Falls among the elderly population cause detrimental physical, mental, financial problems and, in the worst case, death. The increasing number of people entering the higher risk age-range has increased clinicians’ attention to intervene. Clinical tools, e.g., the Timed Up and Go (TUG) test, have been created for aiding clinicians in fall-risk assessment. Often simple to evaluate, these assessments are subject to a clinician’s judgment. Wearable sensor data with machine learning algorithms were introduced as an alternative to precisely quantify ambulatory kinematics and predict prospective falls. However, they require a long-term evaluation of large samples of subjects’ locomotion and complex feature engineering of sensor kinematics. Therefore, it is critical to build an objective fall-risk detection model that can efficiently measure biometric risk factors with minimal costs. We built and studied a sensor data-driven convolutional neural network model to predict older adults’ fall-risk status with relatively high sensitivity to geriatrician’s expert assessment. The sample in this study is representative of older patients with multiple co-morbidity seen in daily medical practice. Three non-intrusive wearable sensors were used to measure participants’ gait kinematics during the TUG test. This data collection ensured convenient capture of various gait impairment aspects at different body locations.

Background Older patients are at increased risk of falling and of serious morbidity and mortality resulting from falls. The ability to accurately identify older patients at increased fall risk affords the opportunity to implement interventions to reduce morbidity and mortality. Geriatricians are trained to assess older patients for fall risk. If geriatricians can accurately predict fallers (as opposed to evaluating for individual risk factors for falling), more aggressive and earlier interventions could be employed to reduce falls in older adult fallers. However, there is paucity of knowledge regarding the accuracy of geriatrician fall risk predictions. This study aims to determine the accuracy of geriatricians in predicting falls. Methods Between October 2018 and November 2019, a convenience sample of 100 subjects was recruited from an academic geriatric clinic population seeking routine medical care. Subjects performed a series of gait and balance assessments, answered the Stay Independent Brochure and were surveyed about fall incidence 6–12 months after study entry. Five geriatricians, blinded to subjects and fall outcomes, were provided the subjects’ data and asked to categorize each as a faller or non-faller. No requirements were imposed on the geriatricians’ use of the available data. These predictions were compared to predictions of an examining geriatrician who performed the assessments and to fall outcomes reported by subjects. Results Kappa values for the 5 geriatricians who used all the available data to classify participants as fallers or non-fallers compared with the examining geriatrician were 0.42 to 0.59, indicating moderate agreement. Compared to screening tools’ mean accuracy of 66.6% (59.6–73.0%), the 5 geriatricians had a mean accuracy for fall prediction of 67.4% (57.3–71.9%). Conclusions This study adds to the scant knowledge available in the medical literature regarding the abilities of geriatricians to accurately predict falls in older patients. Studies are needed to characterize how geriatrician assessments of fall risk compare to standardized assessment tools.

No abstract

Background: Older patients are at increased risk of falling and of serious morbidity and mortality resulting from falls. The ability to accurately identify older patients at increased fall risk affords the opportunity to implement interventions to reduce morbidity and mortality. Geriatricians are trained to assess older patients for fall risk. If geriatricians can accurately predict fallers (as opposed to evaluating for individual risk factors for falling), more aggressive and earlier interventions could be employed to reduce falls in older adult fallers. However, there is paucity of knowledge regarding the accuracy of geriatrician fall risk predictions. This study aims to determine the accuracy of geriatricians in predicting falls. Methods: Between October 2018 and November 2019, a convenience sample of 100 subjects was recruited from an academic geriatric clinic population seeking routine medical care. Subjects performed a series of gait and balance assessments, answered the Stay Independent Brochure and were surveyed about fall incidence 6-12 months after study entry. Five geriatricians, blinded to subjects and fall outcomes, were provided the subjects’ data and asked to categorize each as a faller or non-faller. No requirements were imposed on the geriatricians’ use of the available data. These predictions were compared to predictions of an examining geriatrician who performed the assessments and to fall outcomes reported by subjects. Results: Kappa values for the 5 geriatricians who used all the available data to classify participants as fallers or non-fallers compared with the examining geriatrician were 0.42 to 0.59, indicating moderate agreement. Compared to screening tools’ mean accuracy of 66.6% (59.6-73.0%), the 5 geriatricians had a mean accuracy for fall prediction of 67.4% (57.3-71.9%).Conclusions: This study adds to the scant knowledge available in the medical literature regarding the abilities of geriatricians to accurately predict falls in older patients. Studies are needed to characterize how geriatrician assessments of fall risk compare to standardized assessment tools.