Phentermine-topiramate and Tirzepatide Offer High Cost-Effectiveness for Obesity Treatment; Liraglutide Shows Low Value

The global prevalence of overweight and obesity continues to rise, significantly increasing the risk of serious comorbidities like type 2 diabetes, coronary heart disease, and stroke, ultimately impacting mortality. While various pharmacologic treatments have demonstrated efficacy in weight management, a comprehensive understanding of their cost-effectiveness in a U.S. setting has been lacking. This gap is crucial for informing clinical guidelines, payer decisions, and patient access, ensuring that effective treatments are also economically viable in a healthcare system grappling with rising costs.

Researchers conducted a systematic review of non-industry-sponsored U.S. trial-based and model-based cost-effectiveness evaluations of pharmacologic treatments for adults with overweight or obesity. Data from MEDLINE, Embase, and economic databases were searched up to October 13, 2025. Study quality was assessed using the CHEQUE tool, and certainty of evidence via the GRADE approach. Value was determined by incremental cost-effectiveness ratios (ICERs), with thresholds: high value (<$100,000 per QALY), intermediate ($100,000 to $200,000 per QALY), and low value (>$200,000 per QALY).

Out of 9 included studies, only 4 were at low risk of bias, and none of the 42 reported pairwise comparisons had high certainty of evidence. In the 6 studies with moderate certainty, specific cost-effectiveness profiles emerged. When compared with lifestyle modification, liraglutide consistently demonstrated low value. In contrast, phentermine-topiramate and tirzepatide were found to have high value against lifestyle modification. The review also highlighted comparative cost-effectiveness among active treatments:

Semaglutide showed low value when compared with naltrexone-bupropion and phentermine-topiramate, but demonstrated high value when compared directly with liraglutide. Overall, the evidence base for cost-effectiveness in this area is hampered by methodological limitations and a lack of high-certainty data.