Identifying patient characteristics associated with potentially redirectable paramedic transported emergency department visits in Ontario, Canada: a population-based cohort study

Validating the Emergency Department Avoidability Classification (EDAC): A cluster randomized single-blinded agreement study

INTRODUCTION

The Emergency Department Avoidability Classification (EDAC) retrospectively classifies emergency department (ED) visits that could have been safely managed in subacute primary care settings, but has not been validated against a criterion standard. A validated EDAC could enable accurate and reliable quantification of avoidable ED visits. We compared agreement between the EDAC and ED physician judgements to specify avoidable ED visits.

MATERIALS AND METHODS

We conducted a cluster randomized, single-blinded agreement study in an academic hospital in Hamilton, Canada. ED visits between January 1, 2019, and December 31, 2019 were clustered based on EDAC classes and randomly sampled evenly. A total of 160 ED visit charts were randomly assigned to ten participating ED physicians at the academic hospital for evaluation. Physicians judged if the ED visit could have been managed appropriately in subacute primary care (an avoidable visit); each ED visit was evaluated by two physicians independently. We measured interrater agreement between physicians with a Cohen's kappa and 95% confidence intervals (CI). We evaluated the correlation between the EDAC and physician judgements using a Spearman rank correlation and ordinal logistic regression with odds ratios (ORs) and 95% CIs. We examined the EDAC's precision to identify avoidable ED visits using accuracy, sensitivity and specificity.

RESULTS

ED physicians agreed on 139 visits (86.9%) with a kappa of 0.69 (95% CI 0.59-0.79), indicating substantial agreement. Physicians judged 96.2% of ED visits classified as avoidable by the EDAC as suitable for management in subacute primary care. We found a high correlation between the EDAC and physician judgements (0.64), as well as a very strong association to classify avoidable ED visits (OR 80.0, 95% CI 17.1-374.9). The EDACs avoidable and potentially avoidable classes demonstrated strong accuracy to identify ED visits suitable for management in subacute care (82.8%, 95% CI 78.2-86.8).

DISCUSSION

The EDAC demonstrated strong evidence of criterion validity to classify avoidable ED visits. This classification has important potential for accurately monitoring trends in avoidable ED utilization, measuring proportions of ED volume attributed to avoidable visits and informing interventions intended at reducing ED use by patients who do not require emergency or life-saving healthcare.

Prehospital prediction of hospital admission for emergent acuity patients transported by paramedics: A population-based cohort study using machine learning

INTRODUCTION

The closest emergency department (ED) may not always be the optimal hospital for certain stable high acuity patients if further distanced ED's can provide specialized care or are less overcrowded. Machine learning (ML) predictions may support paramedic decision-making to transport a subgroup of emergent patients to a more suitable, albeit more distanced, ED if hospital admission is unlikely. We examined whether characteristics known to paramedics in the prehospital setting were predictive of hospital admission in emergent acuity patients.

MATERIALS AND METHODS

We conducted a population-level cohort study using four ML algorithms to analyze ED visits of the National Ambulatory Care Reporting System from January 1, 2018 to December 31, 2019 in Ontario, Canada. We included all adult patients (≥18 years) transported to the ED by paramedics with an emergent Canadian Triage Acuity Scale score. We included eight characteristic classes as model predictors that are recorded at ED triage. All ML algorithms were trained and assessed using 10-fold cross-validation to predict hospital admission from the ED. Predictive model performance was determined using the area under curve (AUC) with 95% confidence intervals and probabilistic accuracy using the Brier Scaled score. Variable importance scores were computed to determine the top 10 predictors of hospital admission.

RESULTS

All machine learning algorithms demonstrated acceptable accuracy in predicting hospital admission (AUC 0.77-0.78, Brier Scaled 0.22-0.24). The characteristics most predictive of admission were age between 65 to 105 years, referral source from a residential care facility, presenting with a respiratory complaint, and receiving home care.

DISCUSSION

Hospital admission was accurately predicted based on patient characteristics known prehospital to paramedics prior to arrival. Our results support consideration of policy modification to permit certain emergent acuity patients to be transported to a further distanced ED. Additionally, this study demonstrates the utility of ML in paramedic and prehospital research.

Evaluating emergency department transfers from urgent care centres: insights for paramedic integration with subacute healthcare

OBJECTIVE

Paramedics redirecting non-emergent patients from emergency departments (EDs) to urgent care centres is a new and forthcoming strategy to reduce overcrowding and improve primary care integration. Which patients are likely not suitable for paramedic redirection are unknown. To describe and specify patients inappropriate for urgent care centres, we examined associations between patient characteristics and transfer to the ED after patients initially presented to an urgent care centre.

METHODS

A population-based retrospective cohort study of all adult (≥18 years) visits to an urgent care centre from 1 April 2015 to 31 March 2020 in Ontario, Canada. Binary logistic regression was used to determine unadjusted and adjusted associations between patient characteristics and being transferred to an ED using OR and 95% CIs. We calculated the absolute risk difference for the adjusted model.

RESULTS

A total of 1 448 621 urgent care visits were reported, with 63 343 (4.4%) visits transferred to an ED for definitive care. Being 65 years and older (OR 2.29, 95% CI 2.23 to 2.35), scored an emergent Canadian Triage and Acuity Scale of 1 or 2 (OR 14.27, 95% CI 13.45 to 15.12) and higher comorbidity count (OR 1.51, 95% CI 1.46 to 1.58) had added odds of association with being transferred out to an ED.

CONCLUSION

Readily available patient characteristics were independently associated with interfacility transfers between urgent care centres and the ED. This study can support paramedic redirection protocol development, highlighting which patients may not be best suited for ED redirection.

Validation of a classification to identify emergency department visits suitable for subacute and virtual care models: a randomised single-blinded agreement study protocol

INTRODUCTION

Redirecting suitable patients from the emergency department (ED) to alternative subacute settings may assist in reducing ED overcrowding while delivering equivalent care. The Emergency Department Avoidance Classification (EDAC) was constructed to retrospectively classify ED visits that may have been suitable for safe management in a subacute or virtual clinical setting. The EDAC has established face and content validity but has not been tested against a reference standard as a criterion.

OBJECTIVES

Our primary objective is to examine the agreement between the EDAC and ED physician judgements in retrospectively identifying ED visits suitable for subacute care management. Our secondary objective is to assess the validity of ED physicians' judgement as a criterion standard. Our tertiary objective is to examine how the ED physician's perception of a virtual ED care alternative correlates with the EDAC.

METHODS AND ANALYSIS

A randomised single-centre, single-blinded agreement study. We will randomly select ED charts between 1 January and 31 December 2019 from an academic hospital in Hamilton, Canada. ED charts will be randomly assigned to participating ED physicians who will evaluate if this ED visit could have been managed appropriately and safely in a subacute and/or virtual model of care. Each chart will be reviewed by two physicians independently. We compute our needed sample size to be 79 charts. We will use kappa statistics to measure inter-rater agreement. A repeated measures regression model of physician ratings will provide variance estimates that we will use to assess the intraclass correlation of ED physician ratings and the EDAC.

ETHICS AND DISSEMINATION

This study has been approved by the Hamilton Integrated Research Ethics Board (2022-14625). If validated, the EDAC may provide an ED-based classification to identify potentially avoidable ED visits, monitor ED visit trends, and proactively delineate those best suited for subacute or virtual care models.

Increased demand for paramedic transports to the emergency department in Ontario, Canada: a population-level descriptive study from 2010 to 2019

Purpose

We examined changes in annual paramedic transport incidence over the ten years prior to COVID-19 in comparison to increases in population growth and emergency department (ED) visitation by walk-in.

Methods

We conducted a population-level cohort study using the National Ambulatory Care Reporting System from January 1, 2010 to December 31, 2019 in Ontario, Canada. We included all patients triaged in the ED who arrived by either paramedic transport or walk-in. We clustered geographical regions using the Local Health Integration Network boundaries. Descriptive statistics, rate ratios (RR), and 95% confidence intervals were calculated to explore population-adjusted changes in transport volumes.

Results

Overall incidence of paramedic transports increased by 38.3% (n = 264,134), exceeding population growth fourfold (9.4%) and walk-in ED visitation threefold (13.4%). Population-adjusted transport rates increased by 26.2% (rate ratio 1.26, 95% CI 1.26–1.27) compared to 3.4% for ED visit by walk-in (rate ratio 1.03, 95% CI 1.03–1.04). Patient and visit characteristics remained consistent (age, gender, triage acuity, number of comorbidities, ED disposition, 30-day repeat ED visits) across the years of study. The majority of transports in 2019 had non-emergent triage scores (60.0%) and were discharged home directly from the ED (63.7%). The largest users were persons aged 65 or greater (43.7%). The majority of transports occurred in urbanized regions, though rural and northern regions experienced similar paramedic transport growth rates.

Conclusion

There was a substantial increase in the demand for paramedic transportation. Growth in paramedic demand outpaced population growth markedly and may continue to surge alongside population aging. Increases in the rate of paramedic transports per population were not bound to urbanized regions, but were province-wide. Our findings indicate a mounting need to develop innovative solutions to meet the increased demand on paramedic services and to implement long-term strategies across provincial paramedic systems.