Barraclough 1973.
Methods | Single‐blind, placebo‐controlled trial conducted in Cardiff and London UK. | |
Participants | 116 ambulatory patients, ethnicity was not reported, mean age 55 years, range (45 to 69 years). Male (50%). Baseline mean DBP was 110 mmHg; SBP and pulse pressure were not reported. Inclusion criteria: Men and women between 45 and 69 years with two casual, sitting diastolic blood pressures of between 100 and 120 mmHg on each of two occasions, separated by an interval of at least two weeks were included. Exclusion criteria: Patients were excluded if: (a) there was evidence of renal or cardiac failure or papilloedema; (b) there was a history of cerebrovascular accident or myocardial infarct within the preceding three months; (c) any serious or potentially fatal disease or disability was present that would prevent regular attendances or which contraindicated hypotensive therapy; (d) they were currently receiving antihypertensive therapy; or (e) there was evidence that hypertension was secondary to a surgically remediable condition. Follow‐up: 1.5 years |
|
Interventions | Treatment: bendrofluazide (93%), methlydopa, and debrisoquine Control: placebo (received calcium lactate tablets) |
|
Outcomes | Mortality, CHD, stroke, CHF, and diastolic BP | |
Notes | Doses were not specified, assumed to be high dose bendrofluazide. K supplement was given automatically. | |
Risk of bias | ||
Bias | Authors' judgement | Support for judgement |
Random sequence generation (selection bias) | High risk | Quote: "Random samples of the general population and hospital patients." "The patients in the control and treatment groups were compared for age, weight, the levels of blood glucose, and blood urea at the time of entry to the trial (table I)." Comment: simple random sampling was done. Two groups were comparable at the outset, but within 18 months this comparability had disappeared. |
Allocation concealment (selection bias) | Unclear risk | Quote: "The patients were allocated at random to either the control or treatment group. The series was balanced for age and sex after every 10 allocations." Comment: method of allocation concealment was not specified. |
Blinding of participants and personnel (performance bias) All outcomes | High risk | Quote: "Those in the treatment group were treated with any combination of bendrofluazide with potassium supplement, methyldopa, or debrisoquine, the choice of treatment being at the discretion of the physician. The physicians knew which treatment was given." "Progression from one regimen to the next depended on the blood pressure response and incidence of side effects. If the diastolic blood pressure rose to 130 mmHg or over in a patient in the control group the patient was immediately withdrawn from the trial and given hypotensive treatment." Comment: although stated as single‐blind, neither the physicians nor the participants were blinded. |
Blinding of outcome assessment (detection bias) All outcomes | High risk | Comment: Information on the outcome assessors was not reported in the study. Probably, blinding was not carried out. |
Incomplete outcome data (attrition bias) All outcomes | High risk | Quote: "Total of 42 of the 58 (72.4%) randomised patients left the trial. Six patients in control group were withdrawn from the trial as DBP > 130 mmHg. Seventeen patients left for medical reasons 14 in control group and 3 in treatment group. Nineteen left for non‐medical reasons." "Medical indications in control group for leaving trial were diastolic pressure greater than 130 mmHg, myocardial infarction, pulmonary embolus, cardiac septum infarct, and cardiac failure; while indications for treatment group were senility, fractured skull, and myocardial infarction." Comment: attrition rate was very high and unequal across groups. How data were collected or analyzed in patients who withdrew is not explained or reported. By the end of one year, the two groups lost comparability. Statistical tests used for analysis were not mentioned. |
Selective reporting (reporting bias) | High risk | Comment: Mortality and morbidity data was not clearly stated and list of side effects were not reported. It is not clear whether data from patients who withdrew were collected or reported until end of study. |
Other bias | Unclear risk | Conflict of interest of authors was not reported. |