Geographic Variation in the Utilization of and Mortality After Emergency General Surgery Operations in the Northeastern and Southeastern United States

Objective: To define geographic variations in emergency general surgery (EGS) care, we sought to determine how much variability exists in the rates of EGS operations and subsequent mortality in the Northeastern and Southeastern United States (US). Summary Background Data: While some geographic variations in healthcare are normal, unwarranted variations raise questions about the quality, appropriateness, and cost-effectiveness of care in different areas. Methods: Patients >= 18 years who underwent 1 of 10 common EGS operations were identified using the State Inpatient Databases (2011-2012) for 6 states, representing Northeastern (New York) and Southeastern (Florida, Georgia, Kentucky, North Carolina, Mississippi) US. Geographic unit of analysis was the hospital service area (HSA). Age-standardized rates of operations and in-hospital mortality were calculated and mapped. Differences in rates across geographic areas were compared using the Kruskal-Wallis test, and variance quantified using linear random-effects models. Variation profiles were tabulated via standardized rates of utilization and mortality to compare geographically heterogenous areas. Results: 227,109 EGS operations were geospatially analyzed across the 6 states. Age-standardized EGS operation rates varied significantly by region (Northeast rate of 22.7 EGS operations per 10,000 in population versus Southeast 21.9; P < 0.001), state (ranging from 9.9 to 29.1; P < 0.001), and HSA (1.9-56.7; P < 0.001). The geographic variability in age-standardized EGS mortality rates was also significant at the region level (Northeast mortality rate 7.2 per 1000 operations vs Southeast 7.4; P < 0.001), state-level (ranging from 5.9 to 9.0 deaths per 1000 EGS operations; P < 0.001), and HSA-level (0.0-77.3; P < 0.001). Maps and variation profiles visually exhibited widespread and substantial differences in EGS use and morality. Conclusions: Wide geographic variations exist across 6 Northeastern and Southeastern US states in the rates of EGS operations and subsequent mortality. More detailed geographic analyses are needed to determine the basis of these variations and how they can be minimized.