You can check how good a heart disease computer model is by seeing if it predicts group death rates well and also if it correctly ranks who’s most likely to die.
Claim Context
Micro-simulation models of cardiovascular disease can be externally validated using individual-level longitudinal data by comparing both population-level survival curves and individual-level risk discrimination via ROC analysis, a method not commonly applied to disease simulation models but which provides complementary insights into model performance.
“We validated our micro-simulation model using recent National Health and Nutrition Examination Survey (NHANES) data... using survival curves and, in a method not typically used for disease model validation, receiver operating characteristic (ROC) curves.”
Evidence from Studies
No evidence studies found yet.
What Would Prove This
Per GRADE and EBM methodology, here is what ideal scientific evidence would look like to definitively prove or disprove this claim, ordered from strongest to weakest.
Whether ROC-based validation improves decision-making compared to traditional methods.
A systematic review of CVD model validation studies that evaluates whether models validated with both survival curves and ROC analysis lead to more accurate policy recommendations than those using only one method.
Whether policies based on ROC-validated models improve outcomes.
A cluster-randomized trial assigning 30 health departments to use either ROC-validated or non-ROC-validated CVD models for prevention planning, measuring 5-year changes in CVD mortality and resource allocation efficiency.
Expert consensus on best practices for model validation.
A Delphi survey of 30 health economists and modelers to reach consensus on whether ROC analysis should be required for CVD model validation.