Research Letter
Revisiting Minimum Case Volume Recommendations for Complex Surgery in Contemporary Practice
Akash Kartik, Zhi Ven Fong, Chee-Chee Stucky, et al
JAMA Published Online: October 8, 2025
doi: 10.1001/jama.2025.17274
The volume-outcome relationship informs policies guiding complex cancer surgery—historically, higher case volumes were associated with better postoperative mortality.1 Consequently, multiple organizations have endorsed minimum case volume recommendations (MCVR) for specific procedures.2 However, postoperative mortality has declined over the past 2 decades.3 This trend may have altered the volume-outcome association, challenging the relevance of MCVR in contemporary practice.
Methods
The National Cancer Database, which captures more than 70% of incident cancers in the US, was used to identify patients undergoing a surgical procedure for localized stage I to III esophageal, gastric, pancreatic, liver, lung, or rectal cancer at a Commission on Cancer–accredited hospital from 2005 to 2019.4 We excluded stage IV disease and facilities without continuous reporting. Hospitals were classified into 2 groups: either meeting or below recommended annual case volume thresholds (≥20 for esophageal, pancreatic and gastric; ≥40 for lung and liver; and ≥16 rectal) as recommended by the Leapfrog group. Benchmark 30-day mortality was defined as the value of the intercept of adjusted 30-day mortality rate for hospitals at the recommended minimum annual case volume for that cancer type, including all patients from 2005 to 2019. Mixed-effects logistic regression modeled 30-day mortality against hospital volume, adjusted for age, race, comorbidity, insurance status, cancer stage, surgical procedure, and facility type, with random intercepts to account for clustering effect. Temporal changes in volume-outcome curves between patients in 2005 to 2009 vs 2015 to 2019 were compared by assessing change in case volume needed to meet benchmark mortality. Trends in 30-day mortality between hospitals below and meeting MCVR were assessed using an interaction term between volume group and year (Supplement 1). A 2-sided P < .05 was considered significant. Missing data were excluded from models. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. All statistical analyses were performed using SAS version 9.4 (SAS Institute) and R version 4.3.1 (R Foundation for Statistical Computing). This study was exempt from institutional review board review due to deidentified data.
Results
A total of 840 292 patients were included. The percentage of hospitals meeting MCVR ranged from 6.9% (esophageal) to 26.9% (lung). Benchmark 30-day mortality rates at hospitals meeting exact MCVR were 1.2% for rectal, 2.2% for esophageal and gastric, 2.4% for liver and lung, and 2.9% for pancreatic surgery. The hospital case volume intersecting benchmark 30-day mortality decreased significantly for the 2015 to 2019 vs 2005 to 2009 group: esophagus, from 36 to 16; gastric, 61 to 12; liver, 99 to 29; pancreas, 39 to 12; and rectum, 21 to 10. This was driven by a leftward shift in volume-outcome curves, indicating improved mortality at lower hospital case volumes (Figure 1). These improvements were greater for hospitals below MCVR threshold compared with those meeting MCVR threshold, although the latter continued to have better outcomes (Figure 2).
Figure 1. Volume-Outcome Curves Representing Adjusted 30-Day Mortality
The blue line represents 30-day mortality at the minimum case volume recommendations (MCVR) for the cohort from 2005-2019 and the dotted lines represent the change in case volume needed to meet this benchmark for the 2005-2009 vs 2015-2019 group.
Figure 2. Adjusted 30-Day Postoperative Mortality
Stratified by hospitals below the recommended annual case volume and hospitals meeting recommended annual case volumes, with a comparison for trends over time using an interaction term for volume group and year to compare the slopes of the 2 curves.
Discussion
Current minimum case volume standards for complex surgical procedures, based on older volume outcome studies, do not align with contemporary practice. These findings demonstrate that volume-outcome curves have changed, with fewer cases needed to meet benchmark 30-day mortality over time. These shifts are most pronounced in hospitals below volume thresholds, likely due to diffusion of best practices such as adoption of enhanced recovery protocols, minimally invasive techniques, quality improvement initiatives, and emphasis on a culture of patient safety. Although high-volume hospitals continue to show improved outcomes, the differences between hospitals above and below MCVR have diminished. These findings suggest that even though volume continues to correlate with improved outcomes, this association has evolved in contemporary practice due to the left shift in volume-outcome curves. A reassessment of MCVR is needed to align with contemporary surgical practice.5 Continued emphasis on regionalization,6 which proposes shifting all complex surgery to hospitals meeting MCVR, is not supported by our data. Not only does this policy risk exacerbating disparities in access, but also does not address issues such as care fragmentation and patient choice. Updated MCVR may enable more hospitals to meet volume thresholds and promote equitable access to care while maintaining high-quality surgical outcomes. Limitations include that the data used for were retrospective and demonstrate associations, not causality, and it was not possible to assess changes in surgeon-specific MCVR given the limitations of the National Cancer Database.
