Impact of Hospital Volume on In-Hospital Mortality of Children Undergoing Repair of Congenital Diaphragmatic Hernia
*Brian T Bucher1, *Rebecca M Guth2, *Jacqueline M Saito1, *Tasnim A Najaf1, Brad W Warner1
1Washington University School of Medicine, St. Louis, MO;2St. Louis Children's Hospital, St. Louis, MO
OBJECTIVES: Congenital diaphragmatic hernia (CDH) continues to remain a significant cause of neonatal morbidity and mortality. Previous studies have suggested hospital volume is an independent predictor of in-hospital mortality. We sought to validate this effect using a large National database incorporating 38 free-standing Children’s Hospitals in the United States.
METHODS: Infants whom underwent repair of CDH from 2000-2008 at Pediatric Health Information Systems (PHIS)-member hospitals were evaluated. Using generalized linear mixed models with random effects, we computed the risk adjusted odds ratio of mortality by yearly hospital volume, after adjustment for salient patient and hospital characteristics.
RESULTS: There were 2231 infants whom underwent repair with an overall survival of 82%. Average yearly hospital volume varied from 1.4 to 17.5 cases per year. Smaller birthweight, year of birth, need for governmental assistance, chromosomal abnormalities, longer time to repair, and requirement for Extracorporeal Membrane Oxygenation (ECMO) or Nitric Oxide were all independently associated with mortality (Table). Compared to low volume, medium and high volume hospitals have a significantly lower mortality. The rate of ECMO use at each facility was not independently associated with mortality.
CONCLUSION: This large, population based study suggests that hospitals who perform high volumes of CDH repairs achieve lower in-hospital mortality. This data support the paradigm of regionalized centers of excellence for the management of infants with this morbid condition.
|Univariate Analysis||Multivariate Analysis|
|Risk Factor||Crude OR||Adjusted OR||p-Value|
|Birthweight||per Kg||0.63 (0.53-0.75)||0.49 (0.39-0.63)||<0.0001|
|Year of Birth||per year from 2000||0.99 (0.95-1.05)||0.89 (0.84-0.95)||0.0006|
|Payer Source||Government Insurance||1||1|
|Private Insurance||0.60 (0.46-0.77)||0.60 (0.43-0.84)||0.004|
|Self-Insured||0.55 (0.21-1.46)||0.76 (0.22-2.66)||0.6|
|Other||0.69 (0.50-0.96)||0.93 (0.61-1.44)||0.8|
|Chromosomal Abnormalities||Present vs. Absent||2.34 (1.24-4.43)||4.60 (1.81-11.71)||0.001|
|Time To Repair||per Week||1.34 (1.24-1.46)||1.07 (1.00-1.23)||0.047|
|Need for ECMO||Present vs. Absent||18.27 (13.62-24.52)||10.54 (7.28-15.24)||<0.0001|
|Need for Nitric Oxide||Present vs. Absent||9.59 (7.11-12.92)||4.54 (3.02-6.84)||<0.0001|
|Hospital Volume||Low Volume|
|0.75 (0.50-1.10)||0.55 (0.33-0.91)||0.02|
|0.52 (0.30-0.89)||0.33 (0.14-0.74)||0.008|
|Hospital ECMO Rate||per 10%||1.34 (1.17-1.54)||1.10 (0.90-1.34)||0.25|