In this newly published shoulder arthroplasty meta-analysis, (Journal of Shoulder and Elbow Arthroplasty) the action showed up at two places: 5 cases per year and 25–28 cases per year.
Everything in between was less dramatic, which is not usually how surgeons, hospitals, or patients imagine the learning curve.
Eight retrospective studies representing 332,542 patients were analyzed by researchers at Rothman, Mass General, Hoag Orthopedics, Mayo and the Hand and Upper Extremity Center of San Antonio. The authors tested five annual surgeon-volume thresholds: 5, 10, 14, 20, and 25–28 cases per year.
Only two of those thresholds meaningfully separated outcomes.
The First Gate: Five Cases
At the 5 cases/year cutoff, lower-volume surgeons had higher complication and revision odds than higher-volume surgeons. Complications were 3.7% vs. 2.6%, with an odds ratio (OR) of 1.41 (95% confidence interval [CI], 1.28–1.55; P < .001).
Revisions followed the same pattern: 3.5% vs. 2.6%, with an OR of 1.41 (95% CI, 1.28–1.57; P < .001).
That is the first useful finding. In this dataset, very low annual shoulder arthroplasty volume was not just an academic label. It tracked with worse outcomes.
Then came the boring middle
At 10 cases/year, there was no significant difference in complications or revisions.
At 14 cases/year, revisions were similar. At 20 cases/year, revisions were still similar.
That middle stretch matters because it cuts against the simple story. More volume may be better in the broad sense, but this meta-analysis did not show a clean improvement at every step up the ladder.
The authors described the relationship as nonlinear. That is polite academic phrasing for: The curve did not behave like a hospital administrator’s slide deck.
The Second Gate: 25 to 28 Cases
The next separation showed up around 25–28 cases/year.
At that threshold, lower-volume surgeons again had higher odds of complications and revisions.
Complications were 1.7% vs. 1.5%, with an OR of 1.39 (95% CI, 1.11–1.75; P = .004).
Revisions were 2.5% vs. 2.2%, with an OR of 1.22 (95% CI, 1.15–1.30; P < .001).
The absolute differences look small. The dataset is not. That is where the finding gets its weight.
What To Do With It
This does not prove volume alone makes a surgeon better. It does not create a hard referral rule. It does not separate reverse from anatomic shoulder arthroplasty cleanly enough to hand every procedure its own threshold.
The included studies were retrospective. The analysis could not fully account for implant type, indication, intraoperative variables, case complexity, or the specific complications driving the difference. Many studies did not split anatomic and reverse shoulder arthroplasty outcomes.
Still, the pattern is useful.
The study suggests shoulder arthroplasty experience may have two practical breakpoints: Getting above very low volume, then reaching a more specialized high-volume category.
Five cases a year may move a surgeon out of the basement. Twenty-five to 28 may mark a different neighborhood entirely.
Original Study Title: “Defining surgeon experience thresholds for the reduction in complications and revisions after shoulder arthroplasty: a meta-analysis of 332,542 patients”
Authors: Daher, Mohammad M.D.; Ilyas, Muhammad Hamza M.D.; González-Morgado, Diego M.D.; Steinmann, Scott P. M.D.; Abboud, Joseph A. M.D.; Kassam, Hafiz F. M.D.
