Fw. Moler et Se. Ohmit, Severity of illness models for respiratory syncytial virus-associated hospitalization, AM J R CRIT, 159(4), 1999, pp. 1234-1240
The objective of this investigation was to examine the feasibility of multi
variate severity of illness models for pediatric patients hospitalized with
respiratory syncytial virus (RSV) infection, From a preexisting retrospect
ive cohort study database, all infants and children 2 yr of age or younger
with community-acquired RSV infection admitted to the University of Michiga
n's C. S. Mott Children's Hospital during nine epidemics were examined, The
study group consisted of 802 hospitalized patients younger than 2 yr of ag
e with community-acquired RSV infection; 182 (23%) patients had prolonged h
ospital length of stay defined as 7 d or greater. Multivariate logistic reg
ression modeling of nine variables measurable during the first hospital day
was strongly associated with prolonged hospitalization (p < 0.0001), Recei
ver operator characteristic curve analysis resulted in an area under the cu
rve of 0.894, indicating excellent model discrimination, Goodness-of-fit te
sting indicated excellent model calibration for observed versus predicted o
utcomes (p = 0.216). We conclude that severity of illness models for RSV-as
sociated hospitalization with excellent predictive properties in terms of c
lassification, discrimination, and calibration are possible. Further study
is required to determine if such models are generalizable across multiple c
enters and epidemics.