It is hard to predict your future

Choosing individual health insurance, in the best of situations, is HARD!. It involves navigating a complex bureaucracy that attempts to hide through complex jargon key concepts. It involves digging through dozens if not hundreds of choices on the marketplaces or Medicare Advantage/Part D. It involves trying to predict the future for both technological shocks that may be relevant, policy shocks, and personal health status.

Dr. Abigail Barker et al, ** test several personal predictive models. These models are trying to predict how accurate a first year prediction of personal health status is in predicting second year healthcare expenditures. To me, this is an awesome question as we sort of assume that health is mostly continuous and reasonably predictable from the recent pass unless there is a huge but rare shock event.

Self-reported health status (SHS), whereby patients rate their own health, could improve cost-prediction estimates without requiring individuals to share personal health information or know about undiagnosed conditions. We compared the predictive accuracy of several models: (1) SHS only, (2) a “basic” model adding health-related variables, and (3) a “full” model adding measures of healthcare access. The Medical Expenditure Panel Survey was used to predict 2015 health expenditures from 2014 data….

The full model, which added race and socioeconomic variables, as well as access variables, yielded an adjusted R2 of 0.290 and root MSE of 1.507….

This study found that SHS alone was not a strong predictor of medical expenditures on an individual level, despite strong correlations between better health and lower expenditures at the aggregate level.

I found the last sentence to be fascinating and critical. Self reported health status is a pretty decent metric when applied to a large group in the attempt to predict group costs. It is a MEH metric when applied at the individual level. Adding in some clinical factors improved the model fit but even still, it still only explains 29% of the variance (which is not bad… but not great).

Picking insurance is hard and the idea that self reported health status is an easy hook into reliable decision-support tools to improve plan choice is attractive but likely to lead us down a thorn bracketed path.

** Jan-Dec 2021;58:469580211064118. doi: 10.1177/00469580211064118

The post It is hard to predict your future appeared first on Balloon Juice.