Distribution of pancreatic cancer surgery centers and spatial access to care

Given the well-established relationship between volume and outcomes related to pancreatic cancer surgery, there has been a growing interest in centralizing these procedures at high-volume centers. However, the distribution of high-volume centers and the socioeconomic characteristics of the population that they serve at a national level remain poorly understood.

Centers that performed pancreatic cancer surgery between 2017 and 2021 were identified using the Medicare Standard Analytical Files. Centers with an average annual volume of pancreatic cancer resections meeting the Leapfrog minimum volume thresholds over the study period were defined as high-volume centers. The centers were geolocated based on their address, as reported in the National Plan and Provider Enumeration System database. Service areas representing 90-minute drive-time areas around high-volume centers were defined. Populations within and outside these areas were compared based on their block-level sociodemographic characteristics obtained from the 2021 5-year American Community Survey data.

A total of 832 centers were included in the analysis, of which 76 (9.1%) qualified as high-volume centers. High-volume centers performed a substantially higher number of pancreatectomies over the study period than low-volume centers (median: 70.5 [IQR, 60.2-101.0] vs 5.0 [IQR, 2.0-14.0], respectively) and were responsible for a median of 51.1% (IQR, 50.5%-51.6%) of all resections performed in any given year. A total of 127,519,167 individuals (38.3% of the population) resided outside the service areas of high-volume centers. Compared with individuals residing within the service areas of high-volume centers, those residing outside the service areas of high-volume centers were more often male (49.3% vs 49.8%, respectively), older (40.2 ± 9.7 vs 41.7 ± 10.8, respectively), White (64.0% vs 74.5%, respectively), veterans (6.0% vs 8.1%, respectively), uninsured (8.4% vs 9.3%, respectively), without higher education (54.8% vs 62.5%, respectively), without Internet access (8.9% vs 12.9%, respectively), and with a lower median household income ($85,846 ± 46,180 vs $64,876 ± 32,500, respectively). On multivariate logistic regression, factors associated with a higher likelihood of being included in the service area of a high-volume center were a higher median age (odds ratio [OR], 1.02 [95% CI, 1.01-1.03]), a higher proportion of non-White residents (OR, 1.20 [95% CI, 1.19-1.20]), a higher percentage of households in which English is not the first language spoken at home (OR, 1.23 [95% CI, 1.22-1.24]), and a higher median income (OR, 1.13 [95% CI, 1.13-1.13]). Conversely, factors that led to a lower likelihood of being included in a high-volume service area were a higher proportion of male residents (OR, 0.92 [95% CI, 0.91-0.93]), a higher proportion of Hispanic residents (OR, 0.86 [95% CI, 0.85-0.87]), a higher proportion of veteran residents (OR, 0.62 [95% CI, 0.61-0.63]), more households without Internet access (OR, 0.78 [95% CI, 0.77-0.79]), and a higher rate of noninsured residents (OR, 0.95 [95% CI, 0.94-0.96]).

More than 1 in 3 individuals in the United States may face significant spatial barriers when seeking pancreatic cancer surgery at high-volume centers. These individuals mostly belong to rural, socially vulnerable communities.