Effect of Obesity on Outcomes following Total Hip/Total Knee Arthroplasty among Medicare Beneficiaries with Osteoarthritis

Abstract

This study examined the effect of obesity on immediate rehabilitation outcomes including functional performance, length of stay (LOS) at inpatient rehabilitation facility (IRF), and discharge destination following IRF stay, and on long-term outcomes following discharge from the IRF including 30-day hospital readmissions, the reasons for readmission and mortality within the duration of the study period. The study population was Medicare beneficiaries 65 years and older, with osteoarthritis, who underwent elective primary total hip (THA) or total knee arthroplasty (TKA) during the years 2012 and 2013 and were directly admitted to IRF. The study design was a secondary data analysis of 100% Medicare Claims data. Outcomes deemed undesirable included longer LOS, lower functional status, discharge to a non-community setting, hospital readmission, and/or death, after completion of IRF stay. Reasons for hospital readmission were further classified based on their connection to the index surgical procedure: local complications, systemic complications, or unrelated. Chi-square statistics and one-way ANOVA were used for descriptive statistics. Multivariable linear regression analysis was used for each of the numerical outcomes: discharge motor function, discharge cognition function, and IRF LOS. Multivariable logistic regression was used for the categorical outcomes: community discharge and 30-day hospital readmission. Among each of the THA and TKA sub-cohorts of beneficiaries who were readmitted, multinomial logistic regression was used for the categorical outcome of reason for readmission. Survival analyses using cox proportional hazard modeling were conducted for 30-day hospital readmission and mortality. Normal weight was used as the reference category in all multivariable analyses. Differential effects of obesity were also examined by race/ethnicity and gender. Overweight-obesity was associated with higher discharge motor functional status. Morbid obesity was associated with lower discharge motor functional status, but a higher discharge cognition status. Obesity status was not significantly associated with difference in IRF LOS or with the likelihood of community discharge. In the THA cohort, time to 30-day hospital readmission was significantly different between the three obesity-related categories, and morbid obesity was significantly associated with greater risk for 30-day hospital readmission. Among those who were readmitted, morbid obesity was associated with greater odds of local/procedure-related reasons for readmission. Among the THA cohort, time to death was significantly different between the three obesity-related categories, and morbid obesity had a protective effect for the outcome of risk for mortality.