Does adding MS-specific items to a generic measure (the SF-36) improve measurement?

Objective: The 36-item Short Form Health Survey Questionnaire (SF-36) is a widely used generic health status measure. Recently it has been adapted to produce a disease-specific measure for MS-the 54-item Multiple Sclerosis Qu ality of Life Scale (MSQOL-54)-composed of five unchanged SF-36 scales; thr ee altered SF-36 scales (one item added to each scale); and five new scales incorporating 15 additional items. This study evaluates the impact of thes e additions by comparing the measurement properties of the MSQOL-54 with th e SF-36. Methods: A total of 150 patients with MS, representing a broad spe ctrum of disease severity, completed a range of questionnaires, which inclu ded the MSQOL-54 (from which the SF-36 score was computed). Of these, 44 pe ople completed the measures before and after inpatient rehabilitation to ev aluate responsiveness. Standard psychometric methods were used to evaluate the measurement properties. Results: The measurement properties of the unch anged scales, inevitably, remain identical. Those of the three altered scal es are virtually identical. Of the five new scales, the validity of the two sexual scales is questioned because of the high percentage of missing data , and the validity of the overall quality-of-life scale is limited as demon strated by the low to moderate correlations with other related and unrelate d measures. Responsiveness of the new scales also appears limited. Conclusi on: Modifying existing measures by simply adding clinically chosen items ma y not be as useful. as anticipated in improving the measurement properties of an instrument.