How patients do depends remarkably on the surgeon they choose or get. The evidence has been saying so for two decades.

Across 2.5M Medicare procedures, surgeons in the highest-volume quintile had operative mortality 24–58% lower than the lowest-volume quintile, depending on the procedure. Effect persisted after adjusting for hospital volume and case mix.

Across 17 practicing surgeons, technical skill scores (rated by blinded peers on colectomy video) explained ~25% of the variation in postoperative complication rates. Higher-rated surgeons had significantly lower rates of any complication, reoperation, and a composite of death or serious morbidity.

15,755 patients across Michigan hospitals: complication rate ranged 8.7–30.2% across hospitals. After risk adjustment, patient factors explained 35% of variance, surgeon 2.4%, hospital 1.8% — the residual 60%+ is unexplained by any of the three, leaving room for unmeasured surgeon effect.

2,525 colectomy patients across 276 surgeons at 44 Maryland hospitals. Risk-adjusted postoperative complications varied 0–30% by surgeon — more than by hospital (0–18.2%). Surgeon use of laparoscopy was the single strongest modifiable predictor of lower complications.

20 bariatric surgeons ranked by blinded peer video review. Skill scores spanned 2.6–4.8 / 5.0. Higher-skill surgeons had significantly lower rates of complications, reoperation, and readmission — and the effect extended to long-term (1-year) outcomes, not just perioperative events.

314,872 colonoscopies by 136 gastroenterologists. Adenoma detection rate ranged 7.4–52.5%. Patients of physicians in the highest ADR quintile had a 48% lower risk of interval colorectal cancer (HR 0.52) and a 62% lower risk of fatal interval cancer (HR 0.38) than the lowest quintile.

2,352,174 PCIs at 1,373 hospitals by 3,760 operators in the NCDR CathPCI Registry. 6.5% of operators had risk-standardized mortality >2σ above the mean and 4.1% had RSMR >2σ below — not explained by case mix. PCI operator mortality is a real, measurable axis of variance, though unstable year-to-year.

4,629 radical prostatectomies by 44 surgeons at MSKCC and a second center. Among surgeons with >10 cases, the positive surgical margin rate ranged 10–48%. Surgeon identity was independently associated with margin status after adjusting for PSA, stage, grade, and case volume.

SEER–Medicare analysis restricted to high-volume surgeons. Even after restricting to high-volume practice, complication rates varied substantially between individual surgeons — arguing that case volume alone is a blunt proxy and surgeon technique is a real axis of variance.

All non-federal hospitals in Washington State, 2004–2007. Complication and reoperation rates after lumbar fusion varied significantly across both hospitals and individual surgeons, with surgeon identity explaining a non-trivial share of the variance after case-mix adjustment.