Response to Hoerger et al.
I read with great interest the article by Hoerger et al. regarding cost-effectiveness of screening for pre-diabetes in the U.S. setting (1). The results reported by the authors have potentially global implications and should be viewed in the context of similar analyses conducted elsewhere.
The cost-effectiveness ratio obtained in the study is in broad agreement with recent findings for the U.K. (2). The U.K. analysis was based on a pre-diabetes screening module linked to a validated diabetes model (3). The value of the incremental cost-effectiveness ratio (ICER) was determined as £23,717 for patients with both impaired glucose tolerance and impaired fasting glucose treated with intensive lifestyle intervention. The results were not sensitive to accuracy of the diagnostic tests, testing costs, and prevalence of pre-diabetes but were highly sensitive to the cost of intensive lifestyle intervention, supporting the findings by Hoerger et al. In addition, the U.K. analysis revealed that screening was more expensive because early identification of pre-diabetic patients led to increase in treatment costs. Earlier diagnosis of pre-diabetes, although leading to lower diabetes-related morbidity, also decreased mortality, which resulted in higher costs.
It would be informative to determine the source of the relatively low value of ICER from the Hoerger et al. study versus results in the U.K., where ICER was near the threshold ceiling ratio. Data sources and assumptions in both models were similar, with few exceptions. A time horizon of up to 40 years, rather than full lifespan, was chosen in the U.K. study, but with the mean patient age of 51 years the difference would be negligible. The U.K. analysis, however, was not limited to patients in the 45–75 years age bracket and included patients as young as 25 years. It would be interesting to establish if cost-effectiveness ratios decrease up to a certain age threshold and then increase for older populations, as determined by Hoerger et al.
The U.K. values of ICER are also likely to be higher than those reported by Hoerger et al. because in the U.K. model, the entire population meeting the Diabetes Prevention Program inclusion criteria was subjected to diagnostic tests, with all patients receiving both fasting plasma glucose and oral glucose tolerance tests, conditional on positive results of the former. Using initial capillary blood glucose would lead to lower values of ICER.
Also, ICER determined for intensive lifestyle intervention versus standard lifestyle advice in the U.K. population with diagnosed pre-diabetes was £12,599 per quality-adjusted life-year, considerably higher than the value of $1,100 previously reported for the U.S. (4). In addition, in the U.K. analysis, the probability of intensive lifestyle intervention being cost-effective was only 65% in probabilistic sensitivity analysis. To inform health care resource allocation, the joint uncertainty of all screening-related input variables should also be considered.
Regardless of country-specific diagnostic and treatment pathways and cost consideration, modeling studies involving intensive lifestyle modification only in the context of diabetes are likely to underestimate the degree to which screening and early detection of pre-diabetes may lead to the detection of other chronic diseases, possibly leading to long-term cost savings. Engagement of volunteer lay educators in delivery of intensive lifestyle interventions (5) would likely increase the savings even more.
- DIABETES CARE