin metabolism and concerns regarding toxicity at the 20‐mg dose level."
Interventions	Intervention: rosuvastatin 20 mg/day over a 3‐ to 4‐week treatment period before a 3‐ to 5‐week washout period then cross over to the alternative therapy (n = 19) Control: matching placebo over a 3‐ to 4‐week treatment period before a 3‐ to 5‐week washout period then cross over to the alternative therapy (n = 19)
Outcomes	Changes in the levels of D‐dimer, C‐reactive protein, soluble (s)P‐selectin, factor VIII, thrombin generation, and exploratory biomarkers focusing on endogenous thrombin potential, including tissue factor, activated factor IX and activated factor XI Time frame: 9 weeks
Funding	This project was supported by an intramural grant from the University of Vermont Cancer Center.
Declarations of interest	The authors state that they have no conflict of interest.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number generation by the SAS procedure
Allocation concealment (selection bias)	Low risk	"Randomization was implemented and concealed in our research pharmacy, which distributed the drug or matching placebo to participants."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Randomization was implemented and concealed in our research pharmacy, which distributed the drug or matching placebo to participants."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"Participants, treating teams, and all laboratory personnel involved in the protocol were blinded to group assignment."
Incomplete outcome data (attrition bias) All outcomes	High risk	14 (14/38) participants did not complete the protocol.
Selective reporting (reporting bias)	Unclear risk	No protocol was found.
Other bias	Low risk	No other bias could be found.

Amarenco 2006.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 4.9 yrs Setting: hospital
Participants	Ages eligible for study: > 18 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 4731 Location: 205 sites in USA, Canada, Austria, Denmark, Sweden, Germany, Italy, Switzerland, Portugal, Spain, Australia, New Zealand, South Africa Inclusion criteria: "stroke or TIA within 1‐6 months before study entry, had LDL cholesterol levels of 100 to 190 mg/dl, and had no known CHD" Exclusion criteria: "women who were pregnant or breastfeeding; a history of CHD; significant PVD; atrial fibrillation; prosthetic heart valves; clinically significant mitral stenosis; sinus node dysfunction; uncontrolled hypertension, stroke caused by a revascularization procedure or trauma; subarachnoid haemorrhage; known hypersensitivity to statins or bile‐sequestering resins; active liver disease or hepatic dysfunction, defined as AST, or ALT ≥ 2 x ULN; haematologic conditions that may cause thrombus formation; endarterectomy within 1 month before randomization; severe renal dysfunction or nephrotic syndrome; creatine phosphokinase ≥ 5 ULN; participation in another clinical study within 30 days prior to screening for the present study; diseases or abnormalities that the investigator believed might compromise the patient's safety during the study; unreliability as a study participant, based on the investigator's prior knowledge of the patient, such as alcoholism, drug abuse, or psychiatric illness; use of any drugs known to affect lipid levels or immunosuppressive agents, azole antifungals or drugs associated with rhabdomyolysis in combination with statins"
Interventions	Intervention: 80 mg of atorvastatin per day (n = 2365) Control: placebo (n = 2366)
Outcomes	Nonfatal or fatal stroke; stroke or TIA, major coronary event, major cardiovascular event, acute coronary event, any coronary event, revascularization procedure, and any cardiovascular event. Individual components of the composite endpoints and death from any cause were also prespecified secondary outcomes Time frame: 5 years
Funding	Pfizer
Declarations of interest	"Dr Amarenco reports having received consulting fees from AstraZeneca, Novartis, Pfizer, and Sanofi‐Aventis; lecture fees from Otsuka Pharmaceutical and Pfizer; and grant support from Pfizer. Dr Bogousslavsky reports having received consulting fees from Pfizer and grant support from Pfizer. Dr Callahan reports having received consulting fees from Sanofi, lecture fees from Bristol‐Myers Squibb and Sanofi, and grant support from Pfizer. Dr Goldstein reports having received consulting fees from Pfizer, Bayer, AstraZeneca, Bristol‐Myers Squibb/Sanofi, Glaxo SmithKline, Merck Research Laboratories, Johnson & Johnson Cordis, and Organon; lecture fees from Bayer; and grant support from AGA Medical, Boehringer Ingelheim, the National Institutes of Health, Pfizer, and the Department of Veterans Affairs. Dr Hennerici reports having received grant support from Pfizer and Servier. Dr Rudolph is an employee of Pfizer and reports owning stock in the company. Dr Sillesen reports having received consulting fees from Sanofi‐Aventis; lecture fees from AstraZeneca, Bristol‐Myers Squibb, Merck, and Sanofi‐Aventis; and grant support from Pfizer. Ms Simunovic and Mr Szarek are employees of Pfizer and report owning stock in the company. Dr Welch reports having received consulting fees from Eisai, GlaxoSmithKline, Medpointe, AstraZeneca, NMT Medical, and Ortho‐McNeil; lecture fees from GlaxoSmithKline; and grant support from Pfizer. Dr Zivin reports having received consulting fees from Angel Pharmaceuticals, MEDACorp, MEDIACorp, Pfizer and Sirex; and grant support from PhotoThera and Pfizer."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Multi‐centre, international, double‐blind, randomized, placebo‐controlled trial, but insufficient details provided
Allocation concealment (selection bias)	Unclear risk	No details could be found
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Described as double‐blind, but no details provided
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear ‐ insufficient details provided
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates were high in each group. More participants in the placebo group than in the atorvastatin group withdrew consent after randomization (P = 0.07); permanently discontinued study treatment (atorvastatin: 20.2% vs placebo: 15.4%).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Asselbergs 2004.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: factorial design Masking: double blind Primary purpose: prevention Duration of study: follow‐up 46 months Setting: hospital
Participants	Ages eligible for study: 28 to 75 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 864 Location: Groningen, the Netherlands Inclusion criteria: "persistent microalbuminuria, BP 160/100 mm Hg and no use of antihypertensive medication, and a total cholesterol level 8.0 mmol/L, or 5.0 mmol/L in case of previous MI, 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."
Interventions	Intervention: pravastatin 40 mg/day (n = 433) Control: placebo (n = 431)
Outcomes	The primary endpoint was the combined incidence of cardiovascular mortality and hospitalization for cardiovascular morbidity. Time frame: 4 years
Funding	"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"
Declarations of interest	Not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Investigator‐initiated, single‐center, double‐blind, randomized, placebo‐controlled trial with a 2 x 2 factorial design. 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)	Unclear risk	Unclear ‐ insufficient details provided
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Described as double‐blind, but no details provided
Blinding of outcome assessment (detection bias) All outcomes	Low risk	An independent data and safety monitoring committee regularly monitored the progress of the study.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: 66.4% of participants in the placebo group and 74.1% in the pravastatin group showed compliance above 75% after 4 years of follow‐up.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Chan 2010.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 3.5 (IQR 2.1 to 4.5) yrs Setting: hospital
Participants	Ages eligible for study: 18 to 82 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 269 Location: 23 Canadian centres Inclusion criteria: "men and women between 18 and 82 years of age with asymptomatic mild to moderate aortic valve stenosis defined by maximum aortic valve velocity between 2.5 and 4.0 m/s" Exclusion criteria: "Patients with clinical indications for the use of statins as defined by Canadian guidelines such as coronary artery disease, cerebrovascular disease, peripheral vascular disease, and diabetes were excluded"
Interventions	Intervention: rosuvastatin 40 mg/day (n = 134) Control: matching placebo (n = 135)
Outcomes	The primary endpoint was the onset of stroke and TIA. Secondary endpoints were the onset of each stroke subtype, MI, vascular accident, death, hospitalization, dependence in activities of daily living, degree of disability, the onset of dementia, and severity of cognitive impairment Time frame: 3 to 5 years
Funding	Canadian Institute of Health Research and AstraZeneca Canada Inc
Declarations of interest	"Dr. Chan has received travel grants from Merck Frosst/Schering and AstraZeneca Canada Inc. Dr. Teo has received travel grants from AstraZeneca Canada Inc. Dr. Dumesnil has had consultancies for and/or research funds from St Jude Medical, Edward Life Sciences, and Metronic. Dr Tam has received travel grants from Merck Frosst/Schering and AstraZeneca Canada Inc. A. Ni reports no conflicts."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The randomization schedule was centralized and generated by the computer program, which had no access to the rest of the data.
Allocation concealment (selection bias)	Low risk	When a center was ready to randomize a participant, the site co‐ordinator obtained a randomization number from the study database via a secure Internet line.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants, site coordinators, investigators, and statisticians were all blinded to treatment assignment.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Participants, site coordinators, investigators, and statisticians were all blinded to treatment assignment.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high (intervention vs control: 57/134 vs 66/135).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Colhoun 2004.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 3.9 (IQR 3.0 to 4.7) yrs Setting: hospital
Participants	Ages eligible for study: 40 to 75 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 2838 Location: 132 centers in the UK and Ireland Inclusion criteria: "men and women aged 40‐75 years with type 2 DM diagnosed at least 6 months before study entry were considered for inclusion provided they had at least 1 or more of the following: a history of hypertension, defined as receiving antihypertensive treatment or having systolic BP of 140 mm Hg or greater or diastolic BP of 90 mm Hg or greater on at least 2 successive occasions; retinopathy; microalbuminuria or macroalbuminuria, defined as a positive Micral or other strip test, an albumin creatinine ratio of 2.5 mg/mmol or greater, or an albumin excretion rate on timed collection of 20 g/min or more, all on at least 2 successive occasions; or currently smoking" Exclusion criteria: "any history of MI, angina, coronary vascular surgery, cerebrovascular accident, or severe peripheral vascular disease"
Interventions	Intervention: atorvastatin 10 mg daily (n = 1428) Control: placebo (n = 1410)
Outcomes	The primary endpoint consisted of the first of the following: acute coronary heart disease event, coronary revascularization procedures, or stroke. Prespecified secondary efficacy outcomes were the effect of treatment on total mortality and the effect of atorvastatin on any acute, hospital‐verified cardiovascular endpoint. Time frame: 4 years
Funding	Funded by the UK Department of Health, Diabetes UK, and Pfizer
Declarations of interest	"DJB and HMC have served as consultants to, and received travel expenses and payments for speaking at meetings from, Pfizer. PND has received travel expenses, payment for speaking at meetings, and funding for research from Pfizer. JHF has served as a consultant to and received travel expenses, payment for speaking at meetings, or funding for research from pharmaceutical companies marketing lipid‐lowering drugs, including AstraZeneca and Pfizer. GAH has served as a consultant to and received travel expenses, payment for speaking at meetings, or funding for research from pharmaceutical companies marketing lipid‐lowering drugs, including AstraZeneca and Pfizer. HAWN has served as a consultant to and received travel expenses, payment for speaking at meetings, or funding for research from pharmaceutical companies marketing lipid‐lowering drugs, including AstraZeneca, Merck Sharp and Dohme, and Pfizer. The UCL coordinating center was partly funded by a grant from Pfizer UK and Pfizer Inc to UCL. SJL, MJT, MIM, and VC‐M have no conflicts of interest to declare."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized according to a computer‐generated randomization code
Allocation concealment (selection bias)	Unclear risk	No more details were found.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Investigators, pharmacists, study administrators, and participants were unaware of the randomization code throughout the study.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Investigators, pharmacists, study administrators, and participants were unaware of the randomization code throughout the study.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low (intervention vs control: 15/1412 vs 8/1429).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Cowell 2005.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 25 months Setting: hospital
Participants	Ages eligible for study: > 18 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 155 Location: UK Inclusion criteria: "patients > 18 yrs of age with calcific aortic valve stenosis, an aortic‐jet velocity of at least 2.5 m per second, and aortic‐valve calcification on echocardiography" Exclusion criteria: "child‐bearing potential without contraception, active or chronic liver disease, a history of alcohol or drug abuse, severe mitral‐valve stenosis, severe mitral or aortic regurgitation, left ventricular dysfunction, a planned aortic‐valve replacement, intolerance of statins, statin therapy or a potential benefit from statin therapy, a baseline serum total cholesterol concentration of less than 150 mg/dL, and presence of a permanent pacemaker or cardio‐defibrillator"
Interventions	Intervention: 80 mg of atorvastatin daily (n = 77) Control: placebo (n = 78)
Outcomes	Primary endpoints: progression of stenosis, determined according to changes in aortic jet velocity on Doppler echocardiography, and progression of valvular calcification, as measured by CT Secondary endpoints were a composite of clinical endpoints (death from cardiovascular causes, aortic‐valve replacement, or hospitalization attributable to severe aortic stenosis), aortic‐valve replacement, death from any cause, hospitalization for any cause, and hospitalization for cardiovascular causes Time frame: 7 to 36 months
Funding	Supported by a grant from the British Heart Foundation (PG/2000/044), by an educational award from Pfizer, and by the Wellcome Trust Clinical Research Facility, Edinburgh
Declarations of interest	"Drs Newby, Bloomfield, and Boon report having received unrestricted educational grant support from Pfizer, and Drs Newby, Northridge, and Boon report having received consulting fees from and having served on advisory boards for Pfizer"
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Double‐blind, placebo‐controlled trial. Randomization was determined by the minimization technique with the use of a dedicated, locked computer program.
Allocation concealment (selection bias)	Low risk	Numbered containers were used.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Study co‐ordinator was blinded to group assignment.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear ‐ insufficient details provided
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low. No incomplete outcome data were observed.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Crouse 2007.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: a median follow‐up of 24 months Setting: primary care centers
Participants	Ages eligible for study: 45 to 70 yrs (men) or 55 to 70 yrs (women) Sexes eligible for study: both Accepts healthy volunteers: yes Total number of participants: 984 Location: US and Europe Inclusion criteria: "aged 45‐70 yrs (men) or 55‐70 yrs (women); screening LDL‐C level of 120 to less than 190 mg/dL for those with only age as a CHD risk factor or 120 to less than 160 mg/dL for individuals with 2 or more CHD risk factors and 10‐year risk of CHD events of less than 10%; HDL‐C level of 60 mg/dL or lower; level of triglycerides lower than 500 mg/dL; and maximum CIMT measurements between 1.2 mm and less than 3.5 mm from 2 separate ultrasound examinations" Exclusion criteria: "use of lipid‐lowering therapies in the previous 12 months, clinical evidence of coronary artery disease or other peripheral atherosclerotic diseases, prior revascularization procedures, 10 yrs CHD 10% or more, DM, uncontrolled hypertension or familial hypercholesterolemia, or serum creatinine concentration higher than 2 mg/dL (177 µmol/L)"
Interventions	Intervention: 40 mg dose of rosuvastatin (n = 702) Control: placebo (n = 282)
Outcomes	Annualized rate of change in maximum CIMT based on all scans performed during the 2‐year study period from each of the 12 carotid artery sites (near and far walls of the right and left common carotid artery, carotid bulb, and internal carotid artery) Annualized rate of change in maximum CIMT derived from the near and far walls of the right and left common carotid artery; the right and left carotid bulb; the right and left internal carotid artery; and annualized rate of change in mean CIMT for the near and far walls of the right and left common carotid artery Time frame: 2 years
Funding	AstraZeneca
Declarations of interest	"Dr Crouse reported receiving grant or salary support from Merck, Merck‐Schering Plough, Pfizer, AstraZeneca, and Kos Pharmaceuticals; and giving lectures for Merck, Merck‐Schering Plough, Pfizer, AstraZeneca, Abbott, and Kos Pharmaceuticals. Dr Raichlen reported being an employee of AstraZeneca. Dr Riley reported receiving research contracts from AstraZeneca, Organon, and Pfizer. Mr Evans reported receiving grant support and honoraria from AstraZeneca, Organon, and Pfizer; and being a consultant to AstraZeneca and Pfizer. Dr Palmer reported being an employee of AstraZeneca. Dr O’Leary reported being on data and safety monitoring boards for Pfizer and AstraZeneca; being a consultant to Pfizer, Sankyo Pharma, Sanofi Aventis, GlaxoSmithKline, Eli Lilly, Schering‐Plough, Esperion Therapeutics, and Merck; and being an equity partner in Imagepace LLC. Dr Grobbee reported receiving grant support from and delivering lectures for Pfizer, AstraZeneca, Organon, Servier, and Merck. Dr Bots reported receiving study grants for studies on carotid intima‐media thickness and/or honoraria for professional input on carotid intima‐media thickness issues from AstraZeneca, Icelandic Heart Foundation, Organon, Pfizer, the Netherlands Heart Foundation, the Netherlands Organisation for Health Research and Development, Servier, and Unilever."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Eligible individuals were randomized to either the placebo or rosuvastatin group in blocks of 7 (5 to the rosuvastatin group and 2 to the placebo group) at each clinical site.
Allocation concealment (selection bias)	Unclear risk	This study did not provide details about allocation concealment methods.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinded study medication was supplied in individually numbered bottles prepared prior to the clinic visits and eligible individuals were allocated study medication sequentially. Investigators checked adherence but were unaware of treatment allocations for the duration of the study.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	This study did not provide details about blinding methods.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: 172 participants in the rosuvastatin group (25%) discontinued participation in the study and 74 in the placebo group discontinued participation (26%).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Dinglas 2016.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: triple (participant, care provider, investigator) Primary purpose: prevention Duration of study: a median follow‐up of 1 year Setting: hospital
Participants	Ages eligible for study: 18 yrs and older (adult, older adult) Sexes eligible for study: both Accepts healthy volunteers: yes Total number of participants: 745 Location: United States Inclusion criteria: "patients with sepsis‐associated acute respiratory distress syndrome (ARDS) if they were receiving positive‐pressure mechanical ventilation through an endotracheal tube, had a ratio of the partial pressure of arterial oxygen (Pao2) to the fraction of inspired oxygen (Fio2) of 300 or less, and had bilateral infiltrates on chest radiography that were consistent with pulmonary edema, without evidence of left atrial hypertension. Additional inclusion criteria were a known or suspected infection and either of the following criteria for a systemic inflammatory response: a white‐cell count of more than 12,000 per cubic millimeter or less than 4000 per cubic millimeter or a differential count with more than 10% band forms, or a core body temperature of more than 38°C or less than 36°C." Exclusion criteria: "Major exclusion criteria were the presence of ARDS for more than 48 hours; chronic conditions that could adversely affect survival, impair weaning from the ventilator, or compromise adherence to the protocol; serum levels of creatine kinase, aspartate aminotransferase, or alanine aminotransferase of more than five times the upper limit of the normal range; ingestion of a statin (on an inpatient or outpatient basis) in the 48 hours before randomization; and an inability to obtain consent."
Interventions	Intervention: participants received an initial rosuvastatin 40 mg loading dose followed by 20 mg of study drug daily by mouth or feeding tube for 28 days or until discharged from the study hospital (n = 702). Control: participants received a placebo by mouth or feeding tube daily for 28 days or until discharged from the study hospital (n = 282).
Outcomes	Hospital mortality to day 60; ventilator‐free days at study day 28; organ‐failure‐free days at day 14; ICU‐free days to day 28; percentage of participants with arrhythmia, bowel ischemia, myocardial infarction, ischemic stroke, and thromboembolism; changes in plasma concentrations of CRP from baseline to day 6 and day 14; the age‐ and sex‐adjusted Physical Function and Mental Health domains scores of the Medical Outcomes Study SF‐36 version 2 instrument (SF‐36; range: 0 to 100; higher score is better) at 6‐month follow‐up, with 12‐month scores as a secondary outcome.
Funding	Funded by the National Heart, Lung, and Blood Institute and the Investigator‐Sponsored Study Program of AstraZeneca
Declarations of interest	No
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No details were found.
Allocation concealment (selection bias)	Unclear risk	No details were found.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participant, care provider, and investigator were blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Participant, care provider, and investigator were blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No participants were lost to follow‐up.
Selective reporting (reporting bias)	Low risk	All outcomes were reported.
Other bias	Low risk	No other bias.

Downs 1998.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 5.2 years Setting: hospital
Participants	Ages eligible for study: 45 to 73 yrs (58 ± 7 yrs) Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 6605 Location: "two sites in Texas, Lackland Air Force Base in San Antonio and the University of North Texas Health Science Center in Fort Worth" Inclusion criteria: "met the lipid entrance criteria and had no prior history, signs, or symptoms of definite MI, angina, claudication, CVA, or TIA; Lipid entry criteria (TC, 4.65‐6.82mmol/L; LDL‐C, 3.36‐4.91 mmol/L; HDL‐C, 1.16 mmol/L for men or 1.22 mmol/L for women; and triglycerides, 4.52 mmol/L were to be met at both 4 and 2 weeks prior to randomization, with less than 15% difference in LDL‐C values" Exclusion criteria: "volunteers with uncontrolled hypertension, secondary hyperlipidemia, or type 1 or type 2 diabetes mellitus. Additionally, volunteers were excluded if they had a body weight of more than 50% greater than the desirable limit for their height."
Interventions	Intervention: lovastatin (20 to 40 mg daily) (n = 3304) Control: matching placebo (n = 3301)
Outcomes	First acute major coronary events (i.e. sudden cardiac death, fatal and nonfatal MI, and unstable angina) Fatal and nonfatal coronary revascularization procedures, unstable angina, fatal and nonfatal MI, fatal and nonfatal cardiovascular events, fatal and nonfatal coronary events, cardiovascular mortality, and CHD mortality Time frame: 5 years
Funding	Merck & Co Inc
Declarations of interest	Not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No details could be found, even in the study protocol.
Allocation concealment (selection bias)	Unclear risk	No details could be found, even in the study protocol.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants, investigators, steering committee members, and those providing participant care, monitoring or managing data, or adjudicating endpoints were blinded. Blinding was maintained by titrating equal numbers of randomly selected placebo‐group participants to 2 tablets daily.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Participants, investigators, steering committee members, and those providing participant care, monitoring or managing data, or adjudicating endpoints were blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: "Study drug regimens were maintained until trial termination by 2,335 (71%) of the 3,304 participants randomized to lovastatin and by 2,081 (63%) of the 3,301 randomized to placebo".
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Fassett 2010.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 2.5 yrs Setting: hospital
Participants	Ages eligible for study: 18 to 85 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 132 Location: Northern Tasmania Inclusion criteria: "aged between 18 and 85 yrs with a serum creatinine > 120mol/L, chronic kidney disease" Exclusion criteria: "receiving lipid‐lowering therapy, female of childbearing age, or participating in another intervention study"
Interventions	Intervention: 10 mg of atorvastatin/day (n = 64) Control: placebo (n = 68)
Outcomes	Change in the rate of kidney function decline Time frame: 3 years
Funding	This project was supported by a grant from the Clifford Craig Medical Research Trust, an independent research funding body. Pfizer Pharmaceuticals provided atorvastatin and placebo but did not have any role in the study.
Declarations of interest	None
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized using computer‐generated random numbers
Allocation concealment (selection bias)	Low risk	This sequence was concealed from study investigators. Computer‐generated random numbers were placed in blocks of 10 by the clinical trial pharmacist and related to a series of drug code numbers. Each drug code assignment block referred to one of the stratification groups.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Trial clinicians were blinded to the randomization.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No details were found.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: at randomization, 64 participants were assigned to receive atorvastatin 10 mg and 68 participants placebo. Nine of these, (6 atorvastatin, 3 placebo), withdrew before starting study medication or completing the baseline visit. Thirty‐six participants withdrew after starting study medication, 14 from the atorvastatin and 22 from the placebo group.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Feldman 2010.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 1.5 yrs Setting: hospital
Participants	Ages eligible for study: 50 to 90 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 640 Location: "10 countries and 87 enrolling sites: United States (39), Germany (9), Canada (9), United Kingdom (8), Australia (7), Spain (7), South Africa (5), Sweden (5), Austria (4), and Denmark (4)" Inclusion criteria: "mild to moderate probable Alzheimers disease, aged 50‐90 yrs" Exclusion criteria: "participants experiencing any clinically significant or unstable medical condition"
Interventions	Intervention: atorvastatin 80 mg/day (n = 314) Control: matching placebo (n = 326)
Outcomes	ADAS‐cog, Alzheimer's Disease, and ADCS‐CGIC Time frame: 1.5 years
Funding	Pfizer Inc
Declarations of interest	"Dr Schwam is an employee of Pfizer; and owns equity/ownership in Pfizer. Dr Schindler is an employee of Pfizer; and owns equity/ownership in Pfizer. Dr Hey‐Hadavi is an employee of Pfizer; and owns equity/ownership in Pfizer. Dr DeMicco is an employee of Pfizer; and owns equity/ownership in Pfizer. Dr Breazna is an employee of Pfizer and he and his wife have equity/ownership interests in Pfizer."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization code used
Allocation concealment (selection bias)	Unclear risk	Unclear ‐ insufficient details provided.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blind; blinding of both the investigator and the participants described. "Briefly, at the start of the double‐blind, 72‐week treatment period, participants receiving existing treatment with donepezil 10 mg were randomized to atorvastatin 80 mg or matching placebo."
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear ‐ insufficient details provided.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates were high in each group. The overall study completion rate was very close to predicted at 70.6% with a dropout rate of 29.4%. A larger proportion of participants discontinued the study in the atorvastatin group compared with the placebo group (intervention vs control: 107/314 vs 80/326).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Fellström 2004.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 5.1 yrs Setting: hospital
Participants	Ages eligible for study: 30 to 75 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 2102 Location: "Northern and Central Europe (Belgium, Denmark, Finland, Germany, Norway, Sweden, Switzerland, United Kingdom) and Canada" Inclusion criteria: "renal transplant recipients more than 6 months prior to randomization with stable graft function, and with a total serum cholesterol concentration between 4.0 and 9.0 mmol/L; patients with a history of MI more than 6 months prior to randomization could be enrolled if their total cholesterol levels ranged from 4.0 to 7.0 mmol/L" Exclusion criteria: "statin therapy, had familial hypercholesterolemia, or had experienced an acute rejection episode in the 3 months prior to randomization. In addition, patients with a predicted life expectancy of less than 1 year were excluded."
Interventions	Intervention: fluvastatin, 40 mg/day (n = 1050) Control: matching placebo (n = 1052)
Outcomes	Graft loss, graft loss or doubling of serum creatinine, and graft loss or doubling of serum creatinine or participant death Time frame: 5 to 6 years
Funding	The ALERT trial was supported by a research grant from Novartis Pharma AG, Basel, Switzerland. Novartis Pharma AG also provided fluvastatin and matching placebo for the trial.
Declarations of interest	"The ALERT committee members received financial support from Novartis Pharma AG, Basel, Switzerland, in the form of honoraria (excluding members who were investigators) and support for travel and accommodation expenses incurred by attending committee meetings. DO Solbu is a Novartis employee and was a non‐voting member of the steering committee. All other members of the steering committee have served as consultants for and received travel expenses, payment for lecturing, or funding for research from other pharmaceutical companies marketing lipid‐lowering drugs, including Merck Sharp and Dohme, Bristol‐Myers Squibb, Astra‐Zeneca, Schering, Bayer, and Pfizer."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized, double‐blind, placebo‐controlled study, fixed‐block randomization
Allocation concealment (selection bias)	Unclear risk	No details were found.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blind, matching placebo. Investigators were unaware of participants’ lipid concentrations.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	An independent critical events committee of two nephrologists and two cardiologists who were unaware of the treatment assignment reviewed all primary and secondary endpoints for adjudication.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high (intervention vs control: 15% vs 14%).
Selective reporting (reporting bias)	Low risk	We compared the protocol and the study publications and no outcomes were missed.
Other bias	Low risk	No other bias could be found.

Fellström 2009.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 3.2 yrs Setting: hospital
Participants	Ages eligible for study: 50 to 80 yrs Accepts healthy volunteers: no Total number of participants: 2776 Location: "Australia, Austria, Belgium, Brazil, Bulgaria, Canada, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Korea, Republic of, Mexico, Netherlands, Norway, Poland, Sweden, Switzerland, Turkey, United Kingdom" Inclusion criteria: "Men and women 50 to 80 years of age who had end‐stage renal disease and had been treated with regular hemodialysis or hemofiltration for at least 3 months" Exclusion criteria: "The major exclusion criteria were statin therapy within the previous 6 months, expected kidney transplantation within 1 year, and serious hematologic, neoplastic, gastrointestinal, infectious, or metabolic disease (excluding diabetes) that was predicted to limit life expectancy to less than 1 year. Other exclusion criteria were a history of a malignant condition, active liver disease (indicated by an alanine aminotransferase level that was more than three times the upper limit of the normal range), uncontrolled hypothyroidism, and an unexplained elevation in the creatine kinase level to more than three times the upper limit of the normal range"
Interventions	Intervention: rosuvastatin, 10 mg daily (n = 1389) Control: matching placebo (n = 1384)
Outcomes	Number of randomized participants who: had major cardiovascular event (non‐fatal stroke, non‐fatal myocardial infarction or cardiovascular death); died from any cause; had major cardiovascular event or died from any known cause; died from cardiovascular cause; died from non‐cardiovascular cause; had atherosclerotic cardiac event (non‐fatal myocardial infarction or coronary heart disease death); experienced a procedure as a result of stenosis or thrombosis of the vascular access (arteriovenous (AV) fistulas and grafts only) for haemodialysis; experienced coronary or peripheral revascularisation Time frame: 3.9 years
Funding	AstraZeneca
Declarations of interest	"Dr. Fellström reports receiving consulting fees from AstraZeneca, Novartis, Roche, and Wyeth, lecture fees from Astellas, Novartis, and Roche, and grant support from Novartis, Roche, Merck–Schering‐Plough, and Wyeth and serving as national coordinator for the Study of Heart and Renal Protection (SHARP) study at Oxford University’s Clinical Trial Service Unit; Dr. Jardine, receiving consulting fees from Novartis, AstraZeneca, and Wyeth and lecture fees from Novartis and Astellas; Dr. Schmieder, receiving consulting and lecture fees from AstraZeneca, Novartis, Merck Sharp & Dohme, and Pfizer; Dr. Holdaas, receiving consulting fees from Novartis, AstraZeneca, and Schering‐Plough and lecture fees from Novartis and AstraZeneca, and serving as national coordinator for the SHARP study; Dr. Bannister, receiving consulting fees from Baxter, Amgen, and Genzyme, lecture fees from Servier and Boehringer Ingelheim, and grant support from Amgen; Dr. Beutler, receiving lecture fees from Pfizer; Dr. Chevaile, receiving consulting fees from AstraZeneca; Dr. Cobbe, receiving consulting fees and grant support from AstraZeneca; Dr. Grönhagen‐Riska, receiving grant support from AstraZeneca and serving as national coordinator for the SHARP study; Dr. De Lima, receiving consulting fees from AstraZeneca; Dr. Lins, receiving consulting fees from AstraZeneca and Novartis and lecture fees from Servier and SanofiAventis; Dr. Mayer, receiving consulting fees from AstraZeneca and Genzyme, lecture fees from Genzyme and Amgen, and grant support from Amgen; Dr. McMahon, receiving consulting fees from AstraZeneca and Schering‐Plough and lecture fees from AstraZeneca; Dr. Parving, receiving consulting and lecture fees from AstraZeneca, Merck, and Novartis and having equity ownership and stock options in Novo Nordisk; Dr. Remuzzi, serving on the Prospective Evaluation of Proteinuria and Renal Function in Diabetic Patients with Progressive Renal Disease (PLANET) and AURORA advisory committees for AstraZeneca; Dr. Samuelsson, receiving lecture fees from AstraZeneca; Dr. Sonkodi, receiving lecture fees from AstraZeneca, Richter Gedeon Nyrt, and Egis Nyrt; Dr. Tesar, receiving lecture fees from Amgen and Novartis; Dr. Wiecek, receiving consulting fees from AstraZeneca; Dr. Wüthrich, receiving consulting fees from Amgen, Genzyme, Novartis, and Vifor, lecture fees from Amgen, Vifor, and Wyeth, and grant support from Amgen, Genzyme, Novartis, Vifor, and Wyeth; Mr. Gottlow, being an employee of and owning stock or having other ownership interest in AstraZeneca; Dr. Johnsson, being an employee of and owning stock or having other ownership interest in AstraZeneca; and Dr. Zannad, receiving consulting fees from Servier, Novartis, ResMed, Daiichi Sankyo, Pfizer, AstraZeneca, and Merck and lecture fees from Pfizer, Daiichi Sankyo, Servier, and Medtronic. All authors except for employees of AstraZeneca received honoraria for serving on the steering committee of the AURORA study."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No details could be found in the protocol and final publications.
Allocation concealment (selection bias)	Unclear risk	No details could be found in the protocol and final publications.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double blind (participant and care provider)
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	The executive steering committee designed the trial and supervised its conduct in collaboration with the sponsor, AstraZeneca. The sponsor collected the trial data and analyzed them according to a predefined statistical analysis plan. The analyses were verified by Dr AH Zwinderman, an independent statistician at the Department of Clinical Epidemiology, Biostatistics, and Bioinformatics, University of Amsterdam. So we were unclear about the blinding of outcome assessments.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No participants were lost to follow‐up.
Selective reporting (reporting bias)	Low risk	All outcomes were reported in Clinicaltrials.gov.
Other bias	High risk	The executive steering committee designed the trial and supervised its conduct in collaboration with the sponsor, AstraZeneca. The sponsor collected the trial data and analyzed them according to a predefined statistical analysis plan. The analyses were verified by Dr AH Zwinderman, an independent statistician at the Department of Clinical Epidemiology, Biostatistics, and Bioinformatics, University of Amsterdam. This might introduce bias during the study conduct.

Freeman 2011.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: double‐blind Primary purpose: prevention Duration of study: follow‐up 3.8 yrs Setting: hospital
Participants	Ages eligible for study: 70 to 82 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 5804 Location: 8 UK centers Inclusion criteria: "evidence of pre‐existing vascular disease or at least one major risk factor for vascular disease" Exclusion criteria: "cerebral infarction of determined rare etiology, infarction associated with catheterization or surgery, and preferred use of statins for the treatment of co‐morbid coronary artery disease"
Interventions	Intervention: pravastatin (40 mg per day) (n = 2834) Control: placebo (n = 2865)
Outcomes	Primary outcome was the combined endpoint of definite or suspect death from CHD, non‐fatal MI, and fatal or non‐fatal stroke, assessed in the entire cohort. Secondary outcomes included the examination of the coronary and cerebrovascular components separately. Primary outcome was assessed separately for men and women, and for those with and without pre‐existing disease. Time frame: 3 years
Funding	"Chest, Heart, Stroke Scotland Project Grant R05A89, Chief Scientist Office NHS Clinical Research Grant 2006, and a GlaxoSmithKline Clinical Fellowship and Tenovus Scotland research grant"
Declarations of interest	None
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized, double‐blind, placebo‐controlled trial. "The randomization sequence was generated with a computerized random number generator and consisted of balanced blocks of size four." "Randomization was performed using the method of permuted blocks, stratified by site."
Allocation concealment (selection bias)	Low risk	Randomization was done by telephone call or through fax exchange with the study data center.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Described as double‐blind, but no details provided
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear. No details were found.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low. No participants lost to follow‐up were reported.
Selective reporting (reporting bias)	Low risk	We compared the protocol and the study publications, and no outcomes were missed.
Other bias	Low risk	No other bias could be found.

Glynn 2009.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: a median follow‐up of 1.9 (maximum 5.0) yrs Setting: community
Participants	Ages eligible for study: 50 yrs and older Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 17,802 Location: 1315 sites in 27 countries Inclusion criteria: "men 50 yrs or older, women 60 yrs or older; low to normal levels of LDL‐C (< 130 mg/dL); elevated levels of CRP > 2.0 mg/L" Exclusion criteria: "history of cardiovascular or cerebrovascular events; active liver disease; DM; uncontrolled hypertension or hypothyroidism; history of certain malignancies; chronic inflammatory conditions; history of alcohol or drug abuse"
Interventions	Intervention: rosuvastatin 20 mg daily (n = 8901) Control: placebo 20 mg daily (n = 8901)
Outcomes	Primary outcome measures: rate of major cardiovascular events Secondary outcome measures: mortality, non‐cardiovascular mortality, adverse events, incidence of DM, venous thromboembolic events, and bone fractures Time frame: 5 years
Funding	AstraZeneca
Declarations of interest	"During the period of this project, Dr Glynn reports receiving research grant support from Astra‐Zeneca and Bristol Myers Squibb. Dr Fonseca reports having received research grants, lecture fees, and consulting fees from Astra Zeneca, Pfizer, Schering‐Plough, Sanofi‐Aventis, and Mer.ck. Dr Genest reports having received lecture fees from Astra‐Zeneca, Schering Plough, Merck Schering Plough, Pfizer, Novartis, and Sanofi‐Aventis; and consulting fees from Astra‐Zeneca, Merck, Merck Frosst, Schering‐Plough, Pfizer, Novartis, Resverlogix, and Sanofi‐Aventis. Dr Gotto reports having received consulting fees from Dupont, Novartis, Aegerion, Arisaph, KOWA, Merck, Merck Schering Plough, Pfizer, Genentech, Martek, and Reliant, having served as an expert witness, and as having received publication royalties. Dr Kastelein reports receiving research grant support from Astra‐Zeneca, Pfizer, Roche, Novartis, Merck, Merck Schering Plough, ISIS, Genzyme, and Sanofi‐Aventis; lecture fees from Astra‐Zeneca, Glaxo Smith Kline, Pfizer, Novartis, Merck‐Schering Plough, Roche, ISIS, and Boehringer‐Ingelhiem; and consulting fees from Astra‐Zeneca, Abbott, Pfizer, ISIS, Genzyme, Roche, Novartis, Merck, Merck Schering Plough, and SanofiAventis. Dr Koenig reports receiving research grant support from Anthera, Dade‐Behring and Glaxo‐Smith‐Kline; lecture fees from Astra‐Zeneca, Pfizer, Novartis, Glaxo‐Smith‐Kline, DiaDexus, Roche, and Boehringer‐Ingelheim; and consulting fees from Glaxo‐Smith Kline, Medlogix, Anthera, and Roche. Dr Libby reports receiving lecture fees from Pfizer and lecture or consulting fees from Astra‐Zeneca, Bristol Myers Squibb, Glaxo Smith Kline, Merck, Pfizer, Sanofi‐Aventis, VIA Pharmaceutical, Interleukin Genetics, Kowa Research Institute, Novartis, and Merck Schering Plough. Dr Lorenzatti reports receiving research grant support, lecture fees, and consulting fees from Astra‐Zeneca, Takeda, and Novartis. Dr Nordestgaard reports receiving lecture fees from Astra‐Zeneca, Abbott, Sanofi‐Aventis, Pfizer, Boehringer Ingelheim and Merck; and consulting fees from Astra‐Zeneca and BG Medicine. Dr Shepherd reports receiving lecture fees from Astra‐Zeneca, Pfizer, and Merck; and consulting fees from Astra‐Zeneca, Merck, Roche, Glaxo SmithKline, Pfizer, Nicox, and Oxford Biosciences. Dr. Ridker reports having received investigator‐initiated research grant support from Astra‐Zeneca, Novartis, Roche, and Sanofi‐Aventis, as well as non‐financial research support from Amgen; consulting fees and/or lecture fees from Astra‐Zeneca, Novartis, Merck, Merck Schering Plough, Sanofi‐Aventis, ISIS, Seimens, and Vascular Biogenics; and is listed as a co‐inventor on patents held by the Brigham and Women’s Hospital that relate to the use of inflammatory biomarkers in cardiovascular disease. These patents have been licensed to several entities, including Astra‐Zeneca and Seimens. Dr Willerson, Ms Danielson, and Ms MacFadyen report no conflicts."
Notes	Financially supported by AstraZeneca. The authors of all publications reported that the sponsor collected the trial data and monitored the study sites but played no role in the conduct of the analyses or drafting of the manuscript. This study was meant to continue for five years, but it was stopped early on the advice of the independent data and safety monitoring board, after a median follow‐up of 1.9 years, based on the size and precision of the observed treatment benefit, as well as effects on the rates of death in participants treated with rosuvastatin compared with placebo.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The method of randomization was a computer‐generated list.
Allocation concealment (selection bias)	Low risk	"Randomization was performed with the use of an interactive voice‐response system and was stratified according to center."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"A closeout visit occurred after study termination, at which time participants were unblinded." Double‐blind, but no details were found. Masking: quadruple (participant, care provider, investigator, outcomes assessor)
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"All reported primary endpoints were adjudicated by an independent endpoint committee blinded to randomized treatment assignment. Adverse events were monitored and reported in a blinded manner until the date of the closeout visit and discontinuation of therapy." Masking: quadruple (participant, care provider, investigator, outcomes assessor)
Incomplete outcome data (attrition bias) All outcomes	Low risk	We compared study protocol and study publications; no missing data for outcomes were reported. No participants were withdrawn and all participants were accounted for. There are no incomplete outcome data in this study.
Selective reporting (reporting bias)	Low risk	We compared the protocol and the study publications, and no outcomes were missed.
Other bias	Low risk	No other bias could be found.

HPS study 2002.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: factorial design Masking: blinded (participants) Primary purpose: prevention Duration of study: follow‐up 5.3 yrs Setting: hospital
Participants	Ages eligible for study: 40 to 80 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 20,536 Location: UK Inclusion criteria: "men and women aged about 40‐80 years with non‐fasting blood total cholesterol concentrations of at least 3.5 mmol/L were eligible provided they had a medical history of coronary disease, PAD, cerebrovascular disease, diabetes, or treated hypertension. PAD was defined as a history of intermittent claudication or previous peripheral arterial revascularization procedure, amputation, or aneurysm repair." Exclusion criteria: "people were ineligible if their own doctor considered statin therapy to be clearly indicated or contraindicated, or if they had an MI, stroke, or hospital admission for angina within the previous 6 months; chronic liver disease or evidence of abnormal liver function; severe renal disease or evidence of substantially impaired renal function; inflammatory muscle disease or evidence of muscle problems; concurrent treatment with ciclosporin, fibrates, or high‐dose niacin; child‐bearing potential; severe heart failure; or other conditions that might limit long‐term compliance."
Interventions	Intervention: 40 mg simvastatin daily (n = 10269) Control: matching placebo (n = 10267)
Outcomes	MI, stroke, vascular procedure, cancer or other serious adverse experience, and any of the main reasons for all other hospital admissions (including day cases) Time frame: 5 years
Funding	The UK Medical Research Council, the British Heart Foundation, Merck & Co (manufacturers of simvastatin), and Roche Vitamins Ltd (manufacturers of the vitamins)
Declarations of interest	None. "The Clinical Trial Service Unit has a staff policy of not accepting honoraria or other payments from the pharmaceutical industry, except for the reimbursement of costs to participate in scientific meetings. Members of the writing committee have, therefore, only had such costs reimbursed."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The central telephone randomization system used a minimization algorithm to balance the treatment groups with respect to eligibility criteria and other major prognostic factors.
Allocation concealment (selection bias)	Low risk	The central telephone randomization system used a minimization algorithm to balance the treatment groups with respect to eligibility criteria and other major prognostic factors.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	The placebo was matched in order to make the participants blind to the treatment group.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No details were found.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: 40 mg simvastatin daily (average compliance 85%); matching placebo (average compliance 68%).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Kjekshus 2007.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: single blind Primary purpose: prevention Duration of study: follow‐up 32.8 months Setting: hospital
Participants	Ages eligible for study: > 60 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 5011 Location: 371 sites in 19 European countries, Russia, and South Africa Inclusion criteria: "NYHA class II, III, or IV ischaemic, systolic heart failure" Exclusion criteria: "previous statin‐induced myopathy or hypersensitivity reaction; decompensated heart failure or a need for inotropic therapy; MI within the past 6 months; unstable angina or stroke within the past 3 months; PCI, CABG, or the implantation of a cardioverter‐defibrillator or biventricular pacemaker within the past 3 months or a planned implantation of such a device; previous or planned heart transplantation; clinically significant, uncorrected primary valvular heart disease or a malfunctioning prosthetic valve; hypertrophic cardiomyopathy; acute endomyocarditis or myocarditis, pericardial disease, or systemic disease; acute or chronic liver disease; levels of alanine aminotransferase or thyrotropin of more than 2 times the ULN range; a serum creatinine level of more than 2.5 mg per decilitre; chronic muscle disease or an unexplained creatine kinase level of more than 2.5 times the ULN; previous treatment with cyclosporine; any other condition that would substantially reduce life expectancy or limit compliance with the protocol; or the receipt of less than 80% of dispensed placebo tablets during the run‐in period"
Interventions	Intervention: 10 mg rosuvastatin per day (n = 2514) Control: matching placebo (n = 2497)
Outcomes	The primary outcome was the composite of death from cardiovascular causes, nonfatal MI, and nonfatal stroke. The secondary outcomes were death from any cause, any coronary event, death from cardiovascular causes, and the number of hospitalizations for cardiovascular causes, unstable angina, or worsening heart failure. Time frame: 35 months
Funding	Supported by AstraZeneca
Declarations of interest	"Drs. Kjekshus, Barrios, Böhm, Cornel, Dunselman, Fonseca, Grande, Gullestad, Hjalmarson, Hradec, Kamenský, Mareev, McMurray, Schaufelberger, Vanhaecke, van Veldhuisen, Waagstein, and Wedel report receiving consulting or advisory board fees from AstraZeneca; Drs. Kjekshus, Barrios, Böhm, Cleland, Goudev, Grande, Gullestad, Hjalmarson, Jánosi, Komajda, Korewicki, McMurray, Ranjith, Vanhaecke, and van Veldhuisen, lecture fees from AstraZeneca; Drs. Barrios, Hjalmarson, McMurray, and van Veldhuisen, research grants from AstraZeneca; Dr. Waagstein, having an equity interest in AstraZeneca; and Dr. Wikstrand, being employed as a senior medical advisor at AstraZeneca. No other potential conflict of interest relevant to this article was reported.”
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Optimal assignment procedure (minimization method), with a random element computer‐generated randomization code
Allocation concealment (selection bias)	Unclear risk	No details were found.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Matching placebo. All investigators who were connected with the trial were unaware of study‐group assignments.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All investigators who were connected with the trial were unaware of study‐group assignments except for those on the data and safety monitoring board.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low: 2514 were assigned to receive rosuvastatin and 2497 to receive placebo. During follow‐up, 69 participants who were assigned to the rosuvastatin group and 120 participants who were assigned to the placebo group received open‐label treatment with a statin.
Selective reporting (reporting bias)	Low risk	We compared the protocol and the study publications, and no outcomes were missed.
Other bias	Low risk	No other bias could be found.

Knopp 2006.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 4 yrs Setting: hospital
Participants	Ages eligible for study: 40 to 75 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 2410 Location: "70 centers in 14 countries (Australia, Austria, Canada, Finland, France, Germany, Italy, the Netherlands, New Zealand, Norway, South Africa, Spain, Switzerland, and the US)" Inclusion criteria: "male and female participants, aged 40–75 years, were eligible for inclusion if they had type 2 DM; LDL cholesterol criteria were 1) LDL cholesterol ≤ 140 mg/dl if participants had documented MI or an interventional procedure > 3 months before screening or 2) LDL cholesterol ≤ 160 mg/dl if not. Triglyceride levels were required to be ≤ 600 mg/dl at all visits" Exclusion criteria: "type 1 DM; MI, interventional procedure, or episodes of unstable angina ≤ 3 months before screening; HbA1c (A1C) > 10%; active liver disease or hepatic dysfunction (aspartate or alanine aminotransferase levels ≥ 1.5 the ULN); severe renal dysfunction or nephrotic syndrome; congestive heart failure treated with digoxin; creatine phosphokinase 3 times ULN; BP > 160/100 mmHg; BMI > 35 kg/m2; abuse of alcohol and/or drugs; hypersensitivity to the study medication; participation in another clinical study within 30 days of screening; placebo run‐in compliance rate < 80%; current or planned pregnancy; or use of excluded medications, and drugs associated with increased risk of rhabdomyolysis with statins"
Interventions	Intervention: 10 mg of atorvastatin daily (n = 948) Control: placebo (n = 916)
Outcomes	The primary endpoint was the time to the first occurrence of a composite clinical end point of cardiovascular death (fatal MI, fatal stroke, sudden cardiac death, heart failure, or arrhythmic non‐sudden cardiovascular death), nonfatal or silent MI, nonfatal stroke, recanalization, coronary artery bypass grafting, resuscitated cardiac arrest, or worsening or unstable angina requiring hospitalization. Secondary endpoints included the time to the first occurrence of individual components of the primary composite endpoint, non‐cardiovascular death, TIA, worsening or unstable angina not requiring hospitalization, angina or ischaemic pain requiring hospitalization, surgery for or new diagnosis of PAD, or acute ischaemic heart failure requiring hospitalization. Time frame: 4 years
Funding	Funding was provided by Pfizer.
Declarations of interest	Not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Described as randomized, double‐blind, and placebo‐controlled, but no details provided.
Allocation concealment (selection bias)	Unclear risk	No more details were found.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Double‐blind, parallel‐group study. No more details were found.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	An independent, blinded endpoint committee adjudicated primary and secondary endpoints reported by study investigators, excluding coronary artery bypass grafting and recanalization procedures.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: the double‐blind treatment phase was completed by 78.3% of participants in the atorvastatin group (n = 948) and 76.4% of participants in the placebo group (n = 916); 67.5% in the atorvastatin group and 57.6% in the placebo group were taking study medication at study completion.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Koren 2004.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: open label Primary purpose: prevention Duration of study: follow‐up 51.5 months Setting: hospital
Participants	Ages eligible for study: 31 to 78 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 2442 Location: 16 centers throughout the US Inclusion criteria: "men or women >18 years of age with known CHD defined as a history of acute MI > 3 months before screening, percutaneous transluminal coronary angioplasty > 6 months before screening), coronary artery bypass graft surgery > 3 months before screening, or unstable angina > 3 months before screening" Exclusion criteria: "cerebral infarction of determined rare etiology, infarction associated with catheterization or surgery, and preferred use of statins for the treatment of co‐morbid coronary artery disease"
Interventions	Intervention: atorvastatin dose of 80 mg/day (n = 1217) Control: usual care ‐ participants will be maintained on a cholesterol‐care program prescribed by their primary care physician (n = 1225) "Usual care can represent a varied assortment of lipid‐lowering therapies, including diet, behavior modification, or antihyperlipidemic medication. Atorvastatin will not be excluded from the usual‐care group"
Outcomes	Rate of first acute major coronary event, unstable angina, cardiovascular morbidity and mortality (fatal or nonfatal coronary revascularization procedures, unstable angina, fatal or nonfatal MI, fatal or nonfatal cardiovascular events, fatal or nonfatal coronary events, cardiovascular mortality, and CHD mortality) Time frame: 6 years
Funding	Parke‐Davis and Pfizer Pharmaceuticals
Declarations of interest	"The sponsors initiated the study design and also participated in all aspects of logistical support during the trial, statistical and data analyses, and manuscript preparation. Pfizer Pharmaceuticals reviewed the final manuscript and assisted with editing and revisions, but the authors retained final decisions on content."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No details could be found, even in the study protocol.
Allocation concealment (selection bias)	Unclear risk	No details could be found, even in the study protocol.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open label. No details could be found, even in the study protocol.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	An independent outcomes committee of five cardiologists reviewed and adjudicated all study outcomes.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: a total of 958/1217 participants in the atorvastatin group and 941/1225 in the usual care group completed endpoint assessments.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

LIPID Study 1998.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: open label Primary purpose: prevention Duration of study: follow‐up 6.1 yrs Setting: hospital
Participants	Ages eligible for study: 31 to 75 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 9014 Location: 87 centers (67 in Australia and 20 in New Zealand) Inclusion criteria: "history of MI or hospitalization for unstable angina and initial plasma total cholesterol levels of 155 to 271 mg/dL" Exclusion criteria: "a clinically significant medical or surgical event within 3 months before study entry, cardiac failure, renal or hepatic disease, and the current use of any cholesterol‐lowering agents"
Interventions	Intervention: pravastatin 40 mg daily (n = 4512) Control: placebo (n = 4502) Both groups received advice on following a cholesterol‐lowering diet.
Outcomes	The primary study outcome was mortality from CHD. Secondary outcomes were death from any cause; death from cardiovascular causes; death from CHD or nonfatal MI; MI; stroke; non‐hemorrhagic stroke; coronary revascularization (coronary angioplasty, coronary‐artery bypass surgery, or both); number of days in the hospital; serum lipid levels; and the relation of changes in lipid levels to the occurrence of cardiovascular endpoints. Time frame: 5 years
Funding	Bristol‐Myers Squibb Pharmaceutical Research Institute
Declarations of interest	Not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Double‐blind, randomized trial. A randomized block design was employed, with stratification according to diagnosis of either AMI or UAP.
Allocation concealment (selection bias)	Unclear risk	No details could be found.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Study personnel and participants remained blinded to the results of the central analyses of lipid levels.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	An independent data and safety monitoring committee regularly monitored the progress of the study. Study personnel and participants remained blinded to the results of the central analyses of lipid levels.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low. No participants lost to follow‐up were reported.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Nakamura 2006.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and benefit study Intervention model: parallel assignment Masking: open label Primary purpose: prevention Duration of study: follow‐up 5.3 yrs Setting: hospital
Participants	Ages eligible for study: 40 to 70 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 8214 Location: Japan Inclusion criteria: "men and postmenopausal women aged 40‐70 years with a body weight of 40 kg or more and hypercholesterolemia" Exclusion criteria: familial hypercholesterolemia and a history of CHD or stroke
Interventions	Intervention: diet plus 10 to 20 mg pravastatin daily (n = 3866) Control: diet (n = 3966)
Outcomes	The primary composite endpoint was the first occurrence of CHD, which included fatal and non‐fatal MI, angina, cardiac and sudden death, and a coronary revascularization procedure. Secondary endpoints included stroke, CHD plus cerebral infarction, all cardiovascular events, and total mortality. Data were gathered every 3 to 6 months and recorded on the case report form by the participant’s physician. Time frame: 5 years
Funding	Research funds were provided by the Japanese Ministry of Health, Labor and Welfare for the first 2 years of the study, and thereafter the study was funded by Sankyo Co Ltd, Tokyo.
Declarations of interest	"All authors have received travel grants or speaking honoraria from Sankyo Co Ltd for this study. K Arakawa has stock in Sankyo. H Nakamura, K Arakawa, H Itakura, N Nakaya, S Nishimoto, M Muranaka, A Yamamoto, K Mizuno, and Y Ohashi have received lecture fees from Sankyo or Merk‐Banyu and Pfizer Japan Ltd."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computerised randomization by the permuted‐block method
Allocation concealment (selection bias)	Unclear risk	This study did not report the details of allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open‐labelled, blinded‐endpoint, so participants and personnel were aware of treatment groups. "The continuous follow‐up and treatment of patients is conducted openly in a way that adheres to accepted clinical principles and medical practice."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Open‐labelled, blinded‐endpoint, so outcome assessment blinded. An independent data center monitored data. "Strictly defined endpoints are blinded during the handling procedure, allowing unbiased comparison of therapies and evaluation of the study results. "
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates were high in each group. In the diet group, the proportion of participants who received a statin (mostly pravastatin) was 9%, 25%, and 41% at 1, 5, and 9 years, respectively. This implied that 41% of participants discontinued diet at 9 years. In the diet plus pravastatin group, 95%, 90%, and 89% of participants were receiving pravastatin as assigned at 1, 5, and 9 years, respectively. 11% of participants discontinued at 9 years.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Sever 2003.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: factorial design Masking: double blind Primary purpose: prevention Duration of study: follow‐up 3.3 (IQR 2.7 to 3.7) yrs Setting: hospital
Participants	Ages eligible for study: 40 to 79 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 10,180 Location: "33 sites in Denmark (including Iceland), Finland, Norway, Sweden, UK and Ireland" Inclusion criteria: "men and women aged between 40 years and 79 years were eligible if they had 3 or more risk factors for cardiovascular disease but had no history of MI and were not being treated for angina" Exclusion criteria: "previous MI, currently treated angina, a cerebrovascular event within the previous 3 months, fasting triglycerides higher than 4.5 mmol/L, heart failure, uncontrolled arrhythmias or any clinically important haematological or biochemical abnormality on routine screening"
Interventions	Intervention: atorvastatin 10 mg daily (n = 5168) Control: matching placebo (n = 5137)
Outcomes	AE of interest: muscle‐related, erectile dysfunction, sleep disturbance, and cognitive impairment‐and analyzed all remaining AEs grouped by system organ class; non‐fatal MI plus fatal CHD; total cardiovascular events and procedures; total coronary events; all‐cause mortality; cardiovascular mortality; fatal and non‐fatal stroke; fatal and non‐fatal heart failure Time frame: 4 to 5 years
Funding	The study was supported by the principal funding source, Pfizer, New York, NY, USA. Funding was also provided by Servier Research Group, Paris, France, and Leo Laboratories, Copenhagen, Denmark.
Declarations of interest	All authors declare no competing interests.
Notes	The trialists had planned to follow up for an average of 5 years, but the trial stopped early. "On Sept 2, 2002, the data safety monitoring board recommended that the lipid‐lowering arm of the trial be stopped on the grounds that atorvastatin had resulted in a highly significant reduction in the primary endpoint of CHD events compared with placebo and a significant reduction in the incidence of stroke."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Prospective, randomized, open, blinded endpoint trial with a double‐blinded 2 x 2 factorial component; randomly assigned by computer, with use of minimization procedures at the appropriate co‐ordinating center
Allocation concealment (selection bias)	Low risk	Prospective, randomized, open, blinded endpoint trial with a double‐blinded 2 x 2 factorial component. A statin or fibrate was randomly assigned by computer.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Prospective, randomized, open, blinded endpoint trial with a double‐blinded 2 x 2 factorial component, matching placebo.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Investigators submitted all information relevant to any potential endpoints to the Scandinavian co‐ordinating center for central review of endpoints by the endpoint committee, who were unaware of the treatment assignment.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rate in each group was low: of the remaining 10,240 eligible randomized participants, 60 (33 in the atorvastatin group vs 27 in the placebo group) were excluded from these analyses as they were missing end dates for the blinded phase.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Sola 2006.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 1 yrs Setting: hospital
Participants	Ages eligible for study: > 18 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 108 Location: USA Inclusion criteria: "men and women ages 18 years or older were eligible for enrolment if they had: 1) NYHA functional class II to IV Heart Failure due to a nonischemic etiology; 2) LVEF of 35%, as documented by echocardiography or ventriculography during the 1 year before enrolment; and 3) stable doses of heart failure medications for 3 months before enrolment." Exclusion criteria: "1) had been receiving a statin during the 6 months before enrolment; 2) had a prior adverse event related to statin use; or 3) had DM. Patients were classified as having nonischemic cardiomyopathy if they had no prior clinical history of MI and no coronary artery stenoses > 50% on cardiac catheterization performed during the 1 year before enrolment"
Interventions	Intervention: atorvastatin (20 mg/day) (n = 54) Control: placebo (n = 54)
Outcomes	The primary endpoint of the study was change in LVEF, as determined by transthoracic echocardiography. Secondary endpoints included changes in several markers of inflammation and/or oxidation, including high‐sensitivity CRP, IL‐6, TNF‐ RII, and E‐SOD. Time frame: 1 year
Funding	Not reported
Declarations of interest	"Dr Khan has been an advisory board member for Sanofifi‐Aventis and Bristol Myers Squibb and on the speakers' bureau for Sanofi Aventis, Bristol Myers Squibb, and Takeda Pharmaceuticals."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No details were reported.
Allocation concealment (selection bias)	Unclear risk	No details were reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants were randomized in a double‐blinded fashion to either atorvastatin or matching placebo for a 12‐month period. The images were reviewed by two independent echocardiographers who were blinded to the participant’s clinical status during the study period, and any discrepancies in interpretations were resolved.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The images were reviewed by two independent echocardiographers who were blinded to the participant's clinical status during the study period.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: 54 each were randomized to the placebo and atorvastatin groups. A total of 89 participants (placebo 43, atorvastatin 46) completed the 12‐month treatment period.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Talasaz 2023.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: factorial design Masking: blinded (participants) Primary purpose: prevention Duration of study: follow‐up 90 days Setting: hospital
Participants	Ages eligible for study: 45 to 68 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 605 Location: 11 Iranian hospitals in Tehran, Tabriz, and Karaj Inclusion criteria: "adults aged ≥ 18 years with reverse tPCR confirmed Covid‐19 who were admitted to an ICU and had no definite indication for therapeutic anticoagulation or baseline statin use" Exclusion criteria: "their estimated survival was less than 24 hours, they weighed < 40 kg, they had major bleeding or serious bleeding diathesis within 30 days from enrolment, their liver enzyme test results were > 5 times the ULN, they had active liver disease, and their creatine kinase concentration was > 500 U/L"
Interventions	Intervention: atorvastatin 20 mg once daily (n = 290) Control: matching placebo (n = 297)
Outcomes	Composite of venous or arterial thrombosis, treatment with extracorporeal membrane oxygenation, or all‐cause mortality within 30 days from randomization. Secondary outcomes included the individual components of the primary outcome (VTE, arterial thrombosis, and all‐cause mortality) and ventilator‐free days. Exploratory outcomes included objectively clinically diagnosed type I acute MI, stroke, and acute peripheral arterial thrombosis, the proportion of participants discharged alive from the ICU, length of hospital stay in the ICU, incident atrial fibrillation, and new renal replacement therapy. Time frame: 90 days
Funding	The Rajaie Cardiovascular Medical and Research Center
Declarations of interest	"BB reports that he is a consulting expert, on behalf of the plaintiff, for litigation related to two specific brand models of IVC filters. DJ has served as an advisor or consultant for Bayer HealthCare Pharmaceuticals, Boehringer Ingelheim, Bristol‐Myers Squibb, Daiichi Sankyo, Leo Pharma, Pfizer, ROVI, and Sanofi; served as a speaker or a member of a speakers’ bureau for Bayer HealthCare Pharmaceuticals, Boehringer Ingelheim, Bristol‐Myers Squibb, Daiichi Sankyo, LeoPharma, ROVI, and Sanofi; and received grants for clinical research from Daiichi Sankyo, Sanofi, and ROVI. AG received payment from the Arnold & Porter Law Firm for work related to the Sanofi clopidogrel litigation and from the Ben C Martin Law Firm for work related to the Cook inferior vena cava filter litigation. AG holds equity in a healthcare telecardiology start‐up, Heartbeat Health, and received consulting fees from Edwards LifeSciences. MVM was supported by a grant from the National Institutes of Health/National Heart, Lung, and Blood Institute to Columbia University Irving Medical Center (T32HL007854). SSS reports receiving personal fees from Janssen and Chiesi and grants from the American Heart Association outside the submitted work. SAP reports receiving grants from Abbott Vascular, Boston Scientific, Surmodics, and TriReme Medical; non‐financial support from Cordis, Medtronic, Philips, and Cardiovascular Systems; and personal fees from Terumo, Abiomed, Inari, and Penumbra outside the submitted work. GP has received research grant support to Brigham and Women’s Hospital from EKOS, a BTG International Group company, Bayer, the Bristol Myers Squibb/Pfizer Alliance, Portola, and Janssen. He has received consulting fees from Amgen, Pfizer, Boston Scientific, and Thrombolex. AJK reports receiving institutional funding to Columbia University and/or the Cardiovascular Research Foundation from Medtronic, Boston Scientific, Abbott Vascular, Abiomed, Cardiovascular Systems, CathWorks, Siemens, Philips, and ReCor Medical, including fees paid to Columbia University and/or the Cardiovascular Research Foundation for speaking engagements and/or consulting; consulting fees from Neurotronic; and travel expenses/meals from Medtronic, Boston Scientific, Abbott Vascular, Abiomed, Cardiovascular Systems, CathWorks, Siemens, Philips, ReCorMedical, Chiesi, OpSens, Zoll, and Regeneron. BWVT has received research support from Novartis, Swedish Orphan Biovitrum, Olatec Therapeutics, and Serpin Pharma and has been a consultant for R‐Pharm and Serpin Pharma. GWS reports receiving personal fees from Terumo, Cook, TherOx, Reva, Vascular Dynamics, Robocath, HeartFlow, Gore, Ablative Solutions, Matrizyme, Miracor Neovasc, V‐wave, Abiomed, MAIA Pharmaceuticals, Shockwave, Vectorious, Cardiomech, and Elucid Bio; equity/equity options from Applied Therapeutics, MedFocus, Biostar, Aria, Cagent, and Cardiac Success; personal fees and equity/equity options from Spectrawave, Valfix, and Ancora; and personal fees, equity/equity options, and honorarium from Orchestra Biomed, and outside the submitted work. GYHL reports being a consultant and speaker for BMS/Pfizer, Boehringer Ingelheim and Daiichi‐Sankyo. No fees are received personally. HMK reports receiving personal fees from UnitedHealth, IBM Watson Health, Element Science, Aetna, Facebook, Siegfried & Jensen Law Firm, Arnold & Porter Law Firm, Martin/Baughman Law Firm, F‐Prime, and the National Center for Cardiovascular Diseases, Beijing; being the cofounder of HugoHealth, a personal health information platform, and Refactor Health, an enterprise health care artificial intelligence augmented data management company; receiving contracts from the Centers for Medicare & Medicaid Services, through Yale New Haven Hospital, to develop and maintain measures of hospital performance; and receiving grants from Medtronic, the US Food and Drug Administration, Johnson & Johnson, and the Shenzhen Center for Health Information outside the submitted work. SZG has received research support from Bayer, Boehringer Ingelheim, Bristol Myers Squibb, Boston Scientific, Daiichi‐Sankyo, Janssen, the National Heart, Lung, and Blood Institute, and the Thrombosis Research Institute; and has received consulting fees from Bayer, Agile, Boston Scientific, and Boehringer Ingelheim."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A centralized computer‐based system with a block size of 4
Allocation concealment (selection bias)	Low risk	Access to the allocation sequence was concealed from the site clinicians.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	The study drug and placebo were identical in appearance and given to treatment teams (in the hospital) or the participants or their caregivers on hospital discharge.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The study investigators and participants remained blinded to assigned treatments until the completion of the analyses.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: atorvastatin 20 mg once daily (average compliance 74%); matching placebo (average compliance 77%).
Selective reporting (reporting bias)	Low risk	We compared the protocol and the study publications, and no outcomes were missed.
Other bias	Low risk	No other sources of bias detected.

Tavazzi 2008.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 3.9 yrs Setting: hospital
Participants	Ages eligible for study: aged 18 yrs or older Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 4631 Location: 326 cardiology and 31 internal medicine centers in Italy Inclusion criteria: "men and women aged 18 years or older, with symptomatic heart failure that was classified as NYHA functional class II‐IV, who were being treated according to ESC guidelines" Exclusion criteria: "known hypersensitivity to study treatment; presence of any noncardiac comorbidity that was unlikely to be compatible with a sufficiently long follow‐up; treatment with any investigational agent within 1 month before randomization; acute coronary syndrome or a revascularization procedure within 1 month before randomization; planned cardiac surgery, expected to be done within 3 months after randomization; significant liver disease; serum creatinine concentration greater than 221 μmol/L; alanine and aspartate transaminase concentrations more than 1.5 x ULN; creatine phosphokinase concentrations above the ULN; and pregnant or lactating women or women of childbearing potential who were not adequately protected against becoming pregnant"
Interventions	Intervention: rosuvastatin 10 mg daily (n = 2285) Control: placebo (n = 2289)
Outcomes	Time to death, and time to death or admission to hospital for cardiovascular reasons; cardiovascular mortality, cardiovascular mortality or admission for any reason, sudden cardiac death, admission for any reason, admission for cardiovascular reasons, admission for heart failure, MI, and stroke. Time frame: 54 months
Funding	Società Prodotti Antibiotici (SPA; Italy), Pfizer, Sigma Tau, and AstraZeneca
Declarations of interest	"LT, GT, APM, RM, and MGF received research support and honoraria for lectures from AstraZeneca. GLN and MP received honoraria for lectures from AstraZeneca. RL, SB, and DL received research support from AstraZeneca."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized, double‐blind, placebo‐controlled trial. Concealed, computerized telephone randomization system
Allocation concealment (selection bias)	Low risk	A concealed, computerized telephone randomization system
Blinding of participants and personnel (performance bias) All outcomes	Low risk	All participants and study personnel were blinded to treatment.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All events recorded in the study were adjudicated blindly by an ad hoc committee on the basis of pre‐agreed definitions and procedures. All participants and study personnel were blinded to treatment.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low (intervention vs control: 25/2314 vs 30/2317.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Wanner 2005.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double blind Primary purpose: prevention Duration of study: follow‐up 4 yrs Setting: hospital
Participants	Ages eligible for study: 18 to 80 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 1255 Location: 178 centers in Germany Inclusion criteria: "participants with type 2 DM 18 to 80 years of age who had been receiving maintenance hemodialysis for less than 2 years were enrolled" Exclusion criteria: "levels of fasting serum LDL cholesterol of less than 80 mg/dl or more than 190 mg/dl, triglyceride levels greater than 1000 mg/dl; liver function values more than 3 times the ULN or equal to those in patients with symptomatic hepatobiliary cholestatic disease; hematopoietic disease or systemic disease unrelated to end‐stage renal disease; vascular intervention, congestive heart failure, or MI within the three months preceding the period of enrolment; unsuccessful kidney transplantation; and hypertension resistant to therapy"
Interventions	Intervention: 20 mg of atorvastatin per day (n = 619) Control: matching placebo (n = 636)
Outcomes	The primary endpoint was a composite of death from cardiac causes, fatal stroke, nonfatal MI, or nonfatal stroke, whichever occurred first. Only one event per participant was included in the analysis. Secondary endpoints included death from all causes, all cardiac events combined, and all cerebrovascular events combined. Time frame: 4 years
Funding	Pfizer
Declarations of interest	"The committee members and investigators did not receive remuneration for conducting the study, except for reimbursement of costs to participate in scientific meetings. Dr. Wanner reports having received consulting fees and lecture fees from Genzyme; Dr. März, consulting fees, lecture fees, a research grant and stock options from Pfizer; and Dr. Mann, lecture fees from Aventis, Roche, and Janssen Cilag. Dr Ritz is a member of the safety board of a trial sponsored by AstraZeneca and reports having received consulting fees from the company."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A multicenter, randomized, double‐blind, prospective study. "A computer‐generated randomization code was prepared by a central Pfizer unit that was independent of local study personnel."
Allocation concealment (selection bias)	Low risk	A multicenter, randomized, double‐blind, prospective study. "A computer‐generated randomization code was prepared by a central Pfizer unit that was independent of local study personnel."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"To maintain blinding, a randomly selected participant from the placebo group received an identical dose reduction."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"All committee members were blinded to the treatment assignments until 13 August 2004."
Incomplete outcome data (attrition bias) All outcomes	Low risk	The discontinued rates in each group were low (1/619 vs 0/636).
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

Yusuf 2016.

Study characteristics
Methods	Study type: interventional Study design: RCT Control: placebo control Endpoint classification: safety and efficacy study Intervention model: 2 x 2 factorial trial Masking: double blind Primary purpose: prevention Duration of study: follow‐up 5.6 yrs Setting: hospital
Participants	Ages eligible for study: > 55 yrs Sexes eligible for study: both Accepts healthy volunteers: no Total number of participants: 12,705 Location: 228 centers in 21 countries Inclusion criteria: "men 55 years of age or older and women 65 years of age or older who had at least one of the following cardiovascular risk factors: elevated waist‐to‐hip ratio, history of a low level of HDL cholesterol, current or recent tobacco use, dysgalactia, family history of premature coronary disease, and mild renal dysfunction" Exclusion criteria: "participants with cardiovascular disease and those with an indication for or contraindication to statins, angiotensin‐receptor blockers, angiotensin‐converting‐enzyme inhibitors, or thiazide diuretics"
Interventions	Intervention: rosuvastatin at a dose of 10 mg per day (n = 6361) Control: placebo (n = 6344)
Outcomes	Composite of death from cardiovascular causes, nonfatal MI, or nonfatal stroke, resuscitated cardiac arrest, heart failure, and revascularization, angina with evidence of ischemia Time frame: 5.6 years
Funding	The Canadian Institutes of Health Research and AstraZeneca
Declarations of interest	"Dr. Bosch reports grant support from the Canadian Institutes of Health Research and AstraZeneca during the conduct of the study; and personal fees from Bristol‐Myers Squibb outside the submitted work. Dr. Held reports grant support, personal fees, and non‐financial support from AstraZeneca, and grant and non‐financial support from GlaxoSmithKline, Bristol‐Myers Squibb/Pfizer, and Bayer outside the submitted work. Dr. Khunti reports grant support and personal fees from AstraZeneca, Sanofi, Novartis, Novo Nordisk, Boehringer Ingelheim, Eli Lilly, Merck Sharp & Dohme, and Janssen outside the submitted work. Dr. Leiter reports grant support from GlaxoSmithKline and Pfizer, grant support and personal fees from Amgen, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck, Novo Nordisk, and Sanofi, and personal fees from Aegerion outside the submitted work. Dr. Lewis reports grant support and personal fees from Amgen, Pfizer, Merck, and AstraZeneca outside the submitted work. Dr. Lonn reports grant support from the Canadian Institutes of Health Research and AstraZeneca during the conduct of the study; grant support from Bayer, GlaxoSmithKline, Merck Schering, and Eli Lilly, grant support and personal fees from Amgen and Sanofi, and personal fees from Novartis, Cadila Pharmaceuticals, and Servier Canada outside the submitted work. Dr. McKelvie reports personal fees from AstraZeneca during the conduct of the study. Dr. Parkhomenko reports personal fees from the PHRI during the conduct of the study; and grant support from Sanofi, Amgen, and AstraZeneca outside the submitted work."
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Centrally concealed randomization procedure
Allocation concealment (selection bias)	Low risk	Centrally concealed randomization procedure
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Described as double‐blind, placebo‐controlled. All participants were blinded to the treatment allocations.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No details were found.
Incomplete outcome data (attrition bias) All outcomes	High risk	The discontinued rates in each group were high: in the rosuvastatin group, 88.0% were taking the assigned regimen at 1 year, 83.5% at 3 years, and 75.5% at 5 years; the corresponding rates in the placebo group were 87.8%, 83.0%, and 73.2%.
Selective reporting (reporting bias)	Low risk	Although the incidence of VTE was not reported, this study reported all outcomes outlined in their study protocols. The incidence of VTE was retrieved from meta‐analyses (Kunutsor 2017; Rahimi 2012).
Other bias	Low risk	No other bias could be found.

ACE: angiotensin‐converting enzyme; ADAS‐cog: Alzheimer's Disease Assessment Scale‐cognitive sub‐scale; ADCS‐CGIC: Cooperative Study Clinical Global Impression of Change; AE: adverse event; ALERT: Assessment of Lescol in Renal Transplant; ALT: alanine aminotransferase; AMI: acute myocardial infarction; AST: aspartate aminotransferase; BP: blood pressure; BMI: body mass index; CABG: coronary‐artery bypass grafting; CHD: coronary heart disease; CIMT: carotid intima‐media thickness; CRP: c‐reactive protein; CT: computed tomography; CVA: cerebrovascular accident; DM: diabetes mellitus; ESC: European Society of Cardiology; E‐SOD: erythrocyte superoxide dismutase; GFR: glomerular filtration rate; HbA1c: hemoglobin A1C test to measure the average blood sugar (glucose) level; HDL(‐c): high‐density lipoprotein (‐c); ICU: intensive care unit; IL‐6: interleukin 6; IQR: interquartile range; LDL(‐c): low‐density lipoprotein (‐c); LVEF: left ventricular ejection fraction; MI: myocardial infarction; n: number; NYHA: New York Heart Association; PAD: peripheral arterial disease; PCI: percutaneous coronary intervention; PHRI: Population Health Research Institute; PVD: peripheral vascular disease; RCT: randomized controlled trial; SAS: Statistical Analysis System; TIA: transient ischemic attack; TC: total cholesterol; TNF‐ RII: tumor necrosis factor‐ receptor II; tPCR: transcription polymerase chain reaction; ULN: upper limit of normal; UAP: unstable angina pectoris; UCL: University College London; vs: versus; VTE: venous thromboembolism; yrs: years

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Barrett 2023	Ineligible intervention (acetylsalicylic acid + statin (experimental) and identical placebos (control))
Delluc 2022	Ineligible population (assessed the secondary prevention of VTE)
Orsi 2019	Ineligible population (assessed the secondary prevention of VTE)

PE: pulmonary embolism; VTE: venous thromboembolism

Characteristics of ongoing studies [ordered by study ID]

ChiCTR2100048407.

Study name	A RCT of rosuvastatin in the prevention of DVT in total knee arthroplasty
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: unclear Primary purpose: prevention
Participants	Inclusion criteria: "Patients who are adults requiring total knee replacement for any reason and who consent to this study." Exclusion criteria: "Patients allergic to rosuvastatin; currently on anticoagulation therapy or receiving statin therapy or have used statins within 3 months; history of DVT or PE; preoperative liver transaminase (AST or ALT) levels exceeding 3 times the UNL; CK levels exceed 3 x UNL; family history of thrombosis, currently receiving HRT"
Interventions	Experimental arm: rosuvastatin + fondaparinux sodium Control arm: fondaparinux sodium
Outcomes	Primary outcome measures: DVT; serum thromboelastography; D‐dimer; CRP; interleukin 1,6; TNF‐a Secondary outcome measures: total blood loss; hidden blood loss Time frame: not reported
Starting date	August 2021
Contact information	Name: Gao Zhixiang Address: 190 East Section of Jiannan Road, Youxian District, Mianyang, Sichuan Telephone: + 86 19113689863 Email: jwkgzx@126.com Affiliation: The Third Hospital of Mianyang, Sichuan Mental Health Center
Notes	This study is still recruiting. Funding: this study was funded by Science and technology project of Sichuan province

NCT00259662.

Study name	High‐dose perioperative statins for prevention of DVT
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: double‐blind Primary purpose: prevention
Participants	Inclusion criteria: gynaecologic tumour scheduled for resection Exclusion criteria: "prior reaction to statins; renal insufficiency; liver disease; history of alcoholism; prior history of DVT or hypercoagulability; concurrent medications that significantly affect cytochrome P450 3A4; breast feeding or pregnancy"
Interventions	Experimental arm: atorvastatin Control arm: unclear
Outcomes	Primary outcome measures: decrease in incidence of DVT Secondary outcome measures: decrease in inflammatory mediator release Time frame: not reported
Starting date	November 2005
Contact information	Name: Ala S Haddadin, MD (principal investigator) Address: New Haven Hospital, Yale, New Haven, Connecticut, United States Telephone: not reported Email: ala.haddadin@yale.edu Affiliation: Yale University
Notes	This study has been finished, but the authors did not publish their data. We emailed them to ask for the VTE data, but no response was received. Funding: not reported

NCT01021488.

Study name	Rosuvastatin for prevention of DVT in patients undergoing TKRA: STOP DVT ‐ a prospective RCT
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: open label Primary purpose: prevention
Participants	Inclusion criteria: patients who are going to receive TKRA for any cause; > 19 years old Exclusion criteria: "patients with cancer; patients receiving anticoagulant agents for any cause, current statin users; expecting survival from other co‐morbidity < 1 year; bed‐ridden patient; AST, ALT > 3 times of UNL; CK > UNL; pregnancy; patients who receive HRT"
Interventions	Experimental arm: rosuvastatin + enoxaparin Rosuvastatin 20 mg/d for 7 d before and 7 d after index surgery, TKRA Enoxaparin 40 mg SQ/d 12 hours before TKRA and from 1 d to 7 d after TKRA should be administered at the same time with rosuvastatin Control arm: enoxaparin only (active comparator); enoxaparin 40 mg SQ/d only starting 12 hours before TKRA and from day 1 to 7 after index surgery
Outcomes	Primary outcome measures: development of DVT diagnosed and confirmed by CT angiography at lower extremities (time frame: 7 d after index surgery) Secondary outcome measures: D‐dimer, lipid panel (total cholesterol, TG, HDL, LDL), hsCRP, CK, transaminase, ALP (time frame: 7 d, 1 m, 2 m after index surgery) Time frame: 7 days
Starting date	October 2009
Contact information	Name: Sang‐Ho Jo, MD (principal investigator) Address/affiliation: Hallym University Sacred Heart Hospital, Department of Cardiology and Orthopedic Surgery, Anyang‐si, Gyeonggi‐do, Republic of Korea Telephone: 82‐31‐380‐3722 Email: sophi5@medimail.co.kr Affiliation: Hallym University Medical Center
Notes	This study has finished, but the authors did not publish their outcomes for VTE. We emailed them to ask for the VTE data, but no response was received. Funding: not reported

NCT01063426.

Study name	Re‐STOP DVT: reload of high dose atorvastatin for preventing DVT in statin users
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: safety and efficacy study Intervention model: parallel assignment Masking: open label Primary purpose: prevention
Participants	Inclusion criteria: "patients who are going to receive TKRA from any cause; < 19 years old" Exclusion criteria: "patients with cancer; patients receiving anticoagulant agents for any cause; current statin users; expected survival from other co‐morbidity < 1 year; bed‐ridden patient; AST, ALT > 3 times of UNL, CK > UNL; pregnancy; patients who receive HRT"
Interventions	Experimental: atorvastatin + enoxaparin arm High‐dose atorvastatin arm before TKRA + conventional enoxaparin High‐dose atorvastatin 80 mg/d for 7 d after TKRA. At the same time, enoxaparin 40 mg SQ/day 12 hours before TKRA and from 1 d to 7 d after TKRA should be administered Active comparator: conventional enoxaparin; conventional enoxaparin before 12 hour and on 1 to 7 d after TKRA; enoxaparin 40 mg subcutaneously daily 12 hours before TKRA and on d 1 to d 7 after TKRA should be administered
Outcomes	Primary outcome measures: development of DVT diagnosed and confirmed by CT angiography at lower extremities (time frame: 7 d after index surgery) Secondary outcome measures: D‐dimer, lipid panel (total cholesterol, TG, HDL, LDL), hsCRP, CK, transaminase, ALP (time frame: 7 d, 1 m, 2 m after index surgery) Time frame: 7 days
Starting date	November 2009
Contact information	Name: Sang‐Ho Jo, MD (principal investigator) Address: Hallym University Sacred Heart Hospital, Department of Cardiology and Orthopedic Surgery, Anyang‐si, Gyeonggi‐do, Republic of Korea Email: sophi5@medimail.co.kr Affiliation: Hallym University Medical Center
Notes	This study has been finished, but they did not publish their data. We have emailed the authors for VTE data, however, they did not reply. Funding: not reported

NCT01524653.

Study name	Detecting the impact of statin therapy on lowering risk of venous thrombo‐embolic events (DISOLVE)
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: efficacy study Intervention model: cross‐over assignment Masking: double‐blind (participant, caregiver, investigator, outcomes assessor) Primary purpose: prevention
Participants	Inclusion criteria: "Adult patients > 18 years old with locally advanced or metastatic cancers who are about to start or are already receiving any systemic chemotherapy or targeted therapy; estimated overall survival of ≥ 6 months; anticipated duration of therapy ≥ 9 weeks or ≥12 weeks. Systemic therapy is allowed to change if necessary or to terminate, during this period" Exclusion criteria: "anti‐thrombotic therapy including warfarin, dabigatran, LMWH or UFH; anti‐angiogenic therapy with thalidomide or lenalidomide; patients starting hormonal therapy exclusively; statin use within 3 months prior to enrolment; adjuvant therapy in patients who have already received curative‐intent local therapy; patients with glioblastoma starting adjuvant chemotherapy are an exception given the high likelihood of residual disease and risk of VTE in this population; Asian descent as assessed by history; urinary creatinine clearance of less than 40 mL/min based on reported MDRD GFR, present in Fletcher Allen Health Care metabolic profile reports, during the 14 d screening period; AST or ALT elevation of greater than 3 x UNL during the 14 d screening period; patients with a known history of statin intolerance that was accompanied by severe adverse reaction; patients who are currently participating in another clinical trial involving an investigational medication if there is a known or suspected drug interaction with rosuvastatin or the statin class, or if the investigational agent is known or suspected to be associated with a significantly increased risk of thrombosis"
Interventions	Experimental arm: rosuvastatin first, placebo last Rosuvastatin 20 mg daily Placebo 20 mg daily Control arm: placebo first, rosuvastatin last Rosuvastatin 20 mg daily Placebo 20 mg daily
Outcomes	Primary: decrease in D‐dimer level Secondary: change in Factor VIII; change in soluble P‐selectin; change in C‐reactive protein; change in peak thrombin generation; adverse events (CTCAE v4) associated with rosuvastatin therapy; liver toxicity and rhabdomyolysis; VTE; clinical signs of VTE; change in plasminogen activator inhibitor‐1 activity; change in plasminogen activator inhibitor‐1 protein concentration; change in tissue factor; change in Factor XIa Time frame: 21 weeks; baseline, 3 to 4 weeks, 6 to 9 weeks, 9 to 13 weeks
Starting date	March 2012
Contact information	Name: Steven Ades, MD, MSc (principal investigator) Address: Fletcher Allen Health Care, Burlington, Vermont, United States Telephone: not reported Email: not reported Affiliation: University of Vermont
Notes	This study has been finished, but the researchers did not publish their data. We emailed the authors for VTE data; however, they did not reply. Funding: not reported

NCT02285738.

Study name	Anti‐platelet and statin therapy to prevent cancer‐associated thrombosis
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: efficacy study Intervention model: parallel assignment Masking: none (open label) Primary purpose: prevention
Participants	Inclusion criteria: "Histologic diagnosis of malignancy of a solid organ or lymphoma; Planned to initiate a new systemic chemotherapy regimen (including patients starting on first chemotherapy or patients previously treated but starting on a new regimen); VTE Risk Score ≥1; Written, informed consent." Exclusion criteria: "Hematologic malignancies including acute and chronic leukemias, myelodysplastic syndromes, lymphoma and myeloma; Primary brain tumors; Active bleeding or high risk of bleeding in the opinion of the investigator; Hepatic dysfunction (elevated transaminases or bilirubin > 3 times normal); Planned stem cell transplant; Life expectancy < 6 months; Acute or chronic renal insufficiency with creatinine clearance < 30 mL/min; Pregnancy; Known allergy to or prior intolerance of aspirin and/or simvastatin; Ongoing anticoagulant, statin and/or anti‐platelet therapy."
Interventions	Active comparator: aspirin + asprin/simvastatin + observation (ASO) Aspirin 81 mg/day for 4 weeks followed by 2‐week washout, followed by 4 weeks of aspirin 81 mg/day with daily dose of simvastatin with a 2‐week washout period, ending with 4 weeks of observation Experimental 1: aspirin + observation + asprin/simvastatin (AOS) Aspirin 81 mg/day for 4 weeks followed by 2‐week washout, followed by 4 weeks of observation with 2‐week washout, ending with aspirin 81 mg/day with daily dose of simvastatin Experimental 2: aspirin/simvastatin + observation + asprin (SOA) Aspirin 81 mg/day for 4 weeks with daily dose of simvastatin followed by 2‐week washout, followed by 4 weeks of observation with 2‐week washout, ending with aspirin 81 mg/day Experimental 3: aspirin/simvastatin + asprin + observation (SAO) Aspirin 81 mg/day for 4 weeks with daily dose of simvastatin followed by 2‐week washout, followed by 4 weeks of aspirin 81 mg/day and a 2‐week washout, ending with observation for 4 weeks Experimental 4: observation + aspirin/simvastatin + Asprin (OSA) Observation for 4 weeks with 2‐week washout, followed by aspirin 81 mg/day for 4 weeks with daily dose of simvastatin followed by 2‐week washout, ending with aspirin 81 mg/day for 4 weeks Experimental 5: observation + aspirin + asprin/simvastatin (OAS) Observation for 4 weeks with 2‐week washout, followed by aspirin 81 mg/day for 4 weeks followed by 2‐week washout, ending with aspirin 81 mg/day for 4 weeks with daily dose of simvastatin
Outcomes	Primary outcome measures: change in average sP‐selectin levels; change in sP‐selectin levels as an indicator of measure efficacy (time frame: at 16 weeks of treatment) Secondary outcome measures Frequency of major bleeding complications or clinically significant non‐bleeding complications per participant (time frame: at 17 weeks of treatment); Change in average platelet factor 4; change in average CD40 ligand; change in average serum thromboxane B2; change in average serum VEGF; change in average serum angiopoietin‐2; change in average serum hepatocyte growth factor; change in average serum PECAM; change in average serum PDGF; change in average plasma F1.2; change in average plasma TAT complexes; change in average plasma D‐dimer (time frame: at 16 weeks of treatment); Change in the number of thrombotic events (time frame: at 17 weeks of treatment) Time frame: 17 weeks
Starting date	7 November 2014
Contact information	Name: Alok A Khorana, MD (principal investigator) Address: Cleveland Clinic Taussig Cancer Institute, Case Comprehensive Cancer Center, Cleveland, Ohio, United States, 44195 Telephone: not reported Email: not reported Affiliation: Case Comprehensive Cancer Center
Notes	This study has been finished, but the researchers did not publish their data. We emailed the authors for VTE data, but they did not reply. Funding: not reported

NCT03532139.

Study name	Pilot study of rosuvastatin and enoxaparin thromboprophylaxis following ovarian cancer surgery (O‐STAT Study)
Methods	Study type: interventional Study design: RCT Allocation: randomized Endpoint classification: efficacy study Intervention model: parallel assignment Masking: open label Primary purpose: prevention
Participants	Inclusion criteria: "histologic diagnosis of ovarian, fallopian or primary peritoneal cancer; Age ≥ 18 years; ECOG performance status ≤2; Life expectancy of greater than 6 months; Participants must have normal organ and marrow function." Exclusion criteria: "participants who are receiving any other investigational agents; participants with known brain metastases; active bleeding or high risk of bleeding; history of heparin‐induced thrombocytopenia; any history of significant haemorrhage; the presence of coagulopathy; uncontrolled hypothyroidism; qualifying TSH; Familial bleeding diathesis; known diagnosis of disseminated intravascular coagulation; currently taking statin; currently receiving anticoagulant therapy; current use of aspirin, Clopidogrel, cilostazol, aspirin‐dipyridamole; known Asian descent due to altered metabolism of statins. Concomitant use of the following drugs: cyclosporine, fibrates, niacin, gemfibrozil, ketoconazole, spironolactone, cimetidine, warfarin, erythromycin, or protease inhibitors. Known recent history of heavy alcohol use. History of rhabdomyolysis while on statin therapy. Known active Hepatitis C or active Hepatitis B infection. Uncontrolled intercurrent illness including, but not limited to, ongoing or active infection, symptomatic congestive heart failure, unstable angina pectoris, cardiac arrhythmia, or psychiatric illness/social situations that would limit compliance with study requirements. Pregnant women are excluded from this study due to the potential for teratogenic effects on the human fetus. Because there is an unknown but potential risk of adverse events in nursing infants secondary to the treatment of the mother with rosuvastatin, breastfeeding should be discontinued."
Interventions	Experimental arm 1: enoxaparin; enoxaparin is administered subcutaneous daily Experimental arm 2: enoxaparin + rosuvastatin Enoxaparin is administered subcutaneous daily Rosuvastatin is administered daily orally starting on day 15 Control arm: thromboprophylaxis is administered per clinician discretion, standard of care therapy
Outcomes	Primary outcome measures: comparison of differences in circulating tissue factor bearing microparticles between study arms, concentration of tissue factor‐bearing microparticles Secondary outcome measures: point estimate of the rates of VTE following ovarian surgery in each study arm, VTE rate; comparison of D‐dimer values across study arms; compare the rates of VTE between study arms; compare CRP between study arms; compare concentrations of TFMP, D‐dimer, CRP at study time points; assess incidence of major haemorrhage and clinically relevant bleeding as defined by the International Society of Thrombosis and Haemostasis; estimate the overall rate of any VTE Time frame: 60 days
Starting date	25 July 2018
Contact information	Name: Rushad Patell, MD (principal investigator) Address: Beth Israel Deaconess Medical Center and Massachusetts General Hospital, Boston, Massachusetts, United States Telephone: not reported Email: not reported Affiliation: Beth Israel Deaconess Medical Center
Notes	This is an ongoing study. No results have been posted, as yet. Funding: not reported

ALP: alkaline phosphatase; ALT: alanine aminotransferase; AST: aspartate aminotransferase; CK: creatine kinase; CRP: C‐reactive protein; CT: computed tomography; CTCAE v4: Common Terminology Criteria for Adverse Events V4; d: day(s); DVT: deep vein thrombosis; ECOG: Eastern Cooperative Oncology Group; HDL: high‐density lipoprotein; HRT: hormone replacement therapy; hsCRP: high‐sensitivity C‐reactive protein; GFR: glomerular filtration rate; LDL: low‐density lipoprotein; LMWH: low‐molecular‐weight heparins; m(s): month(s); MDRD: modification of diet in renal disease; PDGF: platelet‐derived growth factor; PECAM: platelet endothelial cell adhesion molecule; PE: pulmonary embolism; RCT: randomized controlled trial; SERM: selective estrogen receptor modulators; SQ: subcutaneous; TAT: thrombin–antithrombin; TKRA: total knee replacement arthroplasty; TFMP: tissue factor bearing microparticles; TG: triglyceride; TNF‐a: tumor necrosis factor alpha; TSH: thyroid stimulating hormone; UFH: unfractionated heparin; UNL: upper limit of normal; VEGF: vascular endothelial growth factor; VTE: venous thromboembolism

Differences between protocol and review

Changes to methods

We have made the following changes to the methods published in our protocol (Zhang 2021).

Types of participants

In the protocol, we classified the population into two categories: healthy individuals and individuals considered to be at risk (those who underwent surgery, non‐surgical cancer patients, and surgical cancer patients). For the review, we grouped the population based on different health statuses. The studies were sorted according to the health conditions of the participants to reduce clinical heterogeneity.

Types of outcome measures

For clarity, we have specified the timing of the outcome assessment as follows: “The timing of the outcome assessment was classified as short‐term (≤ one year), medium‐term (one to five years), and long‐term (> five years) follow‐up”.

Selection of studies

In line with the guidance in the Cochrane Handbook (Lefebvre 2022), we only record a selection of studies excluded at full‐text assessment in the Characteristics of excluded studies table, rather than all studies.

Unit of analysis issues

For clarity, we specified the unit of analysis for cross‐over RCTs and cluster‐RCTs, as well as for studies with multiple interventions. In the case of cross‐over RCTs, the incidence of outcomes within each group from the initial period, prior to the act of crossing over, was combined. For cluster‐randomized trials, groups of individuals were collectively randomized to the same intervention, and the incidence of outcomes within each participant group was combined. For multiple‐arm trials, intervention groups were directly combined.

Dealing with missing data

In the protocol, we wrote, “If data are missing because participants dropped out or there were losses to follow‐up, we plan to conduct a primary analysis based on the provided data, and a sensitivity analysis with missing data imputed based on the worst‐case and best‐case scenario”. For the review, we wrote, “If data were missing because of dropping out of participants or losses to follow‐up, we conducted a primary analysis based on ITT analysis data”, as we thought the worst‐case and best‐case scenarios might not be the best way to deal with the missing data.

Assessment of heterogeneity

In the review, we clarified that we also separately investigated clinical and methodological heterogeneity. We further adjusted the substantial heterogeneity from 40% in the protocol to 40 to 50% in the review for consistency with the Cochrane Handbook (Deeks 2022).

Subgroup analysis and investigation of heterogeneity

We performed a subgroup analysis, categorizing participants by disease (e.g., healthy population, hypercholesterolemia, heart failure, etc.) instead of VTE risk factors to reduce confusion, as different diseases have different VTE risks. This method provides clearer insights, since participants often have multiple risk factors but studies typically focus on one disease type. The findings may help clinicians assess the primary prevention effects of statins across diseases.

We also examined treatment durations (≤ one year, one to five years, > five years) to evaluate the cumulative effects of statins on VTE prevention, and analyzed varying doses of statins for differences in VTE prevention. Additionally, we investigated provoked status (unprovoked vs. provoked VTE) to identify any related prevention effects.

Sensitivity analyses

We incorporated the following post‐hoc sensitivity analyses: firstly, we excluded studies that did not report the incidence of VTE, and secondly, we omitted studies at overall high risk of bias. We added these two analyses to scrutinize the influence of unpublished data and the high risk of bias in included studies on the results of the meta‐analyses. We initially intended to conduct another post‐hoc sensitivity analysis excluding studies that included a subset of participants with previous VTE; however, since no such studies were included, we did not execute this analysis. We also refrained from conducting a sensitivity analysis with missing data imputed based on the worst‐case and best‐case scenarios, given that we based our analysis on the ITT method.

Trial sequence analysis

We included a section on TSA. Considering the incidence of VTE in this review was sparse and could yield type I errors, we employed TSA to determine if the evidence identified for this review was adequate.

Unused methods

We were not able to use all of our pre‐planned outcomes (Zhang 2021), namely those related to measures of treatment effects, as none of the outcomes we analyzed were continuous. Hence, this method was not applicable to this review. In future updates of this review, should studies utilize continuous scales of measurement to evaluate the effects of treatment, we will employ the mean difference (MD), or the standardized mean difference (SMD) if different scales have been utilized to measure the same outcome.

Contributions of authors

Zixin Wang: study selection, data extraction arbitrating, interpretation of results, review writing
Peng Zhang: protocol writing, interpretation of results, review writing
Jinhui Tian: conception of the review, design of the review, protocol writing, study selection arbitrating, data extraction, risk of bias assessment, assessment of the certainty in the body of evidence, interpretation of results, review writing
Peizhen Zhang: protocol writing, data extraction arbitrating, interpretation of results, review writing
Kehu Yang: conception of the review, design of the review, protocol writing, risk of bias assessment, assessment of the certainty in the body of evidence (arbitrating), interpretation of results, review writing
Lun Li: conception of the review, design of the review, co‐ordination of the review, protocol writing, study selection, data extraction, risk of bias assessment, data analysis, assessment of the certainty in the body of evidence, interpretation of results, review writing

Lun Li, Jinhui Tian and Kehu Yang are the guarantors for the review.

Sources of support

Internal sources

• The Changsha Natural Science Foundation, China

  The Changsha Natural Science Foundation (kq2208336) provided support for obtaining the studies. The funder had no role in the design, conduct, or publication of the review update.

• Scientific Research Launch Project for new employees of the Second Xiangya Hospital of Central South University (2022‐086), China

  Scientific Research Launch Project for new employees of the Second Xiangya Hospital of Central South University (2022‐086) paid Lun Li during this work.

• The Hunan Provincial Natural Science Foundation, China

  The Hunan Provincial Natural Science Foundation (2023JJ40831; 2023JJ60441) provided funds for language editing. The funder had no role in the design, conduct, or publication of the review update.

External sources

• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK

  Cochrane Vascular was supported by the Chief Scientist Office up to April 2023.

Declarations of interest

LL is a surgeon in breast cancer at the Second Xiangya Hospital, Central South University. He received funding from the Changsha Natural Science Foundation (kq2208336), the Hunan Provincial Natural Science Foundation (2023JJ40831; 2023JJ60441), and Scientific Research Launch Project for new employees of the Second Xiangya Hospital of Central South University (2022‐086). These funds supported this review, but the funders had no role in the design, conduct, or publication of the review update.
ZxW: none known
PZ is a surgeon at the Second Hospital of Nanyang City. 
JHT: none known
Pz‐Z: none known
KY: none known

New