Context/Problem Admission of babies to NICU for medical care involves the separation of mothers and babies. We questioned what proportion of term babies were admitted in our institution and what interventions they required. We hypothesised that a significant proportion of babies would have minimal intervention on admission and a number would have potentially avoidable and treatable causes that would lend themselves to quality improvement interventions. We aimed to identify causes of potentially preventable admissions as a key performance area to target.

Assessment of problem and analysis of cause We retrospectively reviewed the NICU database to identify all babies ≥ 37 weeks gestation admitted to our tertiary level service over a 12-month period.

We classified term admissions as “expected” when NICU admission was anticipated following an antenatal diagnosis or “unexpected” where there were no concerns.

Study design Through Badgernet, demographic data were collected for all babies and the source of admission, diagnosis, interventions and length of stay (LoS) documented. For those infants in whom Badgernet data was incomplete or missing, retrospective case note analysis was undertaken.

Results There were 5843 babies delivered over the 12 month period of whom 4900 (83.9%) were ≥ 37 weeks. There were 453 (9.2%) term babies admitted to NICU; 65 (14%) of these admissions were expected and 389 (86%) unexpected.

We identified a number of babies who did not meet the recommended standard for achieving high-level neonatal care benchmarked against admission temperature and blood sugar level. Blood sugar level was documented in 174 babies (44.7%). Of these, 20 babies were identified as having a True Blood Glucose (TBG) <2 mmol. 25 babies had admission temperatures to the unit of <36.5°C, 10 with admission temperatures <36°C.

105 of 389 babies were discharged or transferred within 6 h of admission. 11/105 babies required medical intervention prior to transfer for cardiac or surgical management. The remaining 94 were discharged to the postnatal ward, 22 of 94 received IV antibiotics, no other interventions were required in the remaining 72.

72/389 (18.5%) of unexpected admissions at term required no medical interventions and were discharged to the postnatal ward within 6 h, representing 9.5% of all admissions to the unit.

Measurement of improvement Commonest reasons for admission in this group were mild respiratory distress, hypothermia and hypoglycaemia.

Following on from identifying these reversible causes we have implemented a quality improvement temperature bundle for use in labour ward, postnatal wards and NICU. This uses a visual cue to ensure both ambient temperature and infant temperatures are regularly checked. The introduction of this improvement bundle has enabled regular prospective audit of our temperature targets.

Conclusion and lessons learnt 9.2% of term infants were admitted to NICU. A significant proportion of “unexpected” admissions had a brief NICU stay and received minimal intervention. These infants who had minimal interventions represent a substantial share of the workload and admissions to NICU.

A concerning number of infants had hypothermia and hypoglycaemia on admission, a key area to target in implementation of quality improvement strategies.

Provision of simple supportive interventions in a Transitional Care Unit or observation area could potentially have reduced unexpected term admissions by 9.5%.