A data-driven and practice-based approach to identify risk factors associated with hospital-acquired falls: Applying manual and semi- and fully-automated methods

BACKGROUND AND PURPOSE Electronic health record (EHR) data provides opportunities for new approaches to identify risk factors associated with iatrogenic conditions, such as hospital-acquired falls. There is a critical need to validate and translate prediction models that support fall prevention clinical decision-making in hospitals. The purpose of this study was to explore a combined data-driven and practice-based approach to identify risk factors associated with falls. PROCEDURES We conducted an observational case-control study of EHR data from January 1, 2013 to October 31, 2013 from 14 medical-surgical units of a tertiary referral teaching hospital. Patients aged 21 or older admitted to medical surgical units were included in the study. Manual and semi- and fully-automated methods were used to identify fall risk factors across four prediction models. Sensitivity, specificity, and the Area under the Receiver Operating Characteristic (AUROC) curve were calculated for all models using 10-fold cross validation. FINDINGS We confirmed the significance of a set of valid fall risk factors (i.e., age, gender, fall risk assessment, history of falling, mental status, mobility, and confusion) and identified set of new risk factors (i.e., # of fall risk increasing drugs, hemoglobin level, physical therapy initiation, Charlson Comorbity Index, nurse skill mix, and registered nurse staffing ratio) based on the most precise prediction approach, namely stepwise regression. CONCLUSIONS The use of semi- and fully-automated approaches with expert clinical knowledge over expert or data-driven only approaches can significantly improve identifying patient, clinical, and organizational risk factors of iatrogenic conditions, including hospital-acquired falls.

[1]  D. Kosmopoulos,et al.  Automated Pressure Ulcer Lesion Diagnosis for Telemedicine Systems , 2007, IEEE Engineering in Medicine and Biology Magazine.

[2]  M. Mazumdar,et al.  Inpatient Falls after Total Knee Arthroplasty: The Role of Anesthesia Type and Peripheral Nerve Blocks , 2014, Anesthesiology.

[3]  S. Horn,et al.  Practice-based evidence for clinical practice improvement: an alternative study design for evidence-based medicine. , 2010, Studies in Health Technology and Informatics.

[4]  J. Stevens,et al.  Medical Costs of Fatal and Nonfatal Falls in Older Adults , 2018, Journal of the American Geriatrics Society.

[5]  Ewout W Steyerberg,et al.  Validation and updating of predictive logistic regression models: a study on sample size and shrinkage , 2004, Statistics in medicine.

[6]  Y. Shin,et al.  A prediction model of falls for patients with neurological disorder in acute care hospital , 2015, Journal of the Neurological Sciences.

[7]  E. Melo,et al.  Fall risk assessment: retrospective analysis of Morse Fall Scale scores in Portuguese hospitalized adult patients. , 2016, Applied nursing research : ANR.

[8]  J. Golmard,et al.  Two faller risk functions for geriatric assessment unit patients , 2006, AGE.

[9]  Chi Kin Chan,et al.  Evaluation of the Morse Fall Scale: applicability in Chinese hospital populations. , 2007, International journal of nursing studies.

[10]  Evaluation of falls by inpatients in an acute care hospital in Korea using the Morse Fall Scale. , 2014, International journal of nursing practice.

[11]  J. Morse.  The safety of safety research: the case of patient fall research. , 2006, The Canadian journal of nursing research = Revue canadienne de recherche en sciences infirmieres.

[12]  D. Margolis,et al.  Pressure Ulcer Preventive Device Use Among Elderly Patients Early in the Hospital Stay , 2009, Nursing research.

[13]  Gregory L. Alexander,et al.  Effects of a computerized decision support system on pressure ulcers and malnutrition in nursing homes for the elderly , 2011, Int. J. Medical Informatics.

[14]  B. Jacobs,et al.  Recognizing Hypoglycemia in Children Through Automated Adverse-Event Detection , 2011, Pediatrics.

[15]  Nada Nassar,et al.  Predicting falls using two instruments (the Hendrich Fall Risk Model and the Morse Fall Scale) in an acute care setting in Lebanon. , 2014, Journal of clinical nursing.

[16]  Janice M. Morse,et al.  Development of a Scale to Identify the Fall-Prone Patient , 1989, Canadian Journal on Aging / La Revue canadienne du vieillissement.

[17]  Stefan Rust,et al.  Mining geriatric assessment data for in-patient fall prediction models and high-risk subgroups , 2012, BMC Medical Informatics and Decision Making.

[18]  Yan Ma,et al.  In-hospital patient falls after total joint arthroplasty: incidence, demographics, and risk factors in the United States. , 2012, The Journal of arthroplasty.

[19]  George Hripcsak,et al.  Evaluation of an Automated Pressure Ulcer Risk Assessment Model , 2007 .

[20]  Lucila Ohno-Machado,et al.  Building knowledge in a complex preterm birth problem domain , 2000, AMIA.

[21]  L. Ohno-Machado,et al.  “Big Data” and the Electronic Health Record , 2014, Yearbook of Medical Informatics.

[22]  N. Perkins,et al.  The inconsistency of "optimal" cutpoints obtained using two criteria based on the receiver operating characteristic curve. , 2006, American journal of epidemiology.

[23]  Pa-Chun Wang,et al.  Data Mining Techniques for Assisting the Diagnosis of Pressure Ulcer Development in Surgical Patients , 2012, Journal of Medical Systems.

[24]  S. Permpongkosol.  Iatrogenic disease in the elderly: risk factors, consequences, and prevention , 2011, Clinical interventions in aging.

[25]  J. Morse,et al.  A prospective study to identify the fall-prone patient. , 1989, Social science & medicine.

[26]  M. Spaeder,et al.  Monitoring the harm associated with use of anticoagulants in pediatric populations through trigger-based automated adverse-event detection. , 2015, Joint Commission journal on quality and patient safety.

[27]  Brian R. Gaines.  An Ounce of Knowledge is Worth a Ton of Data: Quantitative studies of the Trade-Off between Expertise and Data Based On Statistically Well-Founded Empirical Induction , 1989, ML.

[28]  G. Collins,et al.  Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): The TRIPOD Statement , 2015, Annals of Internal Medicine.

[29]  Marco R. Spruit,et al.  Transitioning to a Data Driven Mental Health Practice: Collaborative Expert Sessions for Knowledge and Hypothesis Finding , 2016, Comput. Math. Methods Medicine.

[30]  A. Rutjes,et al.  Trigger Tool–Based Automated Adverse Event Detection in Electronic Health Records: Systematic Review , 2018, Journal of medical Internet research.

[31]  A. Hendrich,et al.  Hospital falls: development of a predictive model for clinical practice. , 1995, Applied nursing research : ANR.

[32]  Padhraic Smyth,et al.  From Data Mining to Knowledge Discovery: An Overview , 1996, Advances in Knowledge Discovery and Data Mining.

[33]  B. Fitzgerald.  Guidance Regarding Methods for De-identification of Protected Health Information in Accordance with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule , 2015 .

[34]  Stan Matwin,et al.  Using Qualitative Models to Guide Inductive Learning , 1993, ICML.

[35]  M. Aranda-Gallardo,et al.  Instruments for assessing the risk of falls in acute hospitalized patients: a systematic review and meta-analysis , 2013, BMC Health Services Research.

[36]  S. Horn,et al.  Practice-Based Evidence Study Design for Comparative Effectiveness Research , 2007, Medical care.

[37]  Trevor Hastie,et al.  Regularization Paths for Generalized Linear Models via Coordinate Descent. , 2010, Journal of statistical software.

[38]  Norman Fenton,et al.  Integrating Expert Knowledge with Data in Bayesian Networks: Preserving Data-Driven Expectations when the Expert Variables Remain Unobserved , 2016, Expert Syst. Appl..

[39]  Eric S. Kirkendall,et al.  Phenotyping for patient safety: algorithm development for electronic health record based automated adverse event and medical error detection in neonatal intensive care , 2014, Journal of the American Medical Informatics Association : JAMIA.

[40]  S. De Geest,et al.  Evaluation of the Morse Fall Scale in hospitalised patients. , 2006, Age and ageing.

[41]  M. Makary,et al.  Medical error—the third leading cause of death in the US , 2016, British Medical Journal.

[42]  W. E. Hammond,et al.  Data Mining Methods Find Demographic Predictors of Preterm Birth , 2001, Nursing research.

[43]  M. Daniels,et al.  Falls Among Adult Patients Hospitalized in the United States: Prevalence and Trends , 2012, Journal of patient safety.

[44]  A. McDaniel.  Automated adverse event detection. , 2004, Clinical nurse specialist CNS.

[45]  Rok Blagus,et al.  SMOTE for high-dimensional class-imbalanced data , 2013, BMC Bioinformatics.

[46]  Yinji Jin,et al.  Validity of the Morse Fall Scale implemented in an electronic medical record system. , 2014, Journal of clinical nursing.

[47]  Lucila Ohno-Machado,et al.  Automating pressure ulcer risk assessment using documented patient data , 2010, Int. J. Medical Informatics.

[48]  D. Oliver,et al.  Development and evaluation of evidence based risk assessment tool (STRATIFY) to predict which elderly inpatients will fall: case-control and cohort studies , 1997, BMJ.

[49]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.