Effectiveness of structured lactation support and human donor milk banking in German NICUs: a stepped-wedge cluster-randomized trial

Abstract Background Although exclusive human milk feeding is beneficial for very low birth weight infants, rates remain suboptimal globally. This hybrid type 1 effectiveness-implementation trial evaluated structured lactation support combined with human donor milk banking in neonatal intensive care units (NICUs), primarily assessing effectiveness, with secondary assessment of implementation and economic outcomes. Methods We conducted a stepped-wedge cluster-randomized trial in 15 German Level I NICUs (April 2022–March 2024). The Neo-MILK intervention comprised structured lactation counselling, multilingual materials, mobile app support for mothers/parents, and human donor milk banking. Implementation strategies included video-based staff training, guidelines, commitment nudges for medical staff, and support for human donor milk bank establishment. The primary outcome was exclusive mother’s own milk feeding at NICU discharge. Secondary outcomes included feeding status categories, clinical complications, length of stay, and maternal outcomes. Implementation outcomes were acceptability, feasibility, appropriateness, fidelity, and sustainability. Prespecified subgroup analyses examined birth weight (< 1,000 g vs. ≥1,000 g) and illness severity (CRIB ≥ 11 vs. <11). Economic evaluation employed time-driven activity-based costing and budget impact analysis. Results Among 1,627 very low birth weight infants (831 intervention, 796 control), the intervention increased exclusive mother’s own milk feeding at discharge by 11%-points (95% CI: 3.0 to 19.0; p = 0.007) from a 43% baseline. Exclusive formula feeding decreased correspondingly. Subgroup analyses showed no significant effects in extremely low birth weight infants (< 1,000 g) or high-risk infants (CRIB ≥ 11). Clinical complications and length of stay showed no differences. Acceptability, appropriateness, and feasibility were high. Fidelity was sufficient, with prepartal counselling delivered as intended in 71% of cases. Human donor milk banks were established in 40% of sites. Sustainability indicators were promising, with 61% of staff reporting routine integration. Economic evaluation suggested potential net savings nationally in the base-case scenario. Conclusions Structured lactation support with human donor milk banking increased exclusive mother’s own milk feeding across Level I NICUs, with no observed increase in complications or costs. Implementation appeared feasible, with indications of sustainability. Variable uptake and differential effects in highest-risk subgroups highlight the need for ongoing organizational support and tailored strategies to further optimize effectiveness. Clinical trial registration German Clinical Trials Register DRKS00025058.