Predicting elderly patients' mobility using fall history and physician assessment.

This study tested whether a patient's history of falls or an office clinician's general assessment can predict which ambulatory elderly patients will do poorly on mobility testing. Ambulatory patients making routine visits who were age 65 or older, mentally competent, and not acutely ill were eligible. Fifty-two (91%) of these patients participated by completing a fall history questionnaire and undergoing mobility testing. After the visit, the attending physician estimated how the patient would score on the mobility test. Twelve (23%) of the patients reported falls and seven (13.5%) reported fall injuries in the preceding year. Both fall history and physician estimate of mobility score significantly correlated with the measured mobility score. Only the physician estimate of mobility score, however, had adequate sensitivity (94.4%) and specificity (82.4%) to be clinically useful. This estimate will allow the clinician to selectively apply mobility testing to those patients likely to score poorly.