Effectiveness of a large-scale home visiting programme (PIM) on early child development in Brazil: quasi-experimental study nested in a birth cohort.

Background: A large-scale parenting programme with weekly home visits (Primeira Infância Melhor (PIM)) has been implemented in the south of Brazil for nearly two decades, but lacks evaluation of its effects on early childhood development (ECD). This quasi-experimental study aimed to assess the effects of PIM in real-life settings within a population-based birth cohort study.
Methods: Data from the 2015 Pelotas Birth Cohort Study and the state programme information system were linked to identify study children who received PIM. Propensity score matching was used to create a comparable control group (using one-to-one matching) to estimate the effect of PIM on ECD measured at age 4 years. First, the impact of any enrolment in PIM was evaluated; then the intervention group was stratified according to whether enrolment occurred during pregnancy or after birth. Double adjustment was applied in linear regression to analyse child development scores, and Poisson regression for delayed development (below the 10th percentile of whole cohort). Effect modification due to family income was explored.
Results: There was no evidence that any enrolment in PIM (601 pairs) by age 4 years was associated with child development. However, PIM starting during pregnancy (estimated for 121 pairs) predicted higher development scores (0.27 SD; 95% CI 0.02 to 0.52), and 71% lower prevalence of delayed development (prevalence ratio=0.29; 95% CI 0.13 to 0.69), compared with the control group. There was strong statistical evidence (p=0.003, test of interaction) that the effect of PIM starting in pregnancy was larger than when starting after birth (480 pairs). The effect of PIM starting during pregnancy was found to be restricted to poorer families.
Conclusions: In a real-life setting, PIM was effective only when starting during pregnancy and only for poorer families. A higher-quality programme might be more effective with a broader population.
Competing Interests: Competing interests: None declared.
(© Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY. Published by BMJ.)