Exercise, Education to Relieve Knee OA: Mixed Results in Modeling Study

Compared with usual care, an Australian modeling study found that a structured program of education and exercise for people with osteoarthritis (OA) of the knee wasn’t cost-effective over the long term across the entire population, though it did seem to be for those with little to no pain initially.

The hypothetical program, efficacy and costs of which were based on real-life registry data from Australia, was estimated to save the average OA patient A$7,970 (U.S. $5,537) over a lifetime, but it actually led to 0.43 fewer quality-adjusted life-years (QALY) relative to usual care in which patients would proceed more quickly to total knee replacement, according to Sean Docking, PhD, of Monash University in Melbourne, Australia, and colleagues.

But the group’s modeling exercise, reported in JAMA Network Open, also indicated that the education-and-exercise program would be cost-effective across the population for the first 9 years. And it was cost-effective over a lifetime for OA patients when initiated early in their disease trajectory.

“These findings point to opportunities to invest early cost savings in additional care or prevention, including targeted implementation to specific subgroups,” Docking and colleagues concluded.

In Australia, like the U.S., such programs for OA are “poorly funded” by private and public insurers, the group explained. But if cost-effectiveness could be shown, it might boost prospects for increasing financial support for nondrug and nonsurgical interventions that might stave off or at least delay knee replacement.

The model incorporated data on costs and outcomes for such interventions in knee OA as have been implemented in a more piecemeal fashion in Australia, as well as for patients undergoing knee replacement. Effectiveness was expressed in terms of probabilities for “transition” between four health states — little or no pain, moderate pain, severe pain, and death — and the likelihood that a patient receiving the nonsurgical intervention would nevertheless need knee replacement. Docking and colleagues then assembled a sample of 61,394 hypothetical patients, also based on real-world data, with varying ages and degrees of knee pain at baseline. The model then estimated progression over time for each individual given assignment to the nonsurgical program versus usual care.

Just under 20% of patients receiving the program were estimated to avoid knee replacement: if all patients received it, there would be 11,995 fewer primary surgeries and 2,423 fewer revisions.

But this benefit was offset by larger QALY gains with usual care, in which patients proceeded quickly to knee replacement. Cost-effectiveness for the nonsurgical program was evident over a lifetime only in patients with no or mild baseline pain.

Baseline age greater than 75 was associated with the greatest cost savings, but that was an artifact of the model, Docking and colleagues noted: those patients were more likely to die before requiring knee replacement. Usual care was also found to be more cost-effective at the hospital level, at both public and private facilities.

One takeaway from the study, the researchers suggested, is that “[p]rioritizing nonsurgical management of knee osteoarthritis to patients for whom [total knee replacement] is inappropriate may be a more efficient strategy” than pushing it to everyone. Docking and colleagues observed that tools are available to identify those patients although their cost-effectiveness hasn’t been evaluated.

“Future efforts are needed to test the cost-effectiveness of valid and reliable appropriateness criteria that enable the right patient to receive the right care at the right time,” they urged.

Limitations to the study included its use of Australian data, which may not generalize well to the U.S. or other nations, as well as the usual cautions around modeling studies and their underlying assumptions. As well, outcomes could vary substantially with the particulars of the education-and-exercise program.

Please enable JavaScript to view the

comments powered by Disqus.