Model-based pain and function outcome trajectory types for patients undergoing knee arthroplasty: a secondary analysis from a randomized clinical trial

Abstract

Objective: Knee arthroplasty (KA) is an effective surgical procedure. However, clinical studies suggest that a considerable number of patients continue to experience substantial pain and functional loss following surgical recovery. We aimed to estimate pain and function outcome trajectory types for persons undergoing KA, and to determine the relationship between pain and function trajectory types, and pre-surgery predictors of trajectory types. Design: Participants were 384 patients who took part in the KA Skills Training randomized clinical trial. Pain and function were assessed at 2-week pre-and 2-, 6-, and 12-months post-surgery. Piecewise latent class growth models were used to estimate pain and function trajectories. Pre-surgery variables were used to predict trajectory types. Results: There was strong evidence for two trajectory types, labeled as good and poor, for both Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) Pain and Function scores. Model estimated rates of the poor trajectory type were 18% for pain and function. Dumenci's latent kappa between pain and function trajectory types was 0.71 (95% CI: 0.61e0.80). Pain catastrophizing and number of painful body regions were significant predictors of poor pain and function outcomes. Outcome-specific predictors included low income for poor pain and baseline pain and younger age for poor function. Conclusions: Among adults undergoing KA, approximately one-fifth continue to have persistent pain, poor function, or both. Although the poor pain and function trajectory types tend to go together within persons, a significant number experience either poor pain or function but not both, suggesting heterogeneity among persons who do not fully benefit from KA.

What Have We Learned About the Course of Clinical Outcomes After Total Knee or Hip Arthroplasty?

“…The most frequent statistical model reported was a hierarchic linear model, also described as latent growth curve model, generalized linear mixed model, or linear mixed model with random intercept (see Supplementary with similar patterns of change over time. A latent class growth analysis (10), a piece-wise latent class growth analysis (11,12), a growth mixture model (addition of random effects) (13,14) and a latent class mixture model (15) were used to identify subgroups with similar patterns of change over time.…”

Serum Cartilage Oligomeric Matrix Protein in Late-Stage Osteoarthritis: Association with Clinical Features, Renal Function, and Cardiovascular Biomarkers

“…In late-stage OA, joint replacement is mostly the last effective treatment option to alleviate pain and loss of functionality (including joint stiffness) in order to improve the quality of life [4]. While hip and knee arthroplasty are highly successful and cost-effective in most cases [5], about 10% to 20% of patients do not benefit much from this surgical approach, an aspect which is still insufficiently understood [6,7].…”

Classifications of good versus poor outcome following knee arthroplasty should not be defined using arbitrary criteria

“…Variation precludes consensus and prevents meaningful comparisons across study cohorts. We noted an additional problem with evidence classifying outcome as good or poor [4]. Definitions of good versus poor outcome are grounded in the use of arbitrary cutoff values, whether based on final outcome score, percent or absolute change from baseline or the Minimal Clinically Important Difference (MCID) family of change indicators.…”

“…In 2011, we further elaborated on why the cutoff method should not be used to determine patient groupings in scientific research, developed methods originating from discrete latent variable modeling approaches to circumvent problems associated with the arbitrary cutoff method, and provided multiple examples using real-life data to illustrate how new methods could be used to answer scientific questions [10]. In 2019, we used methods originating from a longitudinal discrete latent variable modeling framework to define poor versus good outcomes in KA [4]. For reasons that were unclear to us, given that it met inclusion criteria by te Molder and colleagues, our 2019 study [11] was not included in the review.…”

“…The second remark relates to the reason why the 2019 study by Dumenci et al [4] was not included in the review. As the authors indicated in their correspondence, inclusion criteria for our inventory review focused on predefined dichotomized cutoffs to define poor outcome and, therefore, studies including model-based approaches (such as the study by Dumenci et al) were not included.…”