Hardy 1998.

Methods	Location and time frame: Instituto Materno‐Infantil de Pernambuco (IMIP) in city of Recife (Northeast Brazil); early 1990s. Design: 'Semi‐experimental operational study' with different time series for treatment and control. Sample size calculation (and outcome of focus): no information
Participants	General with N: 698 women who gave birth at the institute (IMIP). Source: Teaching institution providing pediatric and obstetric care to low‐income women and children. Inclusion criteria: Treatment group comprised of first 348 women who delivered after intervention started, returned for 40‐day appointment, and volunteered for study. Control group gave birth at IMIP, received traditional counseling (prior to new intervention), and returned to hospital for regular check‐ups (1 year postpartum). Exclusion criteria: had stillbirth or twin pregnancy, would have difficulty breastfeeding, had surgical sterilization approved, or desired pregnancy soon.
Interventions	Study focus: Introduce lactational amenorrhea method (LAM) to postpartum care Treatment: Provided during family planning (FP) counseling of women at prenatal care or those giving birth at IMIP. LAM introduced as contraceptive method during prenatal visits, in maternity hospital after birth, and on 8th‐day postpartum visit. Comparison or control: Traditional FP counseling prior to development of new intervention (did not include LAM); provided during prenatal care, hospitalization, and day 8 postpartum visit.
Outcomes	Primary: contraceptive use by specific method (we also analyzed as modern or traditional (WHO 2013)); pregnancy (presumably self report; no mention of testing or use of records). Secondary: not applicable Assessment times: Treatment group at visits 3, 6, 9, and 12 months postpartum; Comparison group interviewed at one year postpartum only.
Notes	Total Ns for table on contraceptive use were reversed for experimental and control groups (215 versus 350). All other results indicate experimental group had 215 women at 12 months. Control group consisted of 350 women interviewed at 12 months.
Risk of bias
Bias	Authors' judgement	Support for judgement
Exposed cohort representativeness	High risk	Gave birth at institution serving low‐income women and returned for 40‐day appointment.
Nonexposed cohort selection	High risk	Returned to hospital for regular check‐up at one‐year postpartum.
Exposure ascertainment: method used	High risk	Control: unclear whether records or recall of counseling received. Experimental: presumably program records of clinic encounters.
Comparability of groups: design or analysis	Low risk	Potential confounders in logistic regression: age, years of schooling, # living children, marital status.
Outcome assessment: method used	Unclear risk	Interview at clinic visits; self report
Follow‐up length	Low risk	12 months
Follow‐up adequacy	High risk	Loss to follow up: experimental group at 12 months, 133/348 (38%); control interviewed at 12 months.