Cost-utility analyses in diabetes: a systematic review and implications from real-world evidence

Objectives: To systematically review the cost-effectiveness of diabetes interventions, identify high-value diabetes services, and estimate potential gains from increasing their utilization.

Methods: The study consisted of two steps. First, we reviewed cost-utility analyses (CUAs) related to diabetes published through the end of 2012, using the Tufts Medical Center Cost-Effectiveness Analysis Registry (www.cearegistry.org). We used logistic regression to examine factors independently associated with favorable cost-effective ratios. Second, we used the Humedica electronic medical records to estimate potential savings and health benefits gained by shifting patients currently receiving low-value services to high-value alternatives.

Results: We identified 196 diabetes CUAs, of which 55% examined pharmaceuticals. Most (70%) diabetes CUAs focused on treatment rather than prevention. Most used a health care payer perspective and were industry-sponsored. Of the 497 published cost-utility ratios, 82% examined an intervention recommended by diabetes guidelines. Approximately 73% of the interventions were cost-saving or below $50,000 per quality-adjusted life-year. Logistic regression analysis showed that higher-quality CUAs, CUAs conducted from the US perspective, surgical interventions, and guideline-recommended interventions were more likely to report favorable ratios. Of the 7907 eligible patients with diabetes in our sample, up to 7117 could in principle be shifted to cost-saving treatments, reducing costs by $12.5 million and gaining more than 1938 quality-adjusted life-years over a lifetime.

Conclusions: Most diabetes interventions evaluated by CUAs are recommended by practice guidelines and may provide good value for money. Our results indicate that patients with diabetes and the health care system could potentially benefit from shifting to the greater use of high-value services.