|
Abstract:
|
The assessment of patients' functional status across the continuum of post-acute care requires a common assessment tool. Different assessment tools are implemented in different health care settings and they cannot be easily contrasted, because they require disentangling unintended differences in difficulty among the instruments from the ability of the patients. We consider equating different health assessment questionnaires as a missing data problem, and propose several Bayesian models that can be used to impute unmeasured ordinal responses. Using real data sets, we simulated missing measurements and compared our proposed approaches to existing methods for missing data imputation. We show that, for all of the estimands considered and in most of the experimental conditions that were examined, the proposed approach provides valid inferences, and generally has relatively smaller biases, and shorter interval estimate. We apply our methods to equate two instruments from post-acute care facilities for patients who suffered a stroke, and were discharged from a rehabilitation facility in 2011. Our procedures are general and can be used in other settings as well (e.g. standardized testing).
|
