Impact of the thresholds of surgical volume on mortality in oncological digestive surgery: A French National Cohort

Objective

To report the 90-day mortality benefit associated with the implementation of the new regulatory reform on oncological surgical digestive authorizations.

Summary background data

New thresholds in digestive cancer surgery were applied in 2023, accrediting centers for major interventions. No evidence has been provided to support their justification.

Methods

Any French adult operated for digestive cancer from January 1, 2019 to December 31, 2021 was included from the PMSI. A 90-day mortality logistic regression was performed by adjusting by age, sex, Charlson score, Frailty index, hospital-volume (<5 or ≥5 interventions/year), emergency intervention, specialty.

Results

143,582 patients were identified. Of these, 64,268 underwent surgery of one of the subspecialties subject to the new thresholds (stomach N = 8283, liver N = 18,668, pancreas N = 11,220, esophagus N = 3704, rectum N = 22,393).

4808 (7.5 %) patients underwent surgery in low-volume centers, distributed as follows: stomach 1757/8283 (22.94 %), liver 970/18,668 (5.19 %), pancreas 895/11,220 (7.98 %), esophagus 672/3704 (18.14 %) and rectum 514/22,393 (2.29 %).

In univariate analysis, the 90-day mortality rate was significantly lower in high-volume centers, for all subspecialties, gastric: 127/1757 (7.23 %) vs 330/6526 (5.06 %), p = 0.0004, hepatic: 64/970 (6.6 %) vs 824/17,698 (4.66 %), p = 0.006, pancreatic: 74/895 (8, 27 %) vs 608/10,325 (5.89 %), p = 0.004, esophageal: 58/672 (8.63 %) vs 195/3032 (6.43 %), p = 0.04, rectal 26/514 (5.06 %) vs 639/21,879 (2.92 %), p = 0.005.

The multivariate analysis, showed a mortality reduction for high-volume centers: OR = 0.78 CI95[0.71–0.87], p < 0.001.

Conclusion

The recent implementation of regulatory decrees appears to be justified. The enforcement of these hospital volume thresholds is likely to contribute to a reduction in postoperative mortality following digestive cancer surgery.