Lisa Kuhn1, Linda Worrall-Carter, John Ward, Karen Page. 1. St Vincent's Centre for Nursing Research, School of Nursing, Midwifery and Paramedicine, Faculty of Health Sciences, Australian Catholic University, 115 Victoria Parade, Fitzroy, Melbourne, VIC 3065, Australia. Electronic address: lisa.kuhn@acu.edu.au.
Abstract
BACKGROUND: Minimising time to treatment onset for acute myocardial infarction (AMI) in the emergency department (ED) is essential, yet little is understood about the interactions between variables affecting it. The aim of this study was to develop a regression tree model explicating the influence of patient and non-patient factors on the time taken to commence treatment for patients with AMI in Victorian EDs. METHODS: A regression tree model for variables impacting time to treatment was developed on retrospective data for patients aged 18-85 years with AMI treated in Victorian EDs from 2005 to 2010 (n=21,080). Data were partitioned into three subsets, with a complexity parameter set at 0.0005. RESULTS: Four variables emerged in the final regression tree model: triage score; mode of arrival; area of residence; and patient sex. The variable most influencing time to treatment onset for AMI was triage category. For undertriaged patients, treatment time patterns were affected by arrival mode, residential location and their sex, significantly extending delays to treatment onset. CONCLUSIONS: Interactions between specific variables influenced whether patients with AMI were treated with equity in Victorian EDs, resulting in previously unidentified evidence-practice gaps and an improved understanding of which patient groups were vulnerable to delayed treatment for AMI.
BACKGROUND: Minimising time to treatment onset for acute myocardial infarction (AMI) in the emergency department (ED) is essential, yet little is understood about the interactions between variables affecting it. The aim of this study was to develop a regression tree model explicating the influence of patient and non-patient factors on the time taken to commence treatment for patients with AMI in Victorian EDs. METHODS: A regression tree model for variables impacting time to treatment was developed on retrospective data for patients aged 18-85 years with AMI treated in Victorian EDs from 2005 to 2010 (n=21,080). Data were partitioned into three subsets, with a complexity parameter set at 0.0005. RESULTS: Four variables emerged in the final regression tree model: triage score; mode of arrival; area of residence; and patient sex. The variable most influencing time to treatment onset for AMI was triage category. For undertriaged patients, treatment time patterns were affected by arrival mode, residential location and their sex, significantly extending delays to treatment onset. CONCLUSIONS: Interactions between specific variables influenced whether patients with AMI were treated with equity in Victorian EDs, resulting in previously unidentified evidence-practice gaps and an improved understanding of which patient groups were vulnerable to delayed treatment for AMI.
Authors: Natasha Sobers; Angela M C Rose; T Alafia Samuels; Julia Critchley; Melissa Abed; Ian Hambleton; Arianne Harvey; Nigel Unwin Journal: BMJ Open Date: 2019-01-28 Impact factor: 2.692
Authors: Ki Hong Kim; Jeong Ho Park; Young Sun Ro; Ki Jeong Hong; Kyoung Jun Song; Sang Do Shin Journal: PLoS One Date: 2020-11-05 Impact factor: 3.240