Purpose: Value of information (VOI) techniques provide estimates of the expected benefits from clinical research studies that can inform decisions about the design and priority of those studies. Most VOI studies use decision analytic models to characterize the uncertainty of the effects of interventions on health outcomes, but the complexity of constructing such models poses barriers to some practical applications of VOI. This study explores and develops methods to perform VOI by characterizing uncertainty in health outcomes with "minimal modeling".
Methods: We first develop a conceptual framework to define and classify minimal modeling approaches to VOI. Using this framework, we then review existing VOI studies that apply minimal modeling approaches. Finally, we illustrate and discusses the application of the minimal modeling to three clinical applications to which the approaches appear well suited because clinical trials with comprehensive outcomes provide preliminary estimates of the uncertainty in outcomes.
Results: Minimal modeling approaches can be divided into "limited modeling" and "no modeling" categories. A "no modeling" approach can be applied when prior clinical studies allow charcterization of patient outcomes until any differences between treatments compared are no longer present (e.g., symptom resolution, death). This makes the measured outcomes comprehensive, in the sense that they allow alternatives being compared to be ranked in terms of their net benefit at the individual level without modeling. A "limited modeling" approach can be applied when a clinical trial describes a compenhensive measure over a defined period. However, survival (to death or symptom resolution) must be modeled separately in order to value the benefits of alternative treatments that can then be easily used to produce population-level VOI estimates. We found 12 published studies applying no modeling approaches to VOI and 7 applying limited modeling approaches. We describe three new applications of VOI, one using a limited modeling approach (comaprison of atypical antipsyhoctics) and two using a no modeling approach (erlotinib and gemcitabine versus gemcitabine in pancreatic cancer, and azithromycin vs. amoxicillin/clavulanate in acute bacterial sinusitis).
Conclusion: When appropriate measures of comprehensive outcomes are available, minimal modeling approaches to VOI can be readily applied to estimate the expected benefits of clinical research.
See more of: The 33rd Annual Meeting of the Society for Medical Decision Making