Hospital safety-net burden is associated with increased inpatient mortality after elective total knee arthroplasty: a retrospective multistate review, 2007-2018

BACKGROUND: Total knee arthroplasty (TKA) is among the most common surgical procedures performed in the USA and comprises an outsized proportion of Medicare expenditures. Previous work-associated higher safety-net burden hospitals with increased morbidity and in-hospital mortality following total hip arthroplasty. Here, we examine the association of safety-net burden on postoperative outcomes after TKA.

METHODS: We retrospectively analyzed 1 141 587 patients aged ≥18 years undergoing isolated elective TKA using data from the State Inpatient Databases for Florida, Kentucky, Maryland, New York and Washington from 2007 through 2018. Hospitals were grouped into tertiles by safety-net burden status, defined by the proportion of inpatient cases billed to Medicaid or unpaid (low: 0%-16.83%, medium: 16.84%-30.45%, high: ≥30.45%). Using generalized estimating equation models, we assessed the association of hospital safety-net burden status on in-hospital mortality, patient complications and length of stay (LOS). We also analyzed outcomes by anesthesia type in New York State (NYS), the only state with this data.

RESULTS: Most TKA procedures were performed at medium safety-net burden hospitals (n=6 16 915, 54%), while high-burden hospitals performed the fewest (n=2 04 784, 17.9%). Overall in-patient mortality was low (0.056%), however, patients undergoing TKA at medium-burden hospitals were 40% more likely to die when compared with patients at low-burden hospitals (low: 0.043% vs medium: 0.061%, adjusted OR (aOR): 1.40, 95% CI 1.09 to 1.79, p=0.008). Patients who underwent TKA at medium or high safety-net burden hospitals were more likely to experience intraoperative complications (low: 0.2% vs medium: 0.3%, aOR: 1.94, 95% CI 1.34 to 2.83, p<0.001; low: 0.2% vs high: 0.4%, aOR: 1.91, 95% CI 1.35 to 2.72, p<0.001). There were no statistically significant differences in other postoperative complications or LOS between the different safety-net levels. In NYS, TKA performed at high safety-net burden hospitals was more likely to use general rather than regional anesthesia (low: 26.7% vs high: 59.5%, aOR: 4.04, 95% CI 1.05 to 15.5, p=0.042).

CONCLUSIONS: Patients undergoing TKA at higher safety-net burden hospitals are associated with higher odds of in-patient mortality than those at low safety-net burden hospitals. The source of this mortality differential is unknown but could be related to the increased risk of intraoperative complications at higher burden centers.