Back to 2014 Fall Congress Meeting Abstracts

BACKGROUND: Increasing evidence has suggested that hospital and provider surgical volume may affect surgical outcomes. However, this topic has rarely been addressed in children. Our objective was to investigate whether hospital surgical volume impacts outcomes (post-operative complications) through a representative national database.
MATHODS: We retrospectively reviewed the Nationwide Inpatient Sample (1998-2011) for pediatric (≤ 18 years) admissions for urologic procedures. We used ICD-9-CM codes to identify elective urologic interventions (ureteral reimplant, uretero-ureterostomy, pyeloplasty, radical nephrectomy, partial nephrectomy, bladder exstrophy repair, appendicovesicostomy, bladder augmentation, vesicostomy, sling, and percutaneous nephrolithotomy), and National Surgical Quality Improvement Program (NSQIP) in-hospital post-operative complications. Elixhauser comorbidity index was abstracted using comorbidity software. Annual hospital surgical volume was calculated and dichotomized as high-volume (>90th percentile) or non-high-volume hospitals (<90th percentile).
RESULTS: In total, we identified 158,804 urologic surgery admissions (114,634 high-volume hospital admissions and 44,171 non-high-volume hospital admissions). 75% of hospitals performed fewer than 7 major pediatric urology cases per year. High-volume hospitals treated significantly younger patients (mean 5.4 vs 9.6 years, p<0.001), and were more likely to be teaching hospitals (93% vs 45%, p<0.001). Complication rates did not vary significantly over time (p=0.95 for trend). The overall rate of NSQIP-identified postoperative complications significantly higher at non-high-volume hospitals (11.6%) than at high-volume hospitals (9.3%, p=0.003). After adjusting for age, gender, insurance status, Elixhauser index, treatment year, surgery type, hospital teaching status, hospital size, and geographic region, patients treated at non-high-volume hospitals remained more likely to suffer multiple NSQIP-tracked postoperative complications: acute renal failure (OR 1.4, p=0.04), UTI (OR 1.3, p=0.01), post-op respiratory complications (OR 1.5, p=0.01), systemic sepsis (OR 2.0, p=<0.001), post-op bleeding (OR 2.5, p<0.001), and in-hospital death (OR=2.2, p=0.007).
CONCLUSIONS: Procedures performed at non-high-volume hospitals were associated with a significantly elevated risk of in-hospital, NSQIP-identified postoperative complications, including postoperative death.