Physician estimates as predictors of elderly patients' mobility scores

In a previous study, residency physicians' estimates of patients' mobility scores were strongly associated with the patients' actual scores. The present study tested the hypothesis that community physicians would similarly be able to predict patients' mobility scores. Dysmobility risk-factor questionnaires were given during routine office visits to 120 ambulatory and mentally competent patients older than age 65. After the visit, the attending physician estimated the patient's mobility score (PEMS), and the patient independently underwent mobility testing. The PEMS was 71% sensitive and 80% specific in selecting those patients who had a mobility score of less than 30. Choosing patients with low PEMS or high age (75+) increased sensitivity (96%) while decreasing specificity (53%). Alternatively, selecting only patients over age 75 with low PEMS or problems walking had a sensitivity of 78% and specificity of 86%. This study suggests that by using pretests, the physician can exclude many elderly patients from mobility testing while detecting most of those who are dysmobile.