A systematic review of methods to predict weight trajectories in health economic models of behavioral weight management programs : the potential role of psychosocial factors

Objectives. There is limited evidence on the long-term effectiveness of behavioral weight management interventions, and thus, when conducting health economic modeling, assumptions are made about weight trajectories. The aims of this review were to examine these assumptions made about weight trajectories, the evidence sources used to justify them, and the impact of assumptions on estimated cost-effectiveness. Given the evidence that some psychosocial variables are associated with weight-loss trajectories, we also aimed to examine the extent to which psychosocial variables have been used to estimate weight trajectories and whether psychosocial variables were measured within cited evidence sources.

Methods. A search of databases (Medline, PubMed, Cochrane, NHS Economic Evaluation, Embase, PSYCinfo, CINAHL, EconLit) was conducted using keywords related to overweight, weight management, and economic evaluation. Economic evaluations of weight management interventions that included modeling beyond trial data were included.

Results. Within the 38 eligible articles, 6 types of assumptions were reported (weight loss maintained, weight loss regained immediately, linear weight regain, subgroup-specific trajectories, exponential decay of effect, maintenance followed by regain). Fifteen articles cited at least 1 evidence source to support the assumption reported. The assumption used affected the assessment of cost-effectiveness in 9 of the 19 studies that tested this in sensitivity analyses. None of the articles reported using psychosocial factors to estimate weight trajectories. However, psychosocial factors were measured in evidence sources cited by 11 health economic models.

Conclusions. Given the range of weight trajectories reported and the potential impact on funding decisions, further research is warranted to investigate how psychosocial variables measured in trials can be used within health economic models to simulate heterogeneous weight trajectories and potentially improve the accuracy of cost-effectiveness estimates.