This study examines the existence and sources of variation in the management of pediatric splenic injuries among hospitals in the United States and the factors associated with splenectomy.Methods:
Information on children 15 years of age and younger with a splenic injury diagnosis code was extracted from the Kids’ Inpatient Database 2000, a pediatric inpatient database of 2,784 hospitals in 27 states covering 72% of the nation’s population for the year 2000. Patient variables included age, sex, race, injury diagnoses, grade of splenic injury, splenic procedure code, and calculated Injury Severity Score. Hospital variables included pediatric status (free-standing, unit and adult), teaching status, annual pediatric splenic trauma volume, and national region. A multivariate logistic regression model was used to predict the factors associated with splenectomy based upon patient and hospital characteristics.Results:
In all, 2,191 children with splenic injuries were identified; 253 (12%) underwent splenectomy. The crude rate of splenectomy varied significantly among pediatric hospital types: 3% (11/339) at freestanding children’s hospitals, 9% (45/525) at unit hospitals and 15% (197/1327) at adult hospitals (p < 0.001). Risk of splenectomy increased with the grade of splenic injury, patient age, and the presence of multiple injuries. Teaching hospitals and hospitals with higher patient volume were associated with lower risk for splenectomy. There was no relationship between splenectomy and gender, race, or national region. Despite adjustment for the above noted hospital and patient-specific variables, children treated at an adult hospital had 2.8 times the odds, and those treated at a unit pediatric hospital 2.6 times the odds, of undergoing splenectomy as those cared for at a free-standing pediatric hospital (p = 0.003 and 0.013, respectively).Conclusion:
Nationally, children cared for at freestanding pediatric hospitals have a significantly lower risk of splenectomy than children treated at either adult hospitals or pediatric hospitals within an adult hospital. This may have implications for education, trauma triage and the establishment of practice guidelines.