Composite Metrics to Assess Quality Improvement in Very Low Birth Weight Infants: 2010-2023

Quality improvement collaborations have documented a lack of sustained mortality and morbidity reductions in very low birth weight infants. We calculated mortality/morbidity/hospital stay metrics and major morbidity counts from an affiliated group of neonatal intensive care units (NICUs).

Sixteen NICUs provided Vermont Oxford Network data on births of 401 to 1500 g and/or at most 29 6/7 weeks from 2010 to 2023 to calculate composite mortality, risk-adjusted morbidity, and hospital stay metrics ascertaining proficiency (Benefit Metric) and efficiency (Value Metric).

In 11 795 infants, median age 28.7 (IQR, 26.4-30.6) weeks, the group Benefit Metric did not improve (P = .59). Only 1 NICU improved their Benefit Metric; 2 worsened. Six morbidity rates were unchanged. Chronic lung disease increased (P < .004). Morbidity counts increased with decreasing gestation. Overall, 37% of infants had at least 1 morbidity. Of infants aged 25 0/7 to 27 6/7 weeks, 57% had at least 1 morbidity, and 21% had at least 2. Of infants aged at most 24 6/7 weeks, 55% had at least 2 morbidities, and 23% had at least 3. Mortality excluding early deaths decreased only for infants born at 28 weeks or more (P < .01). All NICUs demonstrated increasing survivor total hospital length of stay (64-71 days, P < .0001). The group NICUs' Value Metric declined (P < .0001).

The Benefit Metric proficiency score from 16 affiliated NICUs showed no progress 2010 to 2023. The Value Metric efficiency score declined because survivor length of stay increased. To improve mortality, morbidity rates, and cost proxies of resource expenditure will require scalable potentially better practices, robust adherence measurements of evidence-based structures and processes, identification of influential culture/environment factors, extensive shared experience, and development of more comprehensive risk-adjustment methodologies that enhance accurate, meaningful NICU comparisons.