Randomised controlled trials (RCTs) that compared LMWH with placebo or an alternative medication (except anti-coagulants) in women with a viable pregnancy at treatment initiation were eligible for inclusion. Trials were required to enrol women with a history of either at least two early miscarriages or at least one late foetal loss and without evidence of antiphospholipid antibodies or therapeutic indication for anticoagulation treatment. A thorough evaluation for underlying cause of recurrent miscarriage was required (details specified in review). Trials were required to report live birth as an outcome.
Most included studies used enoxaparin at 20mg/day or 40mg/day; one study used dalteparin 5,000 IU plus aspirin 81mg/day. Three studies used aspirin at 100mg/day or 81mg/day; other studies used no intervention or placebo. Where reported, treatment was initiated at the eighth week, between six and 12 weeks of gestation or at ultrasound viability. Inclusion criteria for number or type of previous pregnancy loss experienced varied.
Two reviewers independently assessed the studies for inclusion in the review and resolved differences through consensus. A third reviewer confirmed the exclusion of randomised studies.