Demographic and Socioeconomic Determinants Are Associated with Poor Preoperative Patient-Reported Pain and Function in Primary TKA

Abstract

Background: Pain and function, as reflected by patient-reported outcome measures (PROMs), can influence improvement after total knee arthroplasty (TKA) and can reflect the extent of patient access to orthopaedic surgical care. We aimed to (1) categorize patients according to pain and function PROM phenotypes, (2) identify patient characteristics associated with poor preoperative pain and function, and (3) assess relationships between baseline characteristics and PROM phenotypes. Methods: A prospective cohort of 14,079 TKAs was enrolled. Demographics, comorbidities, surgical details, and preoperative PROMs were collected. Outcomes included preoperative Knee injury and Osteoarthritis Outcome Score (KOOS) Pain (P) and Physical Function (PS) subscores, stratified by demographics. Patients were then categorized into 4 PROM phenotypes defined on the basis of the cohort medians: above or equal to the median for both pain and function scores (P+PS+), below the median for both pain and function (P−PS−), above or equal to the median for pain but below the median for function (P+PS−), and below the median for pain but above or equal to the median for function (P−PS+). Descriptive statistics and multivariable regression analyses were calculated. Results: The largest PROM phenotype was P−PS− (39.4%), followed by P+PS+ (38.9%). The cohort with discordantly poor function but high pain scores (less pain) was the smallest cohort (9.9%). Preoperative KOOS-Pain and KOOS-PS scores at or below the 25th percentile were independently associated with younger age, female sex, higher body mass index (BMI), non-White race, current smoking, lower education, non-commercial insurance, and higher Charlson Comorbidity Index (CCI). Multivariate logistic regression showed that patients in the P+PS+ category were older (odds ratio [OR] = 1.56), were more likely to be male (OR = 2.00), had a lower BMI (OR = 0.67), had more education (OR = 1.63), had a lower CCI, and were less likely to be Black (OR = 0.80) or Other (OR = 0.62) race, be a current smoker (OR = 0.62), and have commercial insurance (OR = 0.74), compared with the P−PS− phenotype. Conclusions: Younger age, obesity, non-White race, female sex, current or recent smoking, non-commercial insurance, and higher CCI were associated with worse pre-TKA PROMs and poor pain-function phenotype combinations. Such a pattern may indicate barriers to TKA access among these patient populations leading to advanced levels of impairment at the time of treatment. Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.