Sebastian 2012.

Methods	Location and time frame: rural areas of Meerut district in Uttar Pradesh (India); intervention from September 2006 to January 2007. Design: Evaluation study; described as 'randomized experimental pre‐ and post‐test design.' District had 12 blocks; randomly assigned 2 blocks to intervention and 2 to comparison. Selected 12 villages with population >= 2000 from each block for inclusion. Sample size calculation: Contraceptive prevalence rate of 10% among women aged <= 24 taken as base value. Needed 541 women in each arm to detect 5% increase in contraceptive use at 9 months postpartum. Estimated 600 to compensate for 10% loss to follow up.
Participants	General with N: 959 women Source: Study villages Inclusion criteria: Had 0 or 1 child, < 25 years old, were 4 to 7 months pregnant. Exclusion criteria: first trimester, reportedly because they rarely mention pregnancy to community workers or go for antenatal care.
Interventions	Study focus: Increase knowledge and use of lactational amenorrhea method (LAM) and postpartum contraception through counseling. Treatment: Educational campaign for enrolled pregnant women and mother‐in‐law or oldest female family member. Included healthy timing and spacing of pregnancy, LAM, postpartum contraception; provided in home. According to investigator, community workers continued with messages postpartum, during routine monthly home visits. Also contacted mothers‐in‐law during antenatal and postpartum periods about postpartum contraception. Intervention included an educational campaign for males in community about maternity care and postpartum contraception. Comparison or control: Government‐run health program (antenatal clinic; home‐based counseling of pregnant women and family about antenatal and postnatal care and family planning; preschool program workers also counseled pregnant and lactating women at home.
Outcomes	Primary: modern contraceptive use; pregnancy (self report) Secondary: change in knowledge of healthy timing and spacing of pregnancy (>= 2 messages); change in knowledge of >=2 spacing methods available in national program. Assessment times: 9 months postpartum; knowledge also at 4 months postpartum.
Notes	Investigator provided additional information about the intervention, including time frame for postpartum contact.
Risk of bias
Bias	Authors' judgement	Support for judgement
Exposed cohort representativeness	High risk	Low‐parity women (0 or 1 child), < 25years old, living in selected district, block, and village. In contrast, state fertility rate was 3.8.
Nonexposed cohort selection	Unclear risk	From same district but different block than that of the exposed (treatment) group.
Exposure ascertainment: method used	Low risk	Prenatal: presumably from clinic records of services provided. Postpartum: presumably from community workers' records of home visits and services provided.
Comparability of groups: design or analysis	Low risk	Potential confounders included in logistic regression models: study group, education, age, age at initiation of cohabitation, caste, parity. Analysis does not appear to account for assignment of groups. With 2 blocks per study group, analysis methods are limited.
Outcome assessment: method used	Unclear risk	Interviews conducted during home visits; self‐reported outcomes.
Follow‐up length	Low risk	9 months
Follow‐up adequacy	High risk	Total loss: 20% (238/1197 did not complete all 3 interview rounds)