Background/Purpose: To determine whether frailty is associated with patient outcomes 1 year after total knee replacement (TKR).

Methods: Patients ≥65yo scheduled for elective primary TKR for osteoarthritis were recruited from a musculoskeletal specialty hospital. All were approved for surgery by an internist. Frailty was defined using composite criteria designed for surgical candidates, evaluated at point of care. Knee Injury and Osteoarthritis Outcome Score (KOOS) and PROMIS29 were administered pre-operatively and at 1-year. Depression was measured using the Center for Epidemiologic Studies Depression Scale (CES-D 10). A subset of patients had frailty assessed by their surgeon pre-operatively using the brief, self-report Clinical Frailty Scale for Physicians. Wilcoxon rank-sum and Fisher’s exact tests were used to compare continuous variables and categorical variables, respectively, across frailty status. Multivariable linear and logistic regressions were modeled for each patient-related outcome, adjusting for confounders identified a priori using a directed acyclic graph approach. Spearman correlation coefficients were calculated to evaluate the correlation between frail scales.

Results: 487 knee replacement subjects enrolled, 6% of whom were frail. Frail TKR subjects were more likely to be depressed and have concurrent back pain. 89.9% provided 1-year follow-up. In a multivariable linear regression controlling for age, gender, race, back pain, education, Charlson comorbidity index, and PROMIS29 anxiety and depression, frailty predicted a statistically significant decrease in KOOS Function in Daily Living score at 1-year (β=7.0; p=0.04). Frailty also predicted a significant decrease in PROMIS29 physical function, (β=4.7; p-value 0.002), and a small increase in PROMIS29 Pain Intensity, (β=0.69; p-value 0.047), at 1-year. Frailty did not predict other KOOS subscales or PROMIS29 domains at 1-year. Frailty did not predict cumulative severe adverse events through 1-year or OMERACT-OARSI responder status at 1-year. There was a strong correlation between our frailty definition and the Fried frailty phenotype, (r=0.79; p< 0.001), but a poor correlation with the orthopedic surgeon’s frailty assessment (r=0.18; p=0.6).

Conclusion: Frailty was a strong independent predictor of worse KOOS knee-specific function and PROMIS29 global physical function 1 year after TKR in older adults undergoing elective primary TKR. Frailty was not associated with odds of adverse events or OMERACT-OARSI responder status. Surgeon assessment of frailty is not an accurate proxy for the frailty phenotype. These data can help manage expectations of frail patients contemplating TKR. Future studies should assess if targeted interventions to improve frailty can lead to better post-operative function.

Pre-Op Table -TKR-

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/frail-patients-have-worse-function-after-total-knee-replacement/