Effectiveness of nurse-home visiting in improving child and maternal outcomes prenatally to age two years: a randomised controlled trial (British Columbia Healthy Connections Project)

Abstract

Background: We investigated the effectiveness of Nurse-Family Partnership (NFP), a prenatal-to-age-two-years home-visiting programme, in British Columbia (BC), Canada.

Methods: For this randomised controlled trial, we recruited participants from 26 public health settings who were: <25 years, nulliparous, <28 weeks gestation and experiencing socioeconomic disadvantage. We randomly allocated participants (one-to-one; computer-generated) to intervention (NFP plus existing services) or comparison (existing services) groups. Prespecified outcomes were prenatal substance exposure (reported previously); child injuries (primary), language, cognition and mental health (problem behaviour) by age two years; and subsequent pregnancies by 24 months postpartum. Research interviewers were masked. We used intention-to-treat analyses. (ClinicalTrials.gov, NCT01672060.) RESULTS: From 2013 to 2016 we enrolled 739 participants (368 NFP, 371 comparison) who had 737 children. Counts for child injury healthcare encounters [rate per 1,000 person-years or RPY] were similar for NFP (223 [RPY 316.17]) and comparison (223 [RPY 305.43]; rate difference 10.74, 95% CI -46.96, 68.44; rate ratio 1.03, 95% CI 0.78, 1.38). Maternal-reported language scores (mean, M [SD]) were statistically significantly higher for NFP (313.46 [195.96]) than comparison (282.77 [188.15]; mean difference [MD] 31.33, 95% CI 0.96, 61.71). Maternal-reported problem-behaviour scores (M [SD]) were statistically significantly lower for NFP (52.18 [9.19]) than comparison (54.42 [9.02]; MD -2.19, 95% CI -3.62, -0.75). Subsequent pregnancy counts were similar (NFP 115 [RPY 230.69] and comparison 117 [RPY 227.29]; rate difference 3.40, 95% CI -55.54, 62.34; hazard ratio 1.01, 95% CI 0.79, 1.29). We observed no unanticipated adverse events.

Conclusions: NFP did not reduce child injuries or subsequent maternal pregnancies but did improve maternal-reported child language and mental health (problem behaviour) at age two years. Follow-up of long-term outcomes is warranted given that further benefits may emerge across childhood and adolescence.

Keywords: Child injuries; child mental health; child problem behaviour; cognitive and language development; health inequities; nurse‐home visiting; public health; randomised controlled trial; subsequent pregnancies.