Practical Applications of Cost-Effective Knee Osteoarthritis and Obesity Treatments: Elena Losina, PhD

Tirzepatide was found to be more cost-effective than semaglutide when treating patients with knee osteoarthritis and obesity, according to a recent study published in the Annals of Internal Medicine.

The American Journal of Managed Care^® spoke with corresponding author Elena Losina, PhD, codirector of the Orthopedics and Arthritis Center for Outcomes Research at Brigham and Women’s Hospital, to discuss the study methods and practical implications of the results.

Tirzepatide had an incremental cost-effectiveness ratio of $57,400 per lifetime quality-adjusted life-year (QALY), significantly less than that of semaglutide when combined with usual care (ie, diet and exercise, nonsurgical interventions), but greater than the cost-effectiveness for bariatric surgery.

These data, Losina said, are pivotal for policy makers when making clinical decisions for coverage; however, cost-effectiveness for knee osteoarthritis requires longer studies and larger sample sizes to fully understand the full scope of a cost-effective treatment.

The transcript has been lightly edited. Captions are auto-generated.

What are the biggest strengths and limitations of modeling vs real-world data?

Modeling is a wonderful complementary tool to use alongside clinical trials, because we recognize that clinical trials do provide the highest level of evidence, but treatment decisions often occur in the absence of results from clinical trials. Modeling helps to combine and aggregate data from various reliable sources and national sources to project outcomes in situations where clinical trials are not available.

For example, knee osteoarthritis is a chronic condition, and it lasts for about 26 years. Most clinical trials last for a much shorter period of time. It's important to consider sequential regimens and understand how each treatment fits within the sequential treatment for chronic diseases. That's where modeling comes in handy. Models have been developed based on clinical expertise; they are carefully trying to replicate using the computer algorithm language what’s happening with human beings.

We do extensive validation of such models, and what they do is help to extend beyond the time horizons and availability of results from clinical trials to assist policy decisions, policymakers, and clinicians with clinical decisions, because (A) we live in the era of limited resources, and those decisions need to be made, and (B) perfect data will never be available, but physicians are faced with clinical decisions on a daily basis, so that's why modeling can help complement results of clinical studies to provide more information and more insights into clinical and policy decisions.

Given that tirzepatide had a 64% chance and semaglutide a 34% chance of being effective at $100,000/QALY, what research questions remain unanswered?

The data that we use in the model and the data that are coming out from clinical studies all have a certain level of uncertainty because they’re based on samples. In the decision analysis modeling, we use sensitivity analysis, trying to take into consideration all uncertainty in the data. Based on this uncertainty, we found that if we vary the simultaneous multiple parameters used for our analysis to address the uncertainty, then tirzepatide will be cost-effective in 64% of cases. That means there is a substantial amount of uncertainty that still exists. We would like to kind of get that number as high as 95%. Then, I think, the decisions are more statistically sound. But what that says is we are looking for more data and less uncertainty, especially as it relates to the long-term durability and effectiveness of these medications.

Aside from cost-effectiveness, what patient-centered outcomes (like functional improvement or avoidance of joint replacement) should future research prioritize?

People taking medications or people with knee osteoarthritis and obesity should consider taking these medications to improve their quality of life. They would like to be more mobile, they would like to participate in social life, and so if these medications are also accompanied by some unpleasant adverse events, then the benefits, obviously, are being dampened. I think that the ratio between the benefits and adverse events is something patients are looking at when they are considering taking these medications. Of course, there's the access and the cost too, because if insurance is not covering it, the likelihood that people can pay out of pocket about $800 per month is probably not very realistic.

Reference

Betensky DJ, Smith KC, Katz JN, et al. The cost-effectiveness of semaglutide and tirzepatide for patients with knee osteoarthritis and obesity. Ann Intern Med. Published online September 16, 2025. doi:10.7326/ANNALS-24-03609