Highlight
• In a nationwide cohort of 39,413 infants born at 22–31 weeks’ gestation and discharged home, 67% had at least one hospitalization before age 2.
• Hospitalization frequency and total days hospitalized rose with decreasing gestational age (58.8% at 31 weeks to 87.0% at <24 weeks) and with the presence and number of neonatal morbidities.
• Bronchopulmonary dysplasia (BPD), severe necrotizing enterocolitis (NEC), and neonatal brain injury were each independently associated with 1.5–1.9-fold increases in expected total hospitalization days.
Background
Infants born very preterm (<32 weeks’ gestation) represent a group with elevated short- and long-term healthcare needs. Advances in perinatal and neonatal care have increased survival but have also accentuated the relevance of nonfatal neonatal morbidities—notably bronchopulmonary dysplasia (BPD), necrotizing enterocolitis (NEC), and neonatal brain injury—that influence later health. Quantifying post‑discharge hospital resource use in contemporary, population-level datasets is essential to counsel families, plan community and hospital services, and identify targets for preventive strategies.
Study design
This retrospective population-based cohort study used linked data from the National Neonatal Research Database (NNRD) and national hospital admission records (Hospital Episode Statistics for England and the Patient Episode Database for Wales). The cohort comprised infants born at 22+0 to 31+6 weeks’ gestation between January 1, 2013, and December 31, 2018, who were admitted to neonatal units and subsequently discharged home. The study followed admissions from neonatal discharge up to 2 years corrected age. Analyses adjusted for gestational age, sex, small-for-gestational‑age status, and season at neonatal discharge.
Primary outcomes were: (1) prevalence of ≥1 hospital admission after neonatal discharge and before age 2, and (2) total calendar days hospitalized across all admissions. Neonatal morbidities of interest included BPD, severe NEC, and neonatal brain injury. The investigators used negative binomial regression to estimate adjusted incidence rate ratios (AIRRs) for total hospitalization days.
Key findings
Population: 39,413 infants were included (median gestational age 29 weeks, IQR 27–31; 54.2% male). Two-thirds (66.7%) had no major neonatal morbidity recorded, whereas the remainder had one or more of the predefined morbidities.
Overall hospitalization frequency and burden
Two-thirds (26,498; 67.2%) of cohort infants experienced at least one hospital admission between neonatal discharge and their second birthday. The proportion hospitalized varied strongly by gestational age at birth:
- 31 weeks: 6138/10,444 (58.8%)
- <24 weeks: 450/517 (87.0%)
Median total calendar days hospitalized across all admissions also increased with lower gestation: 1 day (IQR 0–5) for infants born at 31 weeks versus 8 days (IQR 3–21) for those born at <24 weeks.
Impact of specific neonatal morbidities
Neonatal morbidities were independently associated with substantially greater hospitalization burden. Adjusted incidence rate ratios (AIRRs) for total days hospitalized were:
- BPD: AIRR 1.80 (95% CI, 1.72–1.88)
- Severe NEC: AIRR 1.88 (95% CI, 1.65–2.15)
- Neonatal brain injury: AIRR 1.46 (95% CI, 1.36–1.57)
The cumulative effect of multiple morbidities was notable. For example, the regression model estimated a child born at <24 weeks who had all three morbidities would average 40.6 days (95% CI, 34.8–44.3) in hospital before age 2.
Dose–response with number of morbidities
There was a graded relationship between number of neonatal morbidities and hospitalization days: infants with more morbidities accumulated disproportionately more inpatient days. This pattern persisted after adjusting for gestational age and other covariates, indicating independent contributions from the morbidities themselves.
Implications for subgroups and service planning
The high prevalence of post‑discharge hospital use — especially in extremely preterm infants and those with neonatal complications — has implications for transitional care programs, community follow-up capacity, and family support services. Hospital day estimates provide concrete metrics for workforce and bed‑capacity planning, and for cost‑projection models.
Expert commentary
This large, contemporary, population-level study strengthens the evidence that very preterm infants remain a high-utilization group after neonatal discharge, with neonatal morbidities exerting large, additive effects on inpatient resource use. For clinicians counselling families at discharge, the findings provide empiric probabilities to frame expectations: the majority of discharged very preterm infants will require at least one hospital readmission within two years, and risks are highest with lower gestation and with BPD, NEC, and brain injury.
From a mechanistic perspective, the association of BPD with increased hospital days is biologically plausible given ongoing respiratory vulnerability (infant susceptibility to viral lower respiratory infection, recurrent wheeze, need for supplemental oxygen or ventilatory support). NEC and surgical complications predispose to feeding problems, growth faltering, and recurrent sepsis, all of which can drive repeat admissions. Neonatal brain injury increases neurodevelopmental disability and associated comorbidities that may need inpatient management.
Strengths
- Nationwide, near‑complete capture of neonatal admissions with linkage to hospital admission data yields robust population estimates.
- Large sample size permits stratification by gestational age and morbidity combinations with narrow confidence intervals.
- Use of total calendar days as an outcome complements simple readmission counts by reflecting illness severity and resource intensity.
Limitations
- Retrospective linkage of administrative datasets can introduce misclassification (coding errors) and lacks detailed clinical nuance available in prospective cohorts.
- Hospital Episode Statistics and PEDW capture admitted episodes only; emergency department visits not resulting in admission and outpatient health service use were not included.
- Reasons for readmission were not exhaustively detailed in the summary results; targeted analyses of diagnosis-specific drivers (e.g., bronchiolitis, surgical complications, seizures) would refine actionable prevention strategies.
- Findings reflect the England and Wales healthcare context and may not directly generalize to other health systems with different neonatal or community care structures.
Clinical and policy implications
For clinicians and neonatal services, the study offers several practical takeaways:
- Individualized discharge counselling should include clear, data‑driven discussion of the likelihood of readmission and expected hospital days, especially for families of extremely preterm infants and infants with BPD, NEC, or brain injury.
- Enhanced post‑discharge supports — timely neonatal outpatient follow-up, rapid access respiratory clinics, home oxygen services, and multidisciplinary neurodevelopmental follow-up — may mitigate some admissions or enable earlier, community‑based management.
- Preventive strategies targeted during the neonatal admission (e.g., protocols to minimize BPD and NEC, standardized feeding and infection‑prevention bundles, neuroprotective care) remain central to reducing long‑term hospitalization burden.
- At health‑system level, bed‑capacity planning, community nursing workforce estimates, and cost projections should account for the high proportion of infants needing readmission and the strong influence of neonatal morbidities.
Future research directions
Key priorities include:
- Cause-specific analyses to determine primary drivers of readmission (respiratory infections vs feeding issues vs surgical complications) and when these occur in the post‑discharge course.
- Intervention trials or implementation studies testing whether structured transitional care (early home visits, telemedicine follow-up, caregiver education, rapid access clinics) reduce admissions or total inpatient days.
- Health economic evaluations that combine these hospitalization metrics with cost data to quantify potential savings from prevention and enhanced community programs.
- Equity analyses to evaluate how social determinants (socioeconomic status, access to primary care, family resources) modify readmission risk and to target support to high‑need groups.
Conclusion
This nationwide cohort study from England and Wales demonstrates that most infants born very preterm and discharged from neonatal care will experience at least one hospital admission before their second birthday, with lower gestational age and neonatal morbidities markedly increasing total inpatient days. The results provide empiric benchmarks to support family counselling and service planning and emphasize the need for targeted preventive and post‑discharge strategies to reduce hospital burden in this vulnerable population.
Funding and clinicaltrials.gov
The primary paper reports funding and disclosures (see citation). This analysis used routinely collected NHS data and did not report an interventional trial registration number.
References
1. van Hasselt TJ, Dorner RA, Katheria A, Battersby C, Gale C, Lo DKH, Seaton SE; UK Neonatal Collaborative. Neonatal Morbidities and Hospitalization in the First 2 Years of Life Among Infants Born Very Preterm. JAMA Netw Open. 2025 Sep 2;8(9):e2530123. doi: 10.1001/jamanetworkopen.2025.30123 . PMID: 40900591 ; PMCID: PMC12409584 .