Visscher 1972.

Methods	Randomised control trial.
Participants	57 eligible Rh‐negative women who had had spontaneous miscarriage at gestation 8‐24 weeks consented to participate in the study. 9 dropped out pre intervention and no reason was stated. 48 participants randomly received injection of coded ampules within 72 hours of spontaneous miscarriage. 19/57 treatment group: 14/19 had dilatation & curettage (therapeutic dilatation & curettage), 5/19 had spontaneous miscarriage. 29/57 control group: 25/29 had dilatation & curettage (therapeutic dilatation & curettage), 4/29 had spontaneous miscarriage.
Interventions	Intervention: IM 300 µg Rhesus anti‐D immunoglobulin in a coded ampule. Control: IM 1 mL of homogenous gamma globulin placebo in a coded ampule. The coded ampoules were randomly assigned to the eligible participants and both the participants and the clinicians were blinded. Blood was collected after 6 months and the code was revealed. About 19/57 participants received 300 µg of Rhogam and 29/57 participants received 1 mL of homologous gamma globulin placebo.
Outcomes	At 6 months follow‐up, in the treatment group all 19 were non‐sensitised as evidenced by indirect Coombs' test. In subsequent follow‐up, 3 Rh +ve non‐sensitised pregnancies were observed. On 6 months follow‐up in control group, all 29 were non‐sensitised evidenced by indirect Coombs' test. In subsequent follow‐up, 6 Rh +ve non‐sensitised pregnancies were observed.
Notes	ABO incompatibility and parity were not mentioned.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Double blind study with coded ampules. The ampules were randomly assigned to the eligible participants. However, it is not clear how the sequence of the code were generated and how randomisation of the participants was done.
Allocation concealment (selection bias)	Low risk	The ampules containing 300 µg Rh immunoglobulin and gamma globulin placebo were coded and all ampules were homogenous.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	The clinician and the participants were both blinded from pre intervention up to 6 months post intervention when the codes were broken.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Serology tests were done for both groups to detect atypical blood group antibody by indirect Coombs test and the code was revealed. Subsequently onwards, having known who received anti‐D or placebo, the outcome of subsequent Rh‐positive pregnancy was anticipated.
Incomplete outcome data (attrition bias) All outcomes	High risk	All participants were followed up post‐intervention. Of 18/57 subsequent pregnancies, 3/19 from the treatment group and 6/29 from the control group reported Rh‐positive pregnancies.
Selective reporting (reporting bias)	High risk	It is not known whether or not the other 39/57 participants ever got pregnant with a Rh‐postive pregnancy and what was the outcome after the follow‐up period was over. While it is clear that the recruitment was terminated at 2 years, it is not clear how long each of the participants from both groups were followed up in order to see the outcome of the subsequent Rh‐positive pregnancy.
Other bias	Unclear risk	Not noted.

IM: intramuscular