Comparison of the Reliability and Validity of Fall Risk Assessment Tools in Patients with Acute Neurological Disorders

Kim, Sung Reul; Yoo, Sung Hee; Shin, Young Sun; Jeon, Ji Yoon; Kim, Jun Yoo; Kang, Su Jung; Choi, Hea Sook; Lee, Hea Limy; An, Young Hee

doi:10.7475/kjan.2013.25.1.24

Cited by 13 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…24 The sensitivity and specificity of the fall risk assessment tool were lower than those previously reported for the STRATIFY. 25 They were also lower than the results reported in a systematic review by Matarese et al 13 The positive and negative predictive values were similar to those reported in a study 17 where an MFS cutoff score of 50 was used for neurological patients. The sensitivity, specificity, and predictability of the tool were low overall, which suggests the need to develop a tool reflecting the characteristics of the clinical setting to accurately predict falls.…”

Section: Discussionsupporting

confidence: 64%

See 1 more Smart Citation

Factors Influencing Falls in High- and Low-Risk Patients in a Tertiary Hospital in Korea

et al. 2019

View full text Add to dashboard Cite

Objective The purpose of this study was to explore the characteristics and predictors of falls in high- and low-risk inpatients in a tertiary hospital in Korea. Methods Fallers' data were extracted from quality improvement reports and electronic health records from June 1, 2014, to May 31, 2015. Data on nonfallers matched by the length of hospitalization and medical departments of fallers were extracted from electronic health records. Participants were classified into a high- or a low-risk group based on their Morse Fall Scale score, fall risk–related symptoms, and medications known to increase fall risk. Characteristics of falls and risk factors were analyzed using descriptive statistics and logistic regression analysis, respectively. Results In the high-risk group, education, surgery, department, impaired mobility, intravenous catheter placement, use of ambulatory aid, gait disturbance, and some medications were significantly different between the fallers and nonfallers. From these variables, education, operation, department, intravenous catheter placement, gait disturbance, and use of narcotics, vasodilators, antiarrhythmics, and hypnotics were statistically significant factors for falls. In the low-risk group, sex, age, length of hospitalization, surgery, department, diagnosis, and mental status were significantly different between the fallers and nonfallers. From these, sex, age, length of hospitalization, surgery, and liver-digestive diseases were statistically significant factors for falls. Conclusions Characteristics and risk factors for falls differed between the risk groups. Fall prevention strategies need to be tailored to the risk groups and fall risk assessment tools need to be revised accordingly.

show abstract

Section: Discussionsupporting

confidence: 64%

“…The predictive validity of the tools varies depending on study participants and methods. For example, the MFS showed the highest sensitivity in studies with hospitalized patients 9 , 17 and in a meta-analysis. 18 The STRATIFY showed the highest sensitivity in a study with neurological patients.…”

mentioning

confidence: 99%

Factors Influencing Falls in High- and Low-Risk Patients in a Tertiary Hospital in Korea

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Morse Fall Scale (MFS), hendrich Fall Risk Model (hFRM), The Saint Thomas Risk Assessment Tool in Falling Elderly Inpatients (STRATIFY). Odporúčajú sa aj pre posudzovanie rizika pádu u pacientov s neurologickým ochorením v nemocničnom prostredí [3,4,18,19].…”

Section: úVodunclassified

Risk factors of falling in patients with neurological diseases

Miertová

Bóriková

Tomagová

et al. 2018

Kontakt

View full text Add to dashboard Cite

A B S T R A C TPurpose: The aim of the study was to perform a fall risk screening and to identify the risk factors for falling in hospitalized patients with neurological disease aged up to and over 65 years. Design: A quantitative prospective cross-sectional study. Methods: The sample consisted of 103 patients who were hospitalized in the neurological unit. Empirical data was recorded in the research protocol, which included selected fall risk factors and the Morse Fall Scale (MFS). Results: Significant fall predictors in patients older than 65 years were: secondary diagnosis (p = 0.000), poly-pharmacotherapy (p = 0.000), number of medications taken over 24 hours (p = 0.000), a vision disorder (p = 0.005), continence disorder (p = 0.008), female sex (p = 0.005), the use of an ambulatory aid (p = 0.008), sleep disorders (p = 0.025), and hearing disorders (p = 0.033). The mean MFS total fall risk was 45.7 ± 20.2. In the sample of patients at high risk of falling, there were significant predictors of the falls: disorders of gait and balance (p = 0.000), the use of an ambulatory aid (p = 0.000), a history of falling (p = 0.001), age (p = 0.004), poly-pharmacotherapy = 0.008), continence disorder (p = 0.019), and vision disorder (p = 0.027). Conclusion: The identification of fall risk factors is the basis for the implementation of targeted nursing interventions to prevent and reduce falls. S Ú h R N Cieľ: Realizovať skríning rizika pádu a identifikovať rizikové faktory pádu u hospitalizovaných pacientov s neurologickým ochorením vo veku do a nad 65 rokov. Dizajn: Kvantitatívna prospektívna prierezová štúdia. Metodika: Súbor tvorilo 103 pacientov hospitalizovaných na neurologickom oddelení. Empirické údaje boli zaznamenávané do výskumného protokolu, ktorý obsahoval vybrané rizikové faktory pádu a nástroj Morse Fall Scale (MFS). Výsledky: Signifikantné prediktory pádu u pacientov starších ako 65 rokov boli pridružená diagnóza (p = 0,000), polyfarmakoterapia (p = 0,000), počet užívaných liekov * Korespondenční autor: Mgr. Michaela Miertová, PhD., Jesseniova lekárska fakulta UK, Ústav ošetrovateľstva, Malá hora 5, (Print) • ISSN 1804-7122 (Online) Kľúčové slová: Neurológia hospitalizovaní pacienti Pády Riziko Skríning Morse Fall Scale O š e t ř O v a t e l s t v í Článek citujte takto: Miertová M, Bóriková I, Tomagová M, Žiaková K. Risk factors of falling in patients with neurological diseases. Kontakt 2018; 20(3): e217-e222; http://dx.

show abstract

“…The MFS focuses primarily on intrinsic factors which are related to the individual factor such as history of falls or polypharmacy 9 and poses major limitations in two aspects. First, the prediction accuracy of MFS varies significantly among different healthcare settings and patient groups 10,11 making its application challenging due to the need for real-time intervention to response for changing patient conditions. Second, because the MFS does not focus on individual risk factors 12 , nurses must apply a broad fall prevention plan without necessarily focusing on individual patient's unique risk factors.…”

Section: Introductionmentioning

confidence: 99%

Development and validation of interpretable machine learning models for inpatient fall events and electronic medical record integration

Shim

Jekal

et al. 2022

Clin Exp Emerg Med

View full text Add to dashboard Cite

What is already known: Conventional score such as Morse fall scales was developed for fall risk assessment, however it has low accuracy and difficulty in practical implementation.Previous machine learning studies have limitation in its explainability for clinicians and nurses, which is also called as the black box problem. Another challenge is that previous models have not been well-prepared for Electronic Medical Records integration for practical application.What is new in the current study:We developed the interpretable Machine learning model for the prediction of falls by integrating into EMR system. ABSTRACT(1) Background: Falls are one of the most frequently occurring adverse events among hospitalized patients. The Morse fall scales, which have been widely used for fall risk assessment, has two limitations: low specificity and difficulty in practical implementation.The aim of this study was to develop and validate an interpretable machine learning (ML) model for the prediction of falls, which would be integrated in an electronic medical record system (EMR)(2) Methods: This is a retrospective study from a tertiary teaching hospital in Seoul, Korea.Data were collected from January 2018 to March 2020. Based on literatures, known 83 predictors were grouped into seven categories. Interpretable fall event prediction models were developed using multiple machine learning models including gradient boosting and Shapley values.(3) Results: Overall, 191,778 cases with 272 fall events (0.1%) were included for the analysis. With the validation cohort of 2020, the area under the receiver operating curve of the gradient boosting model was 0.817 (95% confidence interval, 0.720 -0.904) which showed better performance than random forest, logistic regression, artificial neural net, and conventional Morse fall score (0.80, 0.80, 0.74, and 0.65, respectively). The model's interpretability was enhanced in both population level and patient level. The algorithm was later integrated in to the current EMR.(4) Conclusion: We developed an interpretable ML prediction model for inpatient fall events using EMR integration formats.

show abstract

Comparison of the Reliability and Validity of Fall Risk Assessment Tools in Patients with Acute Neurological Disorders

Cited by 13 publications

References 21 publications

Factors Influencing Falls in High- and Low-Risk Patients in a Tertiary Hospital in Korea

Factors Influencing Falls in High- and Low-Risk Patients in a Tertiary Hospital in Korea

Risk factors of falling in patients with neurological diseases

Development and validation of interpretable machine learning models for inpatient fall events and electronic medical record integration

Contact Info

Product

Resources

About