Tree-based Approach to Predict Hospital Acquired Pressure Injury

다국어 초록 (Multilingual Abstract)

Despite technical advances in healthcare, the rates of hospital-acquired pressure injury (HAPI) are still highalthough many are potentially preventable. The purpose of this study was to determine whether tree-basedprediction modeling is suitable for assessing the risk of HAPI in ICU patients. Retrospective cohort study hasbeen carried out. A decision tree model was constructed with Age, Weight, eTube, diabetes, Braden score,Isolation, and Number of comorbid conditions as decision nodes. We used RStudio for model training and testing.
Correct prediction rate of the final prediction model was 92.4 and the Area Under the ROC curve (AUC) was0.699, which means there is about 70% chance that the model is able to distinguish between HAPI and non-HAPI. The results of this study has limited generalizability as the data were from a single academic institution.
Our research finding shows that the data-driven tree-based prediction modeling may potentially support ICUsensitive risk assessment for HAPI prevention.

참고문헌 (Reference)

1 Schoonhoven L, "Prospective cohort study of routine use of risk assessment scales for prediction of pressure ulcers" 325 (325): 2002

2 Jan Kottner, "Pressure ulcer risk assessment in critical care: Interrater reliability and validity studies of the Braden and Waterlow scales and subjective ratings in two intensive care units" Elsevier BV 47 (47): 671-677, 2010

3 Pacharmon Kaewprag, "Predictive models for pressure ulcers from intensive care unit electronic health records using Bayesian networks" Springer Science and Business Media LLC 17 (17): 65-, 2017

4 Ridinger M, "Predictive modeling points way to future risk status" 21 (21): 10-12, 2000

5 S. Hyun, "Predictive Validity of the Braden Scale for Patients in Intensive Care Units" AACN Publishing 22 (22): 514-520, 2013

6 Jill Cox, "Predictive Power of the Braden Scale for Pressure Sore Risk in Adult Critical Care Patients" Ovid Technologies (Wolters Kluwer Health) 39 (39): 613-621, 2012

7 Bergstrom N, "Predicting pressure ulcer risk: a multisite study of the predictive validity of the Braden Scale" 47 (47): 261-269, 1998

8 Brickley M, "Neural networks: a new technique for development of decision support systems in dentistry" 26 (26): 305-309, 1998

9 The National Pressure Ulcer Advisory Panel, "National Pressure Ulcer Advisory Panel (NPUAP)announces a change in terminology from pressure ulcer to pressure injury and updates the stages of pressure injury"

10 Dreiseitl S, "Logistic regression and artificial neural network classification models: a methodology review" 35 (35): 352-359, 2002

11 Apilak Worachartcheewan, "Identification of metabolic syndrome using decision tree analysis" Elsevier BV 90 (90): e15-e18, 2010

12 Ellenius J, "Dynamic decision support graph--visualization of ANN-generated diagnostic indications of pathological conditions developing over time" 42 (42): 189-198, 2008

13 Mullins I, "Data mining and clinical data repositories: Insights from a 667,000 patient data set" 36 (36): 1351-1377, 2006

14 Liew P, "Comparison of artificial neural networks with logistic regression in prediction of gallbladder disease among obese patients" 39 (39): 356-362, 2007

15 Kurt I, "Comparing performances of logistic regression classification and regression tree, and neural networks for predicting coronary artery disease" 34 : 366-374, 2008

16 Massimo Tabaton, "Artificial Neural Networks Identify the Predictive Values of Risk Factors on the Conversion of Amnestic Mild Cognitive Impairment" IOS Press 19 (19): 1035-1040, 2010

17 Kuhn M, "Applied Predictive Modeling" Springer 2016

18 Lin S, "A comparison of MICU survival prediction using the logistic regression model and artificial neural network model" 14 (14): 306-314, 2006