Wolff 1966.

Methods	Double‐blind placebo controlled trial conducted in ambulatory patients in USA
Participants	87 participants, mean age 50 years, range not reported. 89.6% patients were African‐American. Male 32%. Baseline mean SBP/DBP was 178/109 mmHg and pulse pressure was 69 mmHg. Inclusion criteria: male and female patients with a diastolic pressure of 100 mmHg or more (taken on three separate occasions at least one week apart) at some time during the recent course of their hypertension; patients with coronary artery disease and cerebrovascular disease Exclusion criteria: patients with a history of, or presently manifesting signs of malignant or accelerated severe hypertension, that is, patients showing very high blood pressures (diastolic pressures > 130 mmHg), retinal deterioration with hemorrhages, exudates and papilledema and evidence of impaired renal function (serum urea nitrogen levels > 60 mg per 100 mL), patients with chronic renal disease with serum urea nitrogen levels above 60 mg per 100 mL, patients with surgically correctable lesions of the adrenal (primary hyper‐aldosteronism, pheochromocytoma, Cushing’s disease) or kidney (renal arterial Iesions) Patients were followed for 2 years
Interventions	Treatment: reserpine 0.25 mg three times daily, chlorthiazide 0.5 g twice daily, or HCTZ 25 mg four times a day plus guanethidine if needed Control: placebo
Outcomes	Mortality, stroke, MI, CHF, systolic BP and diastolic BP Other outcomes: the appearance of the optic fundi, and biochemical tests of renal function, carbohydrate metabolism, serum uric acid and electrolytes
Notes	Treatment failure was decided by 2 physicians: Onset of, or significant increase in, congestive heart failure as evidenced by increasing dyspnea, edema, neck vein distension, hepatomegaly, gallop rhythm, or atrial fibrillation. Occurrence of cerebrovascular accident, encephalopathy, myocardial infarction, or onset of, or increasing symptoms of, angina pectoris or peripheral arterial disease. Appearance of hemorrhages, exudates, or papilledema on funduscopic examination. Increase in the serum urea nitrogen concentration by more than 20 mg per cent, confirmed by at least two determinations, with no other obvious pre‐renal or renal cause. Onset or increase in headaches not responding to aspirin up to 3.6 g daily. Onset of symptomatic diabetes mellitus or other symptoms or conditions which might be construed as serious complications or side effects of therapy.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were placed randomly on either hypotensive drug therapy (reserpine, thiazide, guanethedine) or matched placebos." "A table of random numbers was utilized to divide the patients accepted into the study into two groups (treated and placebo)." Comment: baseline characteristics similar in both groups (age, sex, weight, SBP and DBP).The incidence of past history of hypertension or diabetes, the duration of known hypertension, and the incidence of left ventricular hypertrophy similar in the two groups. The placebo group had more patients with evidence of coronary artery disease at the inception of the study, while more patients in the treatment group had a positive history of cerebrovascular disease and headache.
Allocation concealment (selection bias)	Unclear risk	Method used for allocation concealment was not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "The ‘prescribers’ took the blood pressures on each visit, were aware of the original random allocation of treatment or placebo, and were responsible for the titration of these patients with their medications. The second group of physicians (the ‘examiners’), unaware of the patients’ blood pressure and treatment, examined the patients at intervals ranging from 1 week to 5 months (routinely at 2‐month intervals)." "Placebo group. These patients received matched placebos which were the same shape and size and were administered at the same intervals as those medications received by their matched cases in the treatment group." Comment: participants and physicians were blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Cardiac, cerebral, retinal, renal, and general medical status were followed by observers unaware of the blood pressure readings or drug schedule." Comment: outcome assessors were blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "The cause of the absenteeism of ten patients in the treated group as opposed to one in the placebo group remains obscure." "A follow‐up of six of the treated absentees in their homes by visiting nurses indicated that all were delinquent for social rather than medical reasons. None admitted any increase in symptoms suggestive of treatment failure." "The incidence of complications sufficient to terminate the patients’ participation in the study was significantly higher in the placebo group than in the treated group." Comment: more participants were absent in the treatment group (10/45 = 22.2%) versus placebo group (1/42 = 2.4%).
Selective reporting (reporting bias)	Unclear risk	Protocol was not available to confirm what specific outcome measures were to be reported as primary or secondary outcomes. Comment: morbidity data and target organ function were provided in the results section along with the various reasons for treatment failure.
Other bias	Low risk	Quote: "This study was supported by NH1 Grant HE‐04788‐04. Recipient of the Career Scientist Award of the Health Research Council of the City of New York, under contract i‐342." "The population utilized in this study was primarily a lower‐income Negro group. Negro population may be significantly different from that observed in the white population."

BP: blood pressure; CHD: coronary heart disease; CHF: congestive heart failure; DBP: diastolic blood pressure; ECG: electrocardiogram; SBP: systolic blood pressure; TIA: transient ischemic attack.