Background:
Ethnicity has been noted to have many effects on the delivery of health care. Race has been demonstrated as a factor in the appropriate treatment of many illnesses, a marker for access to care, and an important component of disease outcomes. Although many individual diagnoses have been examined for the effect of race, no studies to date have looked at race as an independent factor in overall length of stay (LOS) in hospitalized children. Our goal was to examine the effect of race on LOS for common medical and surgical DRGs in pediatrics.
Methods:
We used the Kids Inpatient Database (KID) of 2003. Eight of the highest volume DRGs (5 medical, 3 surgical) for children < 18 years old were examined for the effect of race on LOS. Newborn, psychiatric, and oncologic DRGs were not considered. The 8 DRGs were: viral meningitis (vm), cellulitis, simple pneumonia, kidney and UTI, bronchitis and asthma, craniotomy, appendectomy, and lower extremity and humerus. Univariate and multivariate analyses were done to see if race significantly affected LOS. SUDAAN software was used to accommodate the complex sample design of the KID.
Table 1. Multivariate LOS Analysis
Results:
For all except vm, univariate analysis showed that race — Black or Hispanic versus white — was a significant factor of LOS, increasing by up to .5 days for medical conditions and 1.4 days for surgical DRGs, almost 20% in both cases. Multivariate analysis confirmed race as an independent variable for 7 of the DRGs (again not vm), with a difference of more than 1 day for craniotomy and up to 26% longer when significant.
Conclusions:
Across the most common diagnoses for which children are discharged from the hospital, race is a significant determinant of LOS. Therefore, racial mix needs to be accounted for in comparing LOS data. These results are consistent with previous studies that identified race as affecting children's health. The reasons for this need to be uncovered, and this issue needs to be examined from the level of an individual hospital up through national policy.
Author Disclosure:
D. Rauch, Baxter, consultant; Pfizer, consultant.