PS 1-7
SYSTEMATIC OVERVIEW OF DECISION-ANALYTIC MODELING STUDIES IN IODINE DEFICIENCY RELATED DISEASES
Iodine deficiency related disease prevention and treatment remain an important public health problem. Iodine deficiency can have severe health consequences, such as cretinism, goiter, hypothyroidism or thyroid nodules as well as economic implications. Our aim was to give an overview on studies applying a decision-analytic model to evaluate the effectiveness and/or cost-effectiveness of iodine deficiency related treatments, prevention strategies or diagnostic procedures.
Method:
We performed a comprehensive systematic literature search in Pubmed/Medline, Tuft’s CEA Registry, NHS EED and Embase to identify published studies comparing different treatment, prevention or diagnostic strategies with a mathematical decision-analytic model. Studies were required to evaluate patient-relevant health outcomes (e.g., remaining life years, quality-adjusted life years). All studies were screened by at least two independent reviewers and a third person (modeling or clinical expert) decided on inclusion if there was disagreement. We completed a comprehensive evidence table providing an overview on study characteristics (e.g., research question, target population, time horizon, health outcomes, analytic approach, perspective).
Result:
Overall, we found 2241 studies. After removal of duplicates, abstract and title screening, as well as fulltext screening, 16 studies remained included. Ten studies evaluated screening programs (mainly newborns and pregnant women), five studies focused on treatment approaches (Graves’ disease, thyroid nodules) and one study was about a primary prevention (consequences of iodine supplementation on offspring) related to iodine deficiency. Most of the studies were conducted within the U.S. health-care context (n=7), two for Iran and France, and one study for the Czech Republic, Germany, the Netherlands and the U.K. In six studies, a Markov state-transition model was applied and in eight studies a decision tree. The remaining studies did not explicitly state their modeling approach. The analytical time horizon ranged from one year to lifetime. 75% of the studies evaluated quality-adjusted life years as health outcome measure. In all studies, a cost-effectiveness analysis was performed. None of the models compared the same research question for the same health care context; therefore, no comparison of the models was possible. None of the models reported a formal model validation.
Conclusion:
Overall, we mostly found decision-analytic modeling studies evaluating specific screening programs or treatment approaches; however, there was no model evaluating primary prevention programs on a population basis.