Diabetes screening may not be cost-effective unless treatment is highly effective.

“The cost-effectiveness of screening for diabetes is unknown but has been modeled previously. None of these models has taken account of uncertainty. We aimed to describe these uncertainties in a model where the outcome was CHD risk,” scientists in England report.

C. Glumer and colleagues working with Elsie Widdowson Laboratories explained, “Our model used population data from the Danish Inter99 study, and simulations were run in a theoretical population of 1,000,000 individuals. CHD risk was estimated using the U.K.

Prospective Diabetes Study (UKPDS) risk engine, and risk reduction from published randomised clinical trials. Probabilistic sensitivity analysis was used to provide confidence intervals for modeled outputs. Uncertain parameter values were independently simulated from distributions derived from existing literature and deterministic sensitivity analysis performed using multiple model runs under different strategy choices and using extreme parameter estimates.

The researchers determined, “In the least conservative model (low costs and multiplicative risk reduction for combined treatments), the 95% confidence interval of the incremental cost-effectiveness ratio varied from UKP 23,300-82,000. The major contributors to this uncertainty were treatment risk reduction model parameters: the risk reduction for hypertension treatment and UKPDS risk model intercept. Overall cost-effectiveness ratio was not sensitive to decisions about which groups to screen, nor the costs of screening or treatment.

It was strongly affected by assumptions about how treatments combine to reduce risk.”

“Our model suggests,” concluded the authors, “that there is considerable uncertainty about whether or not screening for diabetes would be cost-effective. The most important but uncertain parameter is the effect of treatment. In addition to directly influencing current policy decisions, health care modeling can identify important unknown or uncertain parameters that may be the target of future research.”

Glumer and colleagues published their study in Diabetologia (What determines the cost-effectiveness of diabetes screening? Diabetologia, 2006;49(7):1536-1544).

For more information, contact N.J. Wareham, Medical Research Council, Epidemiology Unit, Elsie Widdowson Laboratories, 120 Fulbourn Rd., Cambridge CB1 9NL, England.

Publisher contact information for the journal Diabetologia is: Springer, 233 Spring Street, New York, NY 10013, USA.