Presented at the Neonatal Society 2007 Autumn Meeting.

Fielder H, Austin N, Heatheridge N (introduced by Dr RJ Thwaites)

Department of Neonatal Medicine, Christchurch Women’s Hospital, Christchurch, New Zealand

Background: Term and borderline premature infants contribute significantly to the workload of neonatal units, especially units without transitional care facilities. Tertiary neonatal units still transfer out or refuse admission to too many extremely premature infants1. Strategies to avoid unnecessary admissions of borderline premature infants could reduce this problem. These infants frequently require only a short period of special care and should ideally be with their mothers. Typically admission of babies of 35 and 36 weeks gestation from labour ward will be at the discretion of junior medical staff. We could find no evidence base to inform their decisions. We hypothesised that babies who spent a total of five days or less in hospital could be most appropriately cared for on the postnatal wards.

Aims of study: To identify perinatal factors which predict length of stay in borderline premature infants with a view to developing a scoring system to help decide which such infants to admit to the neonatal unit and which can be cared for on the postnatal wards.

Methods: We performed a retrospective case note review on all babies born between 35⁺⁰ and 36⁺⁶ weeks gestation from October 2005 to March 2006. Babies were then divided into 2 groups according to total days in hospital; 5 days or less and more than 5 days. Baseline characteristics were then analysed and compared.

Results: We identified 108 babies meeting the criteria. Of these 46 stayed 5 days or less and 62 stayed more than 5 days. There was no relationship between length of stay and gestational age or gender. Factors significantly associated with a shorter stay in hospital included; appropriate birth weight and length (defined by Z score), singleton pregnancy, spontaneous onset labour, cephalic presentation, vaginal delivery, Apgar ≥ 8 at 5 minutes, no need for resuscitation at birth and intent to bottle feed.

Conclusion: There are easily recognisable perinatal factors associated with shorter hospital stays in this group of babies. It should be possible to develop a simple scoring system to aid admission decisions. This would need prospective and local validation.

References
1. Cusack J Arch Dis Child Fetal Neonatal Ed 2007;92:F181-4