Back to Annual Meeting Program

Introduction:
With improved outcomes demonstrated at high volume centers, many complex surgical procedures have migrated to large, specialized hospitals. The purpose of this study is to examine the extent of regionalization and outcomes in anti-reflux surgery.
Methods:
The Nationwide Inpatient Sample (NIS) data were analyzed from 1998-99 (T1) and 2008-09 (T2) for all antireflux operations in patients with gastroesophageal reflux symptoms using ICD-9-CM codes. Hospitals were stratified into high-, mid-, and low-volume centers (HVC, MVC, LVC) based on annual antireflux surgery volume. Complications and outcomes were also compared. Socio-demographic factors were examined as effectors of surgery location.
Results:
A total of 11804 cases were performed in T1 and 8856 in T2. In T1, 41.0% of procedures were performed in a HVC vs 35.4% in T2. LVC rates increased with time: 20.53% vs. 26.87% (p<0.0001). Rural hospitals had decreased surgical volume (19.10% vs. 10.33%, p<0.0001), while all urban hospitals increased volumes: teaching (48.23% vs. 51.03%, p<0.0001) and non-teaching (32.67% vs. 38.64%, p<0.0001). Using multivariate regression, the following were predictors of surgery at a LVC in T1: non-caucasian race (OR 1.42, p<0.0001), emergent admission (OR 2.24, p<0.0001), living in a zip code with low median income (OR 1.52 lowest vs. highest, p=0.0039), increasing age (p=0.0002), and increasing concurrent diagnosis number (p=0.0029). In T2, emergent admission (OR 1.34, p=0.038), low median income (OR 1.69 highest vs lowest, p<0.0001), and number of concurrent diagnoses (p=0.034) were independent predictors of antireflux surgery at a LVC. In T2, mean LOS at a LVC was 4.0 days vs 3.3days at a HVC (p<0.0001), but this was not significant in multivariate analysis. Total charges were lower at a LVC (\ vs \, p=0.0032) in multivariate analysis. Complication rates increased at all centers with time, but were twice as common in LVCs (6.39% vs. 3.16% at HVCs, p<0.0001) in T2. Controlling for confounding variables, complications remained more likely in LVCs (T1: OR 1.71, p<0.0001, T2: OR 1.49, p<0.0001). In hospital mortality decreased in all centers with time and did not differ significantly in either era. Patients at all centers have increased their mean number of concurrent diagnoses over time(3.92 vs 6.70, p<0.0001).
Conclusion:
Despite improved results at HVCs, LVCs have increased their percentage of antireflux operations over time. The urban non-teaching hospitals have experienced the largest gains in caseload. Overall complication rates have increased with time, possibly due to noted increased incidence of comorbidities in the patients seeking antireflux surgery. After controlling for confounding variables, complications remain more likely in LVCs. Regionalization has not occurred over time, but may improve outcomes if supported.