Use of Machine-Learning Approaches to Predict Clinical Deterioration in Critically Ill Patients: A Systematic Review
Journal Title: International Journal of Medical Research & Health Sciences (IJMRHS) - Year 2017, Vol 6, Issue 6
Abstract
Introduction: Early identifcation of patients with unexpected clinical deterioration is a matter of serious concern. Previous studies have shown that early intervention on a patient whose health is deteriorating improves the patient outcome, and machine-learning-based approaches to predict clinical deterioration may contribute to precision improvement. To date, however, no systematic review in this area is available. Methods: We completed a search on PubMed on January 22, 2017 as well as a review of the articles identifed by study authors involved in this area of research following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines for systematic reviews. Results: Twelve articles were selected for the current study from 273 articles initially obtained from the PubMed searches. Eleven of the 12 studies were retrospective studies, and no randomized controlled trials were performed. Although the artifcial neural network techniques were the most frequently used and provided high precision and accuracy, we failed to identify articles that showed improvement in the patient outcome. Limitations were reported related to generalizability, complexity of models, and technical knowledge. Conclusions: This review shows that machine-learning approaches can improve prediction of clinical deterioration compared with traditional methods. However, these techniques will require further external validation before widespread clinical acceptance can be achieved.
Authors and Affiliations
Tadashi Kamio| Department of Anesthesiology and Critical Care Medicine, Jichi Medical University Saitama Medical Center, Saitama, Japan, Institute of Advanced BioMedical Engineering and Science, Tokyo Women’s Medical University, Tokyo, Japan, Corresponding e-mail: tadashi-kamio@mail.goo.ne.jp, Tomoaki Van| Institute of Advanced BioMedical Engineering and Science, Tokyo Women’s Medical University, Tokyo, Japan, Ken Masamune| Institute of Advanced BioMedical Engineering and Science, Tokyo Women’s Medical University, Tokyo, Japan
Keywords
Related Articles
- EP ID EP12379
- DOI -
- Views 381
- Downloads 16
Physicochemical study of some types of Algerian honeys
The modern bee-keeping proposes various types of honeys of floral and geographical origin, of very varied savor and aspect. It is often the only source of sugar of the most withdrawn indigenous populations of the tropic...
TOOTHBRUSH DISINFECTION –A MYTH OR REALITY? A COMPARATIVE EVALUATION OF 4% DISODIUM EDTA, 10% SODIUM PERBORATE IN THE DISINFECTION OF TOOTHBRUSHES: CLINICOMICROBIOLOGICAL STUDY
Aim: The aim of this randomized clinical trial was to evaluate the bacterial survival rate on toothbrushes and efficacy of their decontamination by4% disodium ethyl diamine acetic acid [EDTA], 10% sodium perborate and...
Effects of Muscle Relaxation on Anxiety of Parents Who Have Children with Leukaemia Undergoing Chemotherapy
This research intended to determine the effects of the muscle relaxation techniques on anxiety of parents having children with leukaemia who were undergoing chemotherapy at teaching hospitals of Zahedan in 2015.This was...
BENEFITS VERSUS RISKS OF PROTON PUMP INHIBITORS: ARE WE OPENING THE CAN OF WORMS
Proton pump inhibitors (PPIs) are one of the most commonly used drugs worldwide They are indicated for treatment of Gastro-esophageal Reflux Disease (GERD), acid peptic disorders, stress ulcers and prophylaxis of NSAI...
Examining the prevalence of Hemoglobin A1C level disorder in women affected with polycystic ovaries syndrome (PCOS) at Imam Ali Hospital, Karaj
PCOS in commonly associated with metabolic disorders, therefore it, is reasonable and timely action to identify and manage the disorders. Accordingly in this study we decided to determine, the prevalence of abnormal hem...