Development of an algorithm for assessing fall risk in a Japanese inpatient population

Abstract

Falling is a representative incident in hospitalization and can cause serious complications. In this study, we constructed an algorithm that nurses can use to easily recognize essential fall risk factors and appropriately perform an assessment. A total of 56,911 inpatients (non-fall, 56,673; fall; 238) hospitalized between October 2017 and September 2018 were used for the training dataset. Correlation coefficients, multivariable logistic regression analysis, and decision tree analysis were performed using 36 fall risk factors identified from inpatients. An algorithm was generated combining nine essential fall risk factors (delirium, fall history, use of a walking aid, stagger, impaired judgment/comprehension, muscle weakness of the lower limbs, night urination, use of sleeping drug, and presence of infusion route/tube). Moreover, fall risk level was conveniently classified into four groups (extra-high, high, moderate, and low) according to the priority of fall risk. Finally, we confirmed the reliability of the algorithm using a validation dataset that comprised 57,929 inpatients (non-fall, 57,695; fall, 234) hospitalized between October 2018 and September 2019. Using the newly created algorithm, clinical staff including nurses may be able to appropriately evaluate fall risk level and provide preventive interventions for individual inpatients.

Falls as the result of interplay between nurses, patient and the environment: Using text-mining to uncover how and why falls happen

“…What this study adds to the existing body of knowledge is how nursing practice increases the risk of falling. Traditionally, the risk of falling has been investigated through patient records using mathematical modeling [ 16 , 18 , 21 ]. Hence, how nurses’ perceptions, comprehensions, decisions, and actions affect the risk of falling have rarely been examined.…”

“…A multitude of risk factors for falls have been identified, and these are usually categorized into intrinsic and extrinsic factors [ 13 , 14 ]. The intrinsic factors include the characteristics of a patient, such as sex [ 15 ], the types of disease [ 16 , 17 ], decline in physical function [ 6 , 18 ], elimination and sleeping problems [ 18 , 19 ], confusion and impaired understanding [ 20 ], the types of medication used [ 6 , 17 , 21 ], and patients’ predisposition such as “doing everything on one’s own” [ 20 ]. The extrinsic factors involve the use of unsafe footwear including slippers [ 22 ], tubes and drains attached to patients [ 21 ], and wet and slippery floors [ 23 ].…”

“…The intrinsic factors include the characteristics of a patient, such as sex [ 15 ], the types of disease [ 16 , 17 ], decline in physical function [ 6 , 18 ], elimination and sleeping problems [ 18 , 19 ], confusion and impaired understanding [ 20 ], the types of medication used [ 6 , 17 , 21 ], and patients’ predisposition such as “doing everything on one’s own” [ 20 ]. The extrinsic factors involve the use of unsafe footwear including slippers [ 22 ], tubes and drains attached to patients [ 21 ], and wet and slippery floors [ 23 ]. These intrinsic and extrinsic factors have been incorporated into several fall-screening tools to identify those at risk of falling so that precautions can be taken to prevent falls.…”

Artificial intelligence for falls management in older adult care: A scoping review of nurses' role

“…The study designs were all quantitative in nature, using retrospective case–control or cohort designs, with some not explicitly reporting the methodology used. Nine studies employed secondary analysis on existing falls related datasets from electronic health records, chart reviews, clinical guidelines or published scientific literature, web‐based reporting systems, national surveys or registry data (Cho et al, 2019 , 2021 ; Jung et al, 2020 ; Lee et al, 2010 , 2020 ; Nakanishi et al, 2021 ; Nakatani et al, 2020 ; Yang et al, 2021 ; Yokota et al, 2017 ). Four studies conducted primary research by collecting falls associated data via geriatric or clinical assessments of older adults, physical and mobility assessments using sensors to detect risk factors and self‐reported questionnaires completed by older participants (Beauchet et al, 2018 ; Greene et al, 2014 ; Makino et al, 2021 ; Rabe et al, 2020 ).…”

Simulation Versus Written Fall Prevention Education in Older Hospitalized Adults: A Randomized Controlled Study

“…In a large retrospective study, investigators assessed 36 risk factors of falls in hospitalized older adults. Fall history and delirium were the two most important risks (Nakanishi et al, 2021), and in a systematic review and meta-analysis, although falls had multifactorial etiologies, the primary risk factor was a history of falls (Deandrea et al, 2013). In older adults, even ground-level falls had consequences.…”