scholarly journals Comparing Machine Learning Approaches for Fall Risk Assessment

2017 ◽  
Author(s):  
Joana Silva ◽  
João Madureira ◽  
Cláudia Tonelo ◽  
Daniela Baltazar ◽  
Catarina Silva ◽  
...  
Keyword(s):  
Machine Learning ◽  
Risk Assessment ◽  
Fall Risk ◽  
Learning Approaches ◽  
Fall Risk Assessment

A Machine Learning–Based Fall Risk Assessment Model for Inpatients

2021 ◽  
Vol Publish Ahead of Print ◽  
Author(s):  
Chia-Hui Liu ◽  
Ya-Han Hu ◽  
Yu-Hsiu Lin
Keyword(s):  
Machine Learning ◽  
Risk Assessment ◽  
Fall Risk ◽  
Assessment Model ◽  
Risk Assessment Model ◽  
Fall Risk Assessment

Using Electronic Health Records to Predict Short-Term Falls: Machine learning Applications in Senior Care Facilities (Preprint)

2021 ◽  
Author(s):  
Sepideh Shokouhi ◽  
Rahul Thapa ◽  
Anurag Garikipati ◽  
Myrna Hurtado ◽  
Gina Barnes ◽  
...  
Keyword(s):  
Machine Learning ◽  
Risk Factors ◽  
Risk Assessment ◽  
Electronic Health Records ◽  
Fall Risk ◽  
Vital Signs ◽  
Short Term ◽  
Health Records ◽  
Fall Risk Assessment ◽  
Care Facilities

BACKGROUND Evidence for the best choice of fall risk assessment in long-term care facilities is limited. Short-term fall predictions may enable the implementation of dynamic care practices that specifically address changes in individualized fall risk within senior care facilities. This can be achieved through the use of electronic health records (EHRs), which contain routinely collected information regarding the majority of known fall risk factors. OBJECTIVE We implemented machine learning algorithms that use EHR data to predict a three-month fall risk in residents from a variety of senior care facilities providing different levels of care. METHODS This retrospective study obtained EHR data (2007-2021) from Juniper communities’ proprietary database of 2,785 individuals primarily residing in skilled nursing facilities, independent living facilities, and assisted living facilities across the United States. We assessed the performances of three machine learning (ML)-based fall predictions models and the Juniper Communities fall risk assessment across these different facilities. The ML input features included vital signs and several known risk factors, such as history of fall, comorbidities, and medications. These features were identified within the EHR system based on relevant International Classification of Diseases codes, string searches, or keyword queries. Additional analyses were conducted to examine how the changes in the input features, training datasets, and prediction window affected the performance of these models. RESULTS The extreme gradient boosting (XGB) model exhibited the highest performance with an area under the receiver operating characteristic curve (AUROC) of 0.846, specificity of 0.848, and sensitivity of 0.706 while achieving the best tradeoff in balancing true positive and negative rates. The number of active medications was the most significant feature associated with fall risk, followed by a resident's number of active diseases, and several variables associated with vital signs, including diastolic blood pressure and changes in weight and respiratory rates. The combination of vital signs with traditional risk factors as input features reached a higher prediction accuracy than using either group of features alone. When reducing the prediction window to two months, the XGB model continued to exhibit the highest performance (AUROC = 0.753) in comparison to logistic regression (AUROC = 0.690), multi-layered perceptron (AUROC = 0.678) and Juniper's fall risk assessment (AUROC = 0.582). CONCLUSIONS This study provides novel insights into EHR-based features for predicting short-term fall risk in different types of care facilities. The integration of EHR data into fall prediction models, and particularly vital signs, yields a cost-effective and automated fall risk surveillance. Our XGB model uncovered the impact of a wide range of clinical and pathophysiological fall predictors across heterogenous cohorts while outperforming traditional fall risk assessments and standard ML techniques that are less compatible with EHR data. CLINICALTRIAL N/A

Fall Risk Assessment and Risk Factors of Inpatients in Armed Forces Hospital

2016 ◽  
Vol 34 (1) ◽  
pp. 42-53
Author(s):  
Kyung-Wan Seo ◽  
Jeong-Ok Lee ◽  
Sun-Young Choi ◽  
Min-Jung Park
Keyword(s):  
Risk Factors ◽  
Risk Assessment ◽  
Fall Risk ◽  
Armed Forces ◽  
Fall Risk Assessment

Fall Risk Assessment (FRA) of Korean community-dwelling elderly

2019 ◽  
Vol 39 (4) ◽  
pp. 895-902
Author(s):  
Sohee Shin ◽  
Sato Susumu
Keyword(s):  
Risk Assessment ◽  
Fall Risk ◽  
Community Dwelling ◽  
Fall Risk Assessment

Cross-cultural adaptation and validation of the Spanish version of the Johns Hopkins Fall Risk Assessment Tool

2020 ◽  
pp. 1-8
Author(s):  
Francisco José Ariza-Zafra ◽  
Rita P. Romero-Galisteo ◽  
María Ruiz-Muñoz ◽  
Antonio I. Cuesta-Vargas ◽  
Manuel González-Sánchez
Keyword(s):  
Risk Assessment ◽  
Cultural Adaptation ◽  
Fall Risk ◽  
Assessment Tool ◽  
Cross Cultural ◽  
Spanish Version ◽  
Risk Assessment Tool ◽  
Fall Risk Assessment ◽  
Cross Cultural Adaptation

Is Fall Risk Systematically Evaluated in Memory Clinics? A National Survey of Practice in France

2021 ◽  
pp. 1-9
Author(s):  
Victoire Leroy ◽  
Yaohua Chen ◽  
Naiara Demnitz ◽  
Florence Pasquier ◽  
Pierre Krolak-Salmon ◽  
...  
Keyword(s):  
Risk Assessment ◽  
Standardized Tests ◽  
Fall Risk ◽  
Routine Practice ◽  
Cross Sectional Survey ◽  
Online Questionnaire ◽  
Timed Up And Go ◽  
Major Health Problem ◽  
Fall Risk Assessment ◽  
Memory Clinics

Background: Falls are a major health problem in older persons but are still under-diagnosed and challenging to prevent. Current guidelines do not target high-risk populations, especially people living with dementia. In France, people with neurocognitive disorders are mainly referred to memory clinics (MCs). Objective: We aimed to survey the routine practice of physicians working in MCs regarding fall risk assessment. Methods: We conducted a cross-sectional survey in France from January to May 2019 among physicians working in MCs, through an anonymous online questionnaire: twenty-seven questions about the physician’s background and their practice of fall risk assessment, especially use of clinical and paraclinical tools. We compared the results according to the age and the specialty of the physician. Results: We obtained 171 responses with a majority of women (60%) and geriatricians (78%). All age classes and all French regions were represented. Most of respondents (98.8%) stated that they address gait and/or falls in outpatient clinic and 95.9%in day hospitals. When asked about how they assess fall risk, fall history (83%) and gait examination (68.4%) were the most widely used, while orthostatic hypotension (24%) and clinical standardized tests (25.7%) were less common. Among standardized tests, One-leg Balance, Timed Up and Go Test, and gait speed measurements were the most used. Geriatricians had more complete fall risk assessment than neurologists (e.g., 56%versus 13%for use of standardized tests, p <  0.0001). Conclusion: Almost all physicians addressed the question of fall in MC, but practices are widely heterogeneous. Further investigations are needed to standardize fall risk assessment in MCs.

scholarly journals Fully Automatic Fall Risk Assessment Based on a Fast Mobility Test

Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1338
Author(s):  
Wojciech Tylman ◽  
Rafał Kotas ◽  
Marek Kamiński ◽  
Paweł Marciniak ◽  
Sebastian Woźniak ◽  
...  
Keyword(s):  
Risk Assessment ◽  
Fall Risk ◽  
Low Cost ◽  
Body Movement ◽  
Sole Source ◽  
Timed Up And Go ◽  
True Negative ◽  
Fall Risk Assessment ◽  
Positive Ratio ◽  
Mobility Test

This paper presents a fall risk assessment approach based on a fast mobility test, automatically evaluated using a low-cost, scalable system for the recording and analysis of body movement. This mobility test has never before been investigated as a sole source of data for fall risk assessment. It can be performed in a very limited space and needs only minimal additional equipment, yet provides large amounts of information, as the presented system can obtain much more data than traditional observation by capturing minute details regarding body movement. The readings are provided wirelessly by one to seven low-cost micro-electro-mechanical inertial measurement units attached to the subject’s body segments. Combined with a body model, these allow segment rotations and translations to be computed and for body movements to be recreated in software. The subject can then be automatically classified by an artificial neural network based on selected values in the test, and those with an elevated risk of falls can be identified. Results obtained from a group of 40 subjects of various ages, both healthy volunteers and patients with vestibular system impairment, are presented to demonstrate the combined capabilities of the test and system. Labelling of subjects as fallers and non-fallers was performed using an objective and precise sensory organization test; it is an important novelty as this approach to subject labelling has never before been used in the design and evaluation of fall risk assessment systems. The findings show a true-positive ratio of 85% and true-negative ratio of 63% for classifying subjects as fallers or non-fallers using the introduced fast mobility test, which are noticeably better than those obtained for the long-established Timed Up and Go test.

Sign in / Sign up

Export Citation Format

Share Document