PATS 1996.

Methods	Randomized, double‐blind placebo controlled trial conducted in China from 44 clinical centres.
Participants	5665 Chinese men (72%) and women (28%) with a history of transient ischemic attack, minor stroke or major stroke without severe disability. Mean age + SD was 60 + 8 years. Baseline BP 154/93 mmHg. 16 % of patients were not hypertensive BP <140/90 mm Hg. Exclusion criteria: malignant neoplasm, rheumatic valvular disease, congestive cardiomyopathy, permanent atrial fibrillation, secondary hypertension, hyperthyroidism, severe hepatic or renal disease, haemorrhagic disease, insulin dependent diabetes mellitus etc. Follow‐up: 2 years
Interventions	Treatment: Indapamide 2.5 mg daily Control: placebo
Outcomes	Mortality, stroke, coronary heart disease, blood pressure
Notes	Secondary prevention trial.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Conducted as a double‐blind placebo‐controlled multicentre trial in China." "After a single‐blind, run‐in phase on placebo, eligible patients were randomized to indapamide treatment or to placebo." Comment: method used for the random selection of participants was not mentioned. Baseline characteristics were balanced at study entry.
Allocation concealment (selection bias)	Low risk	Quote: "The sealed envelope system was used to randomize the participants. Investigators in every clinical centre assigned all the eligible patients to either the indapamide treatment group or the placebo group according to the order of the sealed envelopes supplied by the Coordinating Office. Patients would enter into the double‐blind period on the date of randomization." "The hypotensive treatment protocol was fixed, i.e. a tablet of indapamide (2.5 mg) per day in treatment group and a pill of matching placebo per day in the placebo group." Comment: sealed envelopes whether transparent or opaque used for the allocation was not mentioned. Baseline characteristics were balanced at study entry.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "Conducted as a double‐blind placebo‐controlled multicentre trial in China." "A physician might withdraw a patient from the double‐blind treatment when he thought the elevated or lowered blood pressure was harmful to the patient. During the double‐blind phase, if the double‐blind treatment was harmful to the participants because of the elevation or lowering of blood pressure to an intolerable level, treatment can be modified." Comment: participants of the study were blinded but it is unclear from the statement that physicians could withdrew the patients from the treatment or placebo group depending on the patient's medical condition.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Outcome assessors were not blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	162 (5.7%) in treatment group and 150 (5.3%) in placebo group defected from the trial for non‐medical reasons. All patients who left DB period alive were followed to allow ITT analysis of mortality and morbidity. These patients were examined once a year. The loss to follow‐up was not mentioned. The drug withdrawal was more in the placebo group.
Selective reporting (reporting bias)	Low risk	Comment: all outcomes (primary and secondary as well as fatal and non‐fatal events) as mentioned in the objectives analyzed and reported in the results section.
Other bias	Low risk	This study was supported financially by the World bank Office, Ministry of Public Health, China and Clinical Trial service Unit of Oxford University, U.K.