Association of Total Knee Replacement Removal From the Inpatient-Only List With Outpatient Surgery Utilization and Outcomes in Medicare Patients

Abstract

Little is known about the association of total knee replacement (TKR) removal from the Medicare inpatient-only (IPO) list in 2018 with outcomes in Medicare patients. To evaluate (1) patient factors associated with outpatient TKR use and (2) whether the IPO policy was associated with changes in postoperative outcomes for patients undergoing TKR. This cohort study included data from administrative claims from the New York Statewide Planning and Research Cooperative System. Included patients were Medicare fee-for-service beneficiaries undergoing TKRs or total hip replacements (THRs) in New York State from 2016 to 2019. Multivariable generalized linear mixed models were used to identify patient factors associated with outpatient TKR use, and with a difference-in-differences strategy to examine association of the IPO policy with post-TKR outcomes relative to post-THR outcomes in Medicare patients. Data analysis was performed from 2021 to 2022. IPO policy implementation in 2018. Use of outpatient or inpatient TKR; secondary outcomes included 30-day and 90-day readmissions, 30-day and 90-day postoperative emergency department visits, non-home discharge, and total cost of the surgical encounter. A total of 37 588 TKR procedures were performed on 18 819 patients from 2016 to 2019, with 1684 outpatient TKR procedures from 2018 to 2019 (mean [SD] age, 73.8 [5.9] years; 12 240 female [65.0%]; 823 Hispanic [4.4%], 982 non-Hispanic Black [5.2%], 15 714 non-Hispanic White [83.5%]). Older (eg, age 75 years vs 65 years: adjusted difference, -1.65%; 95% CI, -2.31% to -0.99%), Black (-1.44%; 95% CI, -2.81% to -0.07%), and female patients (-0.91%; 95% CI, -1.52% to -0.29%), as well as patients treated in safety-net hospitals (disproportionate share hospital payments quartile 4: -18.09%; 95% CI, -31.81% to -4.36%), were less likely to undergo outpatient TKR. After IPO policy implementation in the TKR cohort, there were lower adjusted 30-day readmissions (adjusted difference [AD], -2.11%; 95% CI, -2.73% to -1.48%; P < .001), 90-day readmissions ( -3.23%; 95% CI, -4.04% to -2.42%; P < .001), 30-day ED visits ( -2.45%; 95% CI, -3.17% to -1.72%; P < .001), 90-day ED visits (-4.01%; 95% CI, -4.91% to -3.11%; P < .001) and higher cost per encounter ($2988; 95% CI, $415 to $5561; P = .03). However, these changes did not differ from changes in the THR cohort except for increased TKR cost of $770 per encounter ($770; 95% CI, $83 to $1457; P = .03) relative to THR. In this cohort study of patients undergoing TKR and THR, we found that older, Black, and female patients and patients treated in safety-net hospitals may have had lesser access to outpatient TKRs highlighting concerns of disparities. IPO policy was not associated with changes in overall health care use or outcomes after TKR, except for an increase of $770 per TKR encounter.