We propose to conduct a multi-therapist, prospective, observational study of the outcomes of usual low vision rehabilitation (LVR) services. The purpose of this project is to study the concordance of three different types of LVR outcome measures (patient-reported, therapist-reported, and functional performance), study the influence of therapists'biases on estimated outcome measures, study the effects of co-morbidities and other patient states on LVR outcomes, and study the relationship of the effectiveness of LVR to its efficacy.
We aim to recruit 10 LVR therapists as clinical collaborators. Each therapist will be expected to recruit 45 patients over a period of 4 years or less. The therapists will provide their usual low vision rehabilitation services between baseline and follow-up outcome measures. Research assistants will interview patients by telephone and administer the adaptive Activity Inventory (AI), which requires patients to rate the importance and difficulty of activity goals and rate the difficulty of subsidiary tasks under each important ad difficult goal, before and after LVR. The therapists will generate Goal Attainment Scale (GAS) ratings of patients'activity goals and administer a battery of functional performance tests before and after LVR. We have developed a measurement model that will be used to estimate interval-scaled measures of patients'functional abilities in different functional domains from patients'difficulty ratings, therapists'GAS ratings, and functional performance measures. Changes after LVR in patient-reported functional ability measures estimated from patients'difficulty ratings of AI goals will be compared to changes after LVR in functional ability measures estimated from therapists'ratings of the same AI goals using the GAS method. We also will compare changes after LVR in patient-reported measures of functional ability (estimated from patients'difficulty ratings of the appropriate subset of AI tasks) to changes after LVR in corresponding performance measure estimates of functional ability (reading speed versus print size measures for reading function, walking speed measures for mobility function, measures of maze tracing errors for visual motor function, and measures of facial expression recognition probability correct for visual information processing function). The objective of the proposed research is to understand sources of bias in, and effects of confounding variables on, the three different types of LVR outcome measures that can be used in clinical research on LVR.
Low vision rehabilitation services are chosen to meet the specific needs and capabilities of each individual patient. Low vision rehabilitation therapists begin by evaluating the patient's functional ability, identifying daily activities that are importat to the patient but are difficult or impossible to perform, and setting and prioritizing rehabilitation goals. Successful rehabilitation is defined by the achievement of rehabilitation goals. Thus, the definition of a successful low vision rehabilitation outcome can be different for each patient. This highly personalized nature of judging outcomes creates a challenge for conducting rigorous clinical research on the effectiveness of low vision rehabilitation. To complicate matters even more, especially for high quality research required by evidence-based practices, low vision rehabilitation therapists have the opportunity to bias the outcome by choosing easily achieved goals that might not address the patient's most important issues, or by underestimating the patient's functional ability at the initial evaluation and overestimating the patient's functional ability upon completion of rehabilitation. The purpose of the proposed research is to apply and test a new measurement theory that enables us to determine therapists'biases in their choice of rehabilitation goals and biases in their estimates of patients'functional abilities. The proposed research will compare low vision rehabilitation outcomes measured from the patient's perspective, from the therapist's perspective, and with objective performance measures. Besides a deeper scientific understanding of outcome measures, if the theory proves successful, we will have powerful new tools for rigorously measuring personalized rehabilitation outcomes and for controlling bias in future clinical studies of the effectiveness of low vision rehabilitation.
|Massof, Robert W; Stelmack, Joan A (2013) Interpretation of low-vision rehabilitation outcome measures. Optom Vis Sci 90:788-98|