Bassaw 1998.

Methods	Randomised clinical trial. Participants were alternately allocated to either the supplemented or to the control groups in order to match for age, parity, ethnic group and BMI. Data from the 'control' group were not used in this analysis. Randomisation was conducted by the pharmacist using a table of random numbers, and supplements were distributed to the participants in sealed envelopes. Clinicians were unaware whether the participants were in the supplemented or control groups, and which supplementation was administered.
Participants	Women attending a hospital in Trinidad between 1992 and 1995. Pregnant women recruited into the study before 20 weeks’ gestation primigravidae, or multigravidae with obstetric history of pre‐eclampsia. Participants were normotensive and urinalysis was negative for albuminuria. None had any underlying medical disorders such as chronic hypertension, renal disease, diabetes mellitus and collagen vascular disorders.
Interventions	2 calcium tablets (1200 mg elemental calcium), a combination of 1 calcium tablet and 1 baby Cafenol (80 mg aspirin) or 1 baby Cafenol daily. All participants, including the controls, received the routine haematinics which were ferrous sulphate (200 mg) and folic acid (5 mg) daily. There were 114 primigravidae amongst the controls. Of the supplemented groups, 45 primigravidae received aspirin, 36 had calcium and aspirin, and calcium tablets were administered to 42 primigravidae. All of these women were less than 24 years of age. For this review we have used only data for calcium (600 mg) and Cafenol (80 mg aspirin) vs Cafenol daily (80 mg aspirin).
Outcomes	DBP (measured by the same observer with the participants in a sitting position, recorded at the onset of muffing ‐phase 4 Korotkoff sound), PIH (BP ≥ 140/90 mmHg), pre‐eclampsia (hypertension plus proteinuria).
Notes	8 participants were unavailable for analysis. Source of funding: not stated. CoI: not stated.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The supplement vs control group were allocated by alternation, but it was clear that various supplemented groups were randomised by the pharmacist using random number tables. In this review we use only data for calcium plus aspirin vs aspirin, which were randomised.
Allocation concealment (selection bias)	Low risk	Supplements were distributed in sealed envelopes.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Clinicians were unaware whether the participants were in the supplemented or control groups, and which supplementation was administered. Participants were not blinded as placebos were not used.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Reported that clinicians recording outcomes were unaware of treatment group
Incomplete outcome data (attrition bias) All outcomes	High risk	8 participants were unavailable for analysis.
Selective reporting (reporting bias)	Low risk	Expected outcomes reported
Other bias	Low risk	None noted