Giving parents support: a randomized trial of peer support for parents after NICU discharge

Peer support in the NICU setting has been associated with reduced parental stress, anxiety, and depression and is recommended as part of family-centered care. However, evidence on whether peer support provided after NICU discharge improves maternal mental health and infant outcomes is limited. NICU hospitalization is highly stressful for parents, with potential downstream effects on parenting practices and infant healthcare utilization. Maternal depression and anxiety are linked to infant feeding problems, lower maternal self-efficacy, delayed well-child care and immunizations, and increased ED use. Up to 30% of NICU parents experience persistent mental health concerns into the first year after birth, and many feel socially isolated post-discharge. The Giving Parents Support study examined whether structured peer support after NICU discharge improves parental mental health (self-efficacy, stress, anxiety, depression) and affects infant health outcomes (ED visits, hospitalizations, immunizations, development) during the 12 months following discharge.

Prior work suggests parent-to-parent support in the NICU can enhance parental competence and reduce maternal distress. For parents of children with chronic conditions, community-based peer support programs also show benefits. Maternal mental health problems are associated with adverse infant healthcare patterns, including reduced adherence to well-child care and immunizations and increased ED use and hospitalizations. Despite recommendations to integrate peer support into family-centered NICU care, evidence for benefits of post-discharge peer support on maternal mental health and infant outcomes remains sparse, motivating this randomized trial.

Design: Randomized controlled trial with intent-to-treat analysis. Setting: Children’s National Hospital Level IV NICU (freestanding children’s hospital without co-located delivery), Washington, DC, USA. Enrollment period: January 4, 2016 to February 24, 2017. Participants: Parents (≥18 years, English-speaking, planning to remain in DC metro area for 12 months) of infants with anticipated discharge within 2 weeks. Initial NICU length-of-stay ≥14 days criterion was broadened to include any LOS given the unit’s case mix. One parent per infant enrolled. IRB-approved; ClinicalTrials.gov registered (NCT02634372). Randomization: 1:1 to intervention or control after baseline surveys; stratified by infant birth weight (<1500 g vs ≥1500 g) with permuted blocks (sizes 2 or 4) for the <1500 g stratum; randomization schedule created in Stat/Ess (IIT College Station, TX). Investigators were blinded where possible. Intervention: Peer navigator program modeled on a primary care parent navigator initiative. Parents with prior NICU experience were hired as peer navigators to provide emotional support, resource connection, assistance with healthcare navigation, and empowerment. Training covered communication, boundaries, NICU terminology, and practical systems navigation (appointments, insurance, equipment/supplies, liaison with clinicians and services). Three navigators supported multiple families. Contact was attempted before discharge, then within 2 weeks post-discharge and monthly for 12 months via phone, email, or in person at appointments. The intervention’s overall effect (availability and outreach), not individual contact dose, was evaluated. Control: Care notebook alone. Measures: Parental outcomes assessed at baseline (in person) and at 1 week, 1 month, 3 months, and 12 months post-discharge (phone/online via REDCap). Instruments included Perceived Stress Scale-10 (PSS-10), Parental Stress Scale (PSS), Perceived Maternal Parenting Self-Efficacy (PMPS-E), Center for Epidemiologic Studies Depression Scale-10 (CES-D 10; with resource letters sent for CES-D ≥10), and State-Trait Anxiety Inventory, State form (STAI Y-1). Infant outcomes: parent-reported ED visits and hospitalizations over 12 months (analyzed as any vs none and counts); immunization status by 12 months for DTP, PCV13, and Hib (obtained from primary care or immunization registries); developmental status at 12 months assessed using the Ages and Stages Questionnaire, Third Edition (ASQ-3). Analysis: Intent-to-treat using generalized estimating equations (GEE) for longitudinal psychosocial outcomes, accounting for within-subject correlation and varying assessment timing; covariance structure chosen by lowest QIC. Group-by-time interactions modeled; predefined time points: 3 months (short-term), 6 months (intermediate), 12 months (long-term). Differences considered statistically significant if 95% CI did not include 0. Missing data handled by scale-specific rules when ≤10% items were missing; GEE used all available data. SAS 9.4 used for analyses.

- Enrollment and follow-up: 300 families randomized (150 intervention, 150 control); overall follow-up rate 88%; only 2% missing all follow-up data. Baseline demographics and patient characteristics were balanced between groups; baseline scale score differences were minor (0.1–1.2 points). - Parental mental health: All outcomes (stress, parental stress, anxiety, depression, self-efficacy) improved significantly over time in both groups (p < 0.0001). No significant differences in trajectories between intervention and control across 12 months. At specific time points (3, 6, 12 months), scores tended to be slightly lower (better) in controls but differences were not clinically meaningful. The only statistically significant between-group difference was lower parental stress (PSS) in controls at 3 months: mean difference −1.58 points (95% CI −3.15, −0.01). - Infant healthcare utilization and development: Full immunization by 12 months was 73.5% overall (control 74.8% vs intervention 72.1%; p = 0.60). Any hospitalization: 22.9% overall (control 21.8% vs intervention 24.1%; p = 0.63). ED use: 44% visited the ED at least once in the first year; proportions did not differ between groups (control 42% vs intervention 47%; p = 0.35). Median number of hospitalizations and ED visits was 0 in both groups. No differences in infant developmental outcomes at 12 months were observed.

Parental mental health improved over the first year post-discharge regardless of assignment, and the peer support intervention did not confer additional benefit on self-efficacy, stress, anxiety, or depression. Infant healthcare utilization and developmental outcomes were also unaffected. Potential reasons include: introduction of peer support very near discharge, limiting relationship-building time; universal enrollment rather than targeting parents with elevated needs; and variability in the amount and type of support delivered due to the flexible, family-driven model. High baseline self-efficacy and overall low rates of ED visits/hospitalizations may have limited detectable effects. The findings suggest that while peer support is valued and recommended in family-centered NICU care, general post-discharge peer support as delivered here may not change measured parental mental health or infant utilization outcomes. Tailoring timing, intensity, and matching of peer navigators to parent needs may be necessary to observe measurable effects.

In this randomized trial, adding a yearlong peer navigator support program beginning at NICU discharge to standard care did not improve parental mental health outcomes or infant healthcare utilization compared with standard care alone, although parental mental health improved over time in both groups. Future studies should evaluate earlier introduction of peer support (during hospitalization), identify and target parents most likely to benefit, and explore matching navigators to families on meaningful characteristics and optimizing intervention dose and standardization.

- Intervention heterogeneity: Amount and type of peer support varied by family; the intervention was not standardized, and individual dose-response could not be evaluated. - Timing: Introduction often occurred immediately before discharge, a busy period that may have limited relationship building and intervention uptake. - Targeting: All eligible parents were enrolled rather than focusing on those with elevated distress or specific needs, potentially diluting effects. - Outcome frequencies: Relatively low rates of ED visits and hospitalizations reduced power to detect differences in infant utilization. - Follow-up: Some loss to 12-month follow-up; those lost tended to be younger, less educated, single, and had higher baseline self-efficacy, which could introduce bias. - Measurement inconsistencies noted between sections for developmental assessment instruments, though overall no between-group differences were detected.