Does the Williams Manual Test predict service use among subjects undergoing geriatric assessment?

This study examines the utility of the Williams Manual Test (WMT) in predicting those older patients who will eventually require formal services. One hundred seventeen older subjects underwent comprehensive geriatric assessment that included administration of the WMT and questionnaires assessing current service use. Study participants returned at 6 and 12 months for repeat measurement of study variables. The WMT was found to be a significant predictor of additional service use at 12 months (chi 2 = 6.4, P less than .01). The test had high specificity (83%) but relatively low sensitivity (40%). The lack of sensitivity in this study group was likely due to high levels of stress among caregivers whose relative was otherwise not predicted to need services. When the WMT was repeated after 6 and 12 months, the results were highly reproducible. The WMT appears to be a useful predictor of increasing formal service use. However, in certain study groups, its predictive ability may be affected by high levels of caregiver stress.