Benz 1998.
| Methods | Randomized, double‐blind, placebo‐controlled trial (parallel arms). Wash‐out period = 2 to 4 weeks; conducted in USA | |
| Participants | DBP 95 to 115 mmHg. Mean age 52 years. Males 57%. Baseline BP was 152.8/101.5 mmHg in the treatment group and 152.7/101.4 mmHg in the control group. Pulse pressure = 51.3 | |
| Interventions | Valsartan 80 mg (N = 99) or 160 mg/d (N = 99), valsartan 80 mg or 160 mg/d + HCTZ 12.5 mg or 25 mg/d (all combined, N = 379), HCTZ 12.5 mg (N = 100) or 25 mg/d (N = 100), or placebo (N = 94) Treatment duration = 8 weeks | |
| Outcomes | Change from the placebo in trough mean sitting SBP and DBP; pulse rate, body weight, ECG, serum chemistry, hematology and urinalysis. Withdrawals due to adverse events were not given in each treatment group | |
| Notes | A sample size calculation was provided based on 85 patients per treatment group to detect a difference in mean sitting DBP of 4 mmHg (standard deviation = ± 8 mmHg) between treatments at a power of 90%. Baseline patient demographics, measurements and medical history were similar across all treatment groups (P value = NS). Standard deviation (SD) of change in BP not given (95% CI given) | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | "The study was ... randomized..." (line 2 under "Study Design" p.862). "Eligible patients were randomized into one of nine double‐blind treatment groups..." (line 10 under "Study Design" p.862). Technique for sequence generation not stated explicitly by study authors |
| Allocation concealment (selection bias) | Unclear risk | Not stated by study authors |
| Blinding (performance bias and detection bias) All outcomes | Low risk | "....a randomized, double‐blind, multiple dose, placebo controlled, multifactorial, parallel trial" (line 2 under "Study Design" p.862). "Study drugs were packaged in double‐dummy fashion to maintain blinding, with each patient taking two capsules per day at 8 am." (line 3 from bottom under "Study Design" p.862). Both low and high doses of HCTZ and placebo were supplied as identical capsules |
| Incomplete outcome data (attrition bias) All outcomes | High risk | "The primary efficacy analysis was an ITT analysis..." (line 7 under "Statistical analysis" p.862). Exclusions: number of patients excluded from study after the single‐blind placebo run‐in period before randomization was not reported Attrition: there were 79/871 (9%) withdrawals due to: 41 ‐ AEs, 9 ‐ unsatisfactory therapeutic effect, 7 ‐ did not meet protocol criteria, 2 ‐ non‐compliance, 15 ‐ withdrew consent, 5 ‐ lost to follow‐up WDAEs: 41/871 (4.7%) of patients withdrew due to an adverse event (description of AE not given). Note that data from each treatment group were pooled and therefore it could not be ascertained how many patients receiving HCTZ monotherapy or placebo were withdrawn |
| Selective reporting (reporting bias) | High risk | Baseline age, height and weight were measured but not reported. Only differences in change in mean sitting DBP or SBP (HCTZ ‐ placebo) were shown at endpoint, not the actual mean change for each individual treatment group (see Table 2, p.864). Standard deviation of change in BP not given. Study authors did not comment on mortality. Serious adverse events: 1 patient from the HCTZ 12.5 mg group, 2 from the HCTZ 25 mg group and 1 from placebo group (reasons not given). Adverse events regardless of relationship to study drug were reported for all treatment groups combined only (464/867), not for individual groups; whereas, drug‐related AEs were reported for each treatment group (PLB = 17/93, HCTZ 12.5 = 23/100, HCTZ 25 = 18/100) |
| Industry sponsorship | High risk | Sponsored by Novartis |