Demographic and outcome evaluation of a paediatric intensive care unit - Baragwanath Hospital ICU - 6 years' experience, 1989-1994

Introduction. This is a retrospective study of demographic and outcome data on 1 528 pediatric patients admitted to Baragwanath intensive care unit over a 6-year period from 1989 to 1994. Aim. To define the population of patients admitted, and to evaluate the pooled demographic and outcome data. To evaluate the performance of the Paediatric Risk of Mortality (PRISM) score in this population. Patients and methods. This is a retrospective, descriptive study that was undertaken at Baragwanath intensive care unit from 1989 to 1994. Paediatric patients comprised 40% of all ICU admissions in the study period, and made up 6% of all admissions to the Department of Paediatrics. Basic demographic information was retrieved from a computerised database and analysed with standard statistics, PRISM score (which provides the predicted risk of mortality) and the actual observed mortality rate were also compared. Results. The overall mortality rate was 32% with a mean PRISM of 29%. Non-surgical cases accounted for 96% of all admissions, and 97% of all admissions were nonelective. The mean duration of admission was 12 days (survivors 14 days and non-survivors 8 days). Respiratory disorders accounted for the majority of admissions to the unit, There was a significant difference between actual mortality and mortality predicted by PRISM score (goodness-of-fit, P < 0.01). Discussion. American and European ICUs report a lower mortality rate (around 8%), but admit older patients and more elective and routine postoperative cases; these patients have a shorter stay in the ICU. Most studies are from middle- and upper-middle-class areas, where malnutrition was not a problem. The high mortality observed and the poor correlation with PRISM is postulated to be a result of: (i) delay in ICU admission; (ii) longer duration of ICU stay and increased nosocomial sepsis; (iii) few surgical patients; and (iv) possible malnutrition. The PRISM score should be validated in the Baragwanath population, and the possible effects of malnutrition on outcome established.