Factors associated with failure of emergency wait-time targets for high acuity discharges and intensive care unit admissions
SHSC-IC-Failed Wait-Time Targets
Sunnybrook Health Sciences Centre (ED)
Objective: Ontario established emergency department length-of-stay (EDLOS) targets but has difficulty achieving them. We sought to determine predictors of target time failure for discharged high acuity patients and intensive care unit (ICU) admissions.
Methods: This was a retrospective, observational study of 2012 Sunnybrook Hospital emergency department data. The main outcome measure was failing to meet government EDLOS targets for high acuity discharges and ICU emergency admissions. The secondary outcome measures examined factors for low acuity discharges and all admissions, as well as a run chart for 2015 - 2016 ICU admissions. Multiple logistic regression models were created for admissions, ICU admissions, and low and high acuity discharges. Predictor variables were at the patient level from emergency department registries.
Results: For discharged high acuity patients, factors predicting EDLOS target failure were having physician initial assessment duration (PIAD)>2 hours (OR 5.63 [5.22-6.06]), consultation request (OR 10.23 [9.38-11.14]), magnetic resonance imaging (MRI) (OR 19.33 [12.94-28.87]), computed tomography (CT) (OR 4.24 [3.92-4.59]), and ultrasound (US) (OR 3.47 [3.13-3.83]). For ICU admissions, factors predicting EDLOS target failure were bed request duration (BRD)>6 hours (OR 364.27 [43.20-3071.30]) and access block (AB)>1 hour (OR 217.27 [30.62-1541.63]). For discharged low acuity patients, factors predicting failure for the 4-hour target were PIAD>2 hours (OR 15.80 [13.35-18.71]), consultation (OR 20.98 [14.10-31.22]), MRI (OR 31.68 [6.03-166.54]), CT (OR 16.48 [10.07-26.98]), and troponin I (OR 13.37 [6.30-28.37]).
Conclusion: Sunnybrook factors predicting failure of targets for high acuity discharges and ICU admissions were hospital-controlled. Hospitals should individualize their approach to shortening EDLOS by analysing its patient population and resource demands.
Authors: Ivy Cheng, Merrick Zwarenstein, Alex Kiss, Maaret Castren, Mats Brommels and Michael Schull
Ivy Cheng - email@example.com
Preliminary data gathering/ baseline