Content-based interpretation aids for health-related quality of life measures in clinical practice. An example for the visual function index (VF-14)

Background: In spite of a well-established development of instruments, difficulty in interpreting health related quality of life scores may limit its use in clinical practice. Objective: To develop generalizable interpretation aids for a measure of perceived functional visual status, the VF-14 index. Design: Item Response Theory (Rasch analysis) was used to analyze the performance of VF-14 items. The 'ruler' aid was derived from the most difficult activity (item) a patient is able to do without difficulty; the 'clinical scenarios' aid, first identified all significantly different clusters of items within the index and then estimated the mean expected difficulty (responses) to perform a benchmark item in each cluster. Setting: The study was conducted in four hospitals and six ambulatory cataract surgery centers in Barcelona, Spain. Patients: One hundred and ninety-eight patients scheduled for first eye cataracts surgery. Measurements: The self-reported VF-14 index and clinical measures were used. Results: All VF-14 items were found unidimensional with three items showing only partial misfit. For a patient with a VF-14 Rasch score of 71, the 'ruler' aid indicated that 'doing fine handwork' would be the most requiring activity he/she would perform without difficulty. The 'clinical scenarios' aid estimated that such a patient would be unable to drive at night', would have some difficulty 'reading small print' and no difficulty 'doing fine handwork', 'watching TV' or 'recognizing people'. Concordance between modeled and observed responses was fair to substantial. Conclusions: Simple content-based interpretation aids for the VF-14 scores were developed that should facilitate its use in clinical practice. These aids should be easily generalizable to other quality of life instruments.