Utilizing the assessment of motor and process skills ability measures to predict level of community dependence

Objective. To determine if ADL motor and ADL process ability measures can be used to predict level of community dependence.;Participants. Potential participants included all available people in the Assessment of Motor and Process Skills (AMPS) database who (a) had been assessed after January 2000, (b) had not previously been associated with rater scoring error as evidenced by artificially high ADL motor or ADL process ability measures, (c) were not scored by more than 10 raters during rater calibration, and (d) did not have missing or unknown functional level ratings.;Method. Receiver operating characteristic (ROC) curves were generated to (a) verify the accuracy of the current AMPS cutoff measures, (b) determine if AMPS ability measures can be used to determine level of community dependence, and (c) globally examine the relative predictive validity of AMPS ability measures across different diagnostic groups.;Results/Conclusion. Results indicate both the ADL motor and ADL process ability measures can be utilized to help occupational therapists determine level of community dependence. Proposed ADL motor and ADL process cutoff measures demarcating independence were 1.50 and 1.10 logits respectively, and proposed ADL motor and ADL process cutoff measures demarcating the need for moderate to maximal assistance were 1.00 and 0.70 logits respectively. The ability measures between the proposed cutoff measures demarcate a risk zone. For all diagnostic groups the estimates of the area under the ROC curve (global estimates of predictive accuracy) indicated that ADL process ability measure is a fair to good indicator of community dependence. Area under the curve estimates for ADL motor ability were fair to good for all diagnostic groups, except for the psychiatric group, where the area under the curve estimate indicated that ADL motor ability is a poor indicator of community dependence.