Utilizing AMPS ability measures to predict level of community dependence.

The aim of the study was to examine the efficacy of utilizing Assessment of Motor and Process Skills (AMPS) ability measures as evidence for determining level of community dependence. The sample for the study was drawn from existing data from the AMPS International Project database, and consisted of adults from numerous world regions, with varying diagnoses and levels of ability (n=64,466). Findings support using ADL motor and process ability measures as evidence of a client's level of community dependence. When using ADL motor or process ability measures to identify the need for moderate to maximal assistance, the areas under the ROC curves were 0.74 (fair discrimination power) and 0.82 (good discrimination power), respectively. New ADL motor and process cutoff measures demarcating maximum assistance were set at 1.00 logit (sensitivity=0.70, specificity=0.66) and 0.70 logit (sensitivity=0.79, specificity=0.69), respectively. While ADL process ability continues to be the better predictor, the highest accuracy estimates occurred when motor and process decisions matched (n=30,835). For the first time, multiple AMPS cutoff measures, demarcating independent in the community and in need of maximal assistance, have been proposed.