AMSTERDAM – Patients with milder knee osteoarthritis symptoms or better quality of life before undergoing total knee replacement surgery gained less benefit from the surgery than did those who had more severe symptoms in two separate analyses of British and U.S. patients.
Additional evidence from total knee replacements (TKRs) performed on U.S. participants of the Osteoarthritis Initiative also suggest that as the use of TKR has increased to include less symptomatic patients, the overall cost-effectiveness of the procedure has declined.
“Knee replacements are one of those interventions that are known to be very effective and very cost-effective,” Rafael Pinedo-Villanueva, Ph.D., of the University of Oxford (England) said during his presentation of National Health Service data from England at the World Congress on Osteoarthritis. Indeed, knee replacements are associated with significant improvements in pain, function, and quality of life, he said, but that is if you look at the mean values.
As the deciles for baseline knee pain and function decrease in severity, there are diminishing mean improvements and an increasing proportion of patients who do worse after the operation, he reported. Up to 17% of patients had unchanged or improved knee pain scores and up to 27% had lower quality of life scores. If minimally important differences were considered, these percentages rose to 40% and 48% of patients being worse off, respectively.
“The significant improvements seem to be overshadowing what happens to those patients who are doing worse,” Dr. Pinedo-Villanueva suggested. “So essentially cost-effectiveness is being driven by the magnitude of the change in those who do improve, and we don’t really see much about what is happening to those who are doing worse.”
Of over 215,000 records of knee replacement collected from all patients undergoing TKR in England during 2008-2012, Dr. Pinedo-Villanueva and his study coauthors found 117,844 had data on pre- and post-operative knee pain assessed using the Oxford Knee Score (OKS) and quality of life measured with the EQ-5D instrument. The majority of replacements were in women (55%) and almost three quarters of patients had one or no comorbidities. Overall, the mean change in OKS was 15 points, improving from 19 to 34 (the higher the score the lesser the knee pain). EQ-5D scores also improved by a mean difference of 0.30 (from 0.41 to 0.70 where 1.0 is perfect health). Although the vast majority of patients had improved OKS and EQ-5D scores after surgery, unchanged or decreased scores were seen in 8% and 22% of patients, respectively.
“As we breakdown these data by deciles of baseline pain and function we see clearly that those starting at the lower decile improved the most, and that’s to be expected; they’ve a lot more to improve than the ones that came into the operation at the higher decile,” Dr. Pinedo-Villanueva said at the Congress, sponsored by the Osteoarthritis Research Society International. But there were patients who fared worse at every decile, he noted.
Dr. Pinedo-Villanueva concluded that outcome prediction models were needed to try to reduce the number of patients who are apparently worse off after knee replacement and improve the efficiency of resource allocation.
The value of TKR in a contemporary U.S. population was the focus of a separate presentation by Dr. Bart Ferket of Mount Sinai Hospital in New York. Dr. Ferket reported the results of a study looking at the impact of TKR on patients’ quality of life, lifetime costs, and quality-adjusted life years (QALYs) while varying the use of TKR by patients’ functional status at baseline.
“In the United States, the rate at which total knee replacement is performed has doubled in the last two decades,” Dr. Ferket observed. This “disproportionate” increase has been attributed to expanding the eligibility criteria to include less symptomatic patients.
Using data collected over an 8-year period on 1,327 participants from the Osteoarthritis Initiative, Dr. Ferket and his associates at Mount Sinai and Erasmus University Medical Center in Rotterdam (The Netherlands) discovered that the increased uptake of TKR might have affected the likely benefit and reduced the overall cost-effectiveness of the procedure.
At baseline, 17% of the participants, who all had knee osteoarthritis, had had a prior knee replacement.
Quality of life measured on the physical component scores (PCS) of the 12-item Short Form (SF-12) were generally improved after TKR but decreased in those who did not have a knee replaced. The effect on the mental component of the SF-12 was less clear, with possibly a decrease seen in some patients. Changes on the Western Ontario and McMaster Universities Arthritis Index (WOMAC) and Knee injury and Osteoarthritis Outcome Score (KOOS) showed a considerable benefit for knee replacement and there was a general decrease in pain medication over time in those who had surgery. The overall effect was more pronounced if patients with greater baseline symptoms were considered.