Article Text
PDF
Abstract
Aims and background The admission numbers to PICUs in the UK is rising. This increase is most probably attributable to rising birth rates, as admission rates have remained stable. The increase in admission numbers has been accompanied by a reduction in crude mortality. There is, however, a perception amongst some paediatric intensivists that the length of stay of children who die in PICU is increasing over time.
Our aim was to investigate whether there is a clinically and statistically significant increase in the length of time that some children spend in a PICU before their death.
Methods Data on all admissions from November 2002 to September 2013 were extracted from the PICANet database. Mean length of stay (LOS) in days was plotted against individual year for survivors and non-survivors. Linear regression was used to assess the relationship between LOS and year. The following parameters available on the database were analysed to establish if LOS for children who die on PICU is increasing in one section of the PICU population: diagnostic group, sex, ethnicity, deprivation, expected probability of mortality (PIM2) and individual PICU.
Results Using simple linear regression, there was a significant increase in LOS of 0.245 days per year (95% CI; 0.111–0.379) for non-survivors in PICU. There was a very slight increase in LOS for children discharged alive (0.039 days per year, 95% CI; 0.022–0.059) – Figure 1. Random effects model reduced these estimates to 0.159, (95% CI; 0.021–0.297) and 0.033 (95% CI 0.015–0.052) days per year respectively reflecting a high degree of variability between PICUs. There was also considerable variability in the coefficients estimated by subgroup analysis, although there was a statistically significant increase in LOS for non-survivors with respiratory conditions (0.297 days per year, 95% CI 0.0376–0.557) and for those aged 1–4 years (0.32, 95% CI 0.119–0.549).
Conclusion The results support the hypothesis that length of stay in non-survivors is increasing over time. This simple observation does not appear to have a simple explanation. It may reflect an emerging societal trend, manifesting as changing physician behaviour and parental expectations.