Adjusting Population Risk for Functional Health Status.

Abstract

Risk adjustment accounts for differences in population mix by reducing the likelihood of enrollee selection by managed care plans and providing a correction to otherwise biased reporting of provider or plan performance. Functional health status is not routinely included within risk-adjustment methods, but is believed by many to be a significant enhancement to risk adjustment for complex enrollees and patients. In this analysis a standardized measure of functional health was created using 3 different source functional assessment instruments submitted to the Medicare program on condition of payment. The authors use a 5% development sample of Medicare claims from 2006 and 2007, including functional health assessments, and develop a model of functional health classification comprising 9 groups defined by the interaction of self-care, mobility, incontinence, and cognitive impairment. The 9 functional groups were used to augment Clinical Risk Groups, a diagnosis-based patient classification system, and when using a validation set of 100% of Medicare data for 2010 and 2011, this study found the use of the functional health module to improve the fit of observed enrollee cost, measured by the R(2) statistic, by 5% across all Medicare enrollees. The authors observed complex nonlinear interactions across functional health domains when constructing the model and caution that functional health status needs careful handling when used for risk adjustment. The addition of functional health status within existing risk-adjustment models has the potential to improve equitable resource allocation in the financing of care costs for more complex enrollees if handled appropriately. (Population Health Management 2016;19:136-144).

DOI: 10.1089/pop.2015.0043

Cite this paper

@article{Fuller2016AdjustingPR, title={Adjusting Population Risk for Functional Health Status.}, author={Richard Fuller and John S. Hughes and Norbert I. Goldfield}, journal={Population health management}, year={2016}, volume={19 2}, pages={136-44} }