Abstract
Purpose :
To compare low vision rehabilitation (LVR) outcomes by patient-report, performance measures and therapist ratings.
Methods :
Prior to LVR, self-reported visual ability was estimated by administering the Activity Inventory (AI), an adaptive visual function questionnaire, by telephone to new low vision (LV) patients. Baseline performance of five Timed Instrumental Activities of Daily Living (TIADL) in areas of communication, finance, shopping, cooking and managing medicine was administered to the patient in their home by a tester. A list of the patient-reported AI goals without importance and difficulty rankings were shared with the occupational therapist (OT), and after the in-home evaluation, the OT rated his/her estimate of the patient’s ability relative to each AI goal using the Functional Independence Measure (FIM). After usual LVR care and discharge, the tester re-administered the TIADL’s, the OT assigned an updated FIM score for each goal, and the AI was re-administered by phone. Rasch analysis was performed on patient AI responses, TIADL’s, and FIM ratings at baseline and follow up. AI item difficulty measures were anchored at baseline and change scores were calculated for each outcome measure. The percent of patients achieving a minimum clinically importance difference (MCID = ratio of the change in measure to 1.95 * standard error of the estimate) after LVR was calculated for each outcome measure.
Results :
Baseline findings showed moderate correlations (r=0.55) between TIADL’s and AI task difficulty for low vision patients (n=40). Baseline FIM scores and AI goal difficulty ratings (filtered vs. unfiltered) showed moderate correlations with r=0.57 and r=0.46 respectively. LVR outcomes showed effect sizes (Cohen’s d) of 0.34 for TIADL’s and 0.49 for AI patient-reported outcomes. FIM outcomes as rated by the OT showed an effect size of 3.29. Poor correlation was observed between TIADL and AI outcomes (r=0.12), and the percent of patients achieving a MCID was 38% vs. 60% respectively.
Conclusions :
Despite relatively good agreement at baseline between different functional assessments of visual ability in LV patients, there is poor agreement between these measures post-treatment. TIADL’s may underestimate LVR effects because testing a specific activity may not be relevant to the individual’s goals.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.