Introduction
There is an expanding role for anesthesiologists in the preoperative optimization and postoperative management of patients, often in the context of a so-called perioperative surgical home. Such efforts typically include enhanced recovery after surgery (ERAS) protocols and often an anesthesiologist-led team for perioperative management. Studies of the cost-effectiveness of such approaches have generally been conducted at single institutions, with most patients cared for by small numbers of surgeons. This limitation creates generalizability issues as to whether improvement was related mostly to organizational culture or the studied surgeons' practices (non-generalizable) versus the procedures (generalizable). We studied whether other organizations can rely on achieving similar benefits following the adoption of a studied process improvement strategy at a single institution.
Methods
All patients undergoing elective major therapeutic inpatient surgery discharged between October 2015 and June 2022 at non-federal hospitals in the state of Florida were included. For each discharge, the United States Medicare Severity Diagnosis-Related Group (MS-DRG) weighting factor (i.e., the multiplier for the hospital’s base rate for admissions that determines reimbursement) and the Clinical Classification Software Refined (CCSR) code for the principal procedure were determined at admission and discharge from the state's inpatient healthcare database based on the diagnoses present at those time points. An increase in the weighting factor from admission to discharge represents societal costs from perioperative complications. Statewide, by hospital, and by surgeon, we calculated the total increase for each CCSR's weighting factor. Our primary hypothesis was that surgeon variability would be statistically greater than CCSR variability but that the incremental effect would be <5%. If CCSR and surgeon variability were comparable, this would be supportive of generalizability. In contrast, if there were a predominant effect related to the surgeon, results from one institution might not be applicable to others.
Results
Among the 1,482,344 discharges studied, the pooled (N=7 years) contributions to MS-DRG weighting factor increases from the upper 20% of surgeons were 2.8% more than from the upper 20% of CCSRs (95% CI 1.9%-3.9%,
p
=0.0006). Those CCSRs accounted for 85.5% (95% CI 79.4%-91.7%,
p
<0.0001) of the total increase in the MS-DRG weighting factor. The average contribution of the top two surgeons at each hospital to that hospital's increase in the weighting factor ranged among CCSRs from 68% to 97%. The median and 75th percentile of surgeons performing at least 10% of the total number of cases at each hospital was similar to those values for the contributions to the increases in the MS-DRG weighting factor, median 2.0 to 3.0, and 75th percentile 1.75 to 4.0.
Conclusions
Because variability among ...