Inequal benefits from regionalization of cancer care: The pancreatic cancer surgery paradigm.

Abstract

4059 Background: Regionalization has been proposed for high-level care, including multidisciplinary cancer treatment and complex procedures. Pancreatic resections can serve as a marker for both. Using Massachusetts Division of Health Care Finance and Policy (DHCFP) data, we investigated regionalization of surgery for pancreatic cancer (PCa), its potential effect on perioperative outcomes, and disparities in access to high-volume PCa surgery centers. Methods: Using MA DHCFP Hospital Inpatient Discharge Data, 2005-2009, 10,524 discharges for PCa were identified, of which 746 were associated with pancreatic resection. Discharges with missing or out-of-state residence were excluded (n=704). Using geodetic methods and ZIP codes, center-to-center distances were calculated between patient (pt) and treating hospital. Median ZIP income was estimated from 2009 census data. High volume hospitals (4 of 25 performing pancreatic resections in MA) were defined using Leapfrog Criteria (> 11 per year (87th percentile for MA). Chi-square and logistic regression analyses were performed using SAS software. Results: Median age was 65. Pts were predominantly White (87.2%), with median ZIP income of $54,677. Pts travelled in-state up to 112 miles (median 15.4), with the majority resected at high volume hospitals (76%). Median length of stay (LOS) was 8.0 days, with LOS>1 week associated with low volume hospitals (p=0.0002). Of 14 in-hospital deaths, 7 were at low volume hospitals (4.14% of 169 pts) compared to 7 at high volume hospitals (1.31% of 535 pts) (p=0.0214). Predictors of shorter travel distance were: Black race (OR 4.45 (95% CI 1.66-11.93)), operation at low volume hospital (OR 2.62 (95% CI 1.81-3.77), and increased age (per year) (OR 1.02 (95% CI 1.00-1.03), but not sex or median income. Conclusions: Using MA statewide discharge data, regionalization of pancreatic cancer surgery to high-volume, better-outcome centers is seen to be occurring. However, it is not uniform, and disparities exist between groups of cancer pts that do and do not travel for their care. In the current era of scrutiny on cost, quality, and access to cancer care, further study into predictors of pts receiving optimal care is warranted.