Hospital admissions and gestational age at birth: 18 years of follow up in Western Australia.

Background: Infants born moderate to late preterm are twice as likely to be rehospitalised within the first few weeks following discharge from the birth admission. It is not understood how rehospitalisation risk changes with age or how risks have changed over time.

Methods: A retrospective birth cohort study of all live, singleton births in Western Australia 1 January 1980-31 December 2010, without congenital anomalies, followed to 18 years of age. Rehospitalisation rates for gestational age categories (<28, 28-31, 32-33, 34-36, 37-38 and ≥42 weeks) were compared with term births (39-41 weeks) using negative binomial regression. To assess whether rehospitalisation risk changed with age or over time, analyses were conducted for different age intervals and for 5-year birth cohorts.

Results: Rehospitalisation rates were higher up to 18 years for all preterm and early term categories including early term (37-38 weeks) [130.2/1000 person-years at risk (pyr); 95% confidence interval 129.1, 131.4]; late preterm (34-36 weeks) (164.2/1000 pyr; 161.1, 167.4), and post-term (≥42 weeks) (115.3/1000 pyr; 111.7, 119.0) compared with term births (109.1/1000 pyr; 108.5, 109.7). The effect of gestational age on rehospitalisation was highest during the first year of life and declined by adolescence [e.g. 34-36 weeks: rate ratio = 2.10 (2.04, 2.15) for 29 days-1 year; 1.14 (1.11, 1.18) for 12-18 years]. The risk of rehospitalisation up to 1 year of age has declined since 1980, except for those born <32 weeks.

Conclusions: Rehospitalisation risk is greater for singleton children born at all gestational ages compared with those born full term. This effect of gestational age on rehospitalisation is highest in the first year post-discharge, but has almost disappeared by adolescence.