Paediatric admissions to the general intensive care unit at palmerston north hospital.

OBJECTIVES To report Palmerston North Hospital's (PNH) recent experiences with paediatric admissions to the general Intensive Care Unit (ICU), and to identify any aspects relevant to regionalisation of paediatric intensive care. METHODS Retrospective review of the ICU database and admission register to enable clinical and demographic profiling of all paediatric (aged < 15 years) admissions to PNH ICU from 1996 to 1998. Observed and predicted mortalities were compared using the Paediatric Index of Mortality (PIM). RESULTS Seventy-five paediatric patients had 76 admissions, comprising 7.4% of ICU admissions during the study period. Forty-eight (64%) were male, and 59 (79%) of the patients received mechanical ventilatory support. The median ICU stay was 21.5 hours (range 0.1 to 568 hours). There were 3 deaths in the ICU, and one following ICU discharge but prior to hospital discharge, compared with 6.37 predicted deaths. The median PIM score overall was 5.3% predicted probability of death (range 0.8 - 61.4%), with 6 patients (7.9%) having a PIM score of greater than 30%. Fifteen patients (20%) were transferred to tertiary services. CONCLUSIONS If the PNH experience reflects that of other similar institutions, then non-tertiary ICUs admit small numbers of critically ill paediatric patients who tend to be of low to moderate severity, but who cover the full spectrum of severity. Most cases can be well managed locally, but appropriate referral and transfer is an important component in the delivery of a rational and integrated paediatric intensive care service.