Hulley 1985.

Methods	Randomized, double‐blind, placebo‐controlled trial (parallel arms) ‐ a pilot study of the SHEP program‐isolated systolic hypertension in the elderly. Wash‐out period = 8 weeks. Multicenter, conducted in USA
Participants	SBP 160 to 219 mmHg. DBP < 90 mmHg. Age > 70 years: 61% of patients. Males 36%. Baseline BP was 172/75 in the treatment group and 174/77 in the control group. The study included elderly (60 years or older) patients with isolated systolic hypertension only (i.e. SBP 160 to 219 mmHg and DBP < 90 mmHg)
Interventions	Chlorthalidone 25 mg daily (N = 443) or placebo (N = 108) for first 4 weeks. A step‐up protocol was used wherein poor response (i.e. BP goal not reached) after 4 weeks led to a doubling of drug dosage from 1 to 2 capsules of chlorthalidone per d; poor response in patients receiving placebo led to a simulated randomization with a doubling of placebo capsules. If the goal BP was not reached after 12 additional weeks, patients on chlorthalidone were re‐randomized to step II drugs including reserpine, metoprolol, or hydralazine
Outcomes	Trough mean sitting SBP and DBP; serum biochemistry including potassium, uric acid, creatinine, glucose and cholesterol levels
Notes	A sample size calculation was provided based on a sample of 500 patients to detect a mean difference in SBP of 6 mmHg at 90% power. BP results were taken from the first 4 weeks of fixed monotherapy. The study authors stated that the baseline patient demographics, characteristics and medical history were "reasonably well distributed" across the 2 groups except for angina and carotid bruit which were statistically more common in the placebo group (P value < 0.05). BP data of the 1st 4 weeks was used from figure 1 on page 916 (SD not given). Change in serum potassium level given but no SD Additional publications: Smith WM et al. Drugs 1986; 31(Suppl 4):154‐64; Hulley SB et al. J Am Ger Society 1986 [data from first 3 months], SHEP. JAMA 1991; 265(24):3255‐64 and Kostis JB et al. JAMA 1997;278(3):2126.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Elderly men and women who met eligibility criteria at 4 baseline examinations were randomised in a double‐blind fashion to chlorthalidone or matching placebo." (line 1 under "Methods" p.914). "We used an adaptive randomisation procedure that varied treatment assignment probabilities by 10% in one or the other direction in order to balance the step I study groups within race, sex, age and baseline systolic BP strata." (line 15 from top of p.914, right column). No further information was given
Allocation concealment (selection bias)	Low risk	"Each randomisation was carried out by telephone between the clinic staff and the coordinating centre data manager, who checked that eligibility criteria were met before assigning the participant to chlorthalidone or placebo." (line 9 from top of p.914, right column)
Blinding (performance bias and detection bias) All outcomes	Low risk	"Elderly men and women who met eligibility criteria at 4 baseline examinations were randomised in a double‐blind fashion to chlorthalidone or matching placebo." (line 1 under "Methods" p.914). "Upon randomisation into the study, participants entered the step‐up protocol and received 25 mg/day of chlorthalidone or placebo (supplied as identical capsules by USV Pharmaceutical Corp.) (line 1 under "Methods‐Blood pressure treatment, p.914)
Incomplete outcome data (attrition bias) All outcomes	High risk	The analysis of efficacy was based on an intention‐to‐treat (ITT) technique Exclusions: 27,199 patients were screened; 2130 of these were included in the first baseline clinic visit. 1579/2130 (74%) of patients were excluded from the study during the 3 baseline clinic visits following screening, but prior to randomization Attrition: total number of withdrawals were not clearly documented WDAEs: not given. 7 patients receiving chlorthalidone (3 ‐ dizziness on standing, 1 ‐ syncope on standing, 1 ‐ rash, 1 ‐ hyperglycemia and 1 ‐ sexual dysfunction) and 2 patients receiving placebo (1 ‐ asthma and 1 ‐ escaping BP) had their medications terminated due to drug‐related events; the time point was not given, therefore these data were not useful (note that only the 1st 4 weeks out of the 12‐month trial were useable)
Selective reporting (reporting bias)	High risk	BP variability was not given. BP data from the placebo group were in graph form only (Fig. 1, p.916). Except for serum potassium levels, all other biochemical data were reported at baseline and 12 months only, not at 4 weeks (therefore not useable). Mortalities, SAEs and total AEs were not reported. Only more severe AEs characterized as "troublesome" or "intolerable" were reported in the study (AEs determined to be "not troublesome" were not mentioned)
Industry sponsorship	Low risk	Supported by grants from the National Heart, Lung and Blood Institute, the National Institute of Aging and the National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland