Association between evidence-based standardized protocols in emergency departments with childhood asthma outcomes: a Canadian population-based study

Assessment of Machine Learning-Based Medical Directives to Expedite Care in Pediatric Emergency Medicine

IMPORTANCE

Increased wait times and long lengths of stay in emergency departments (EDs) are associated with poor patient outcomes. Systems to improve ED efficiency would be useful. Specifically, minimizing the time to diagnosis by developing novel workflows that expedite test ordering can help accelerate clinical decision-making.

OBJECTIVE

To explore the use of machine learning-based medical directives (MLMDs) to automate diagnostic testing at triage for patients with common pediatric ED diagnoses.

DESIGN, SETTING, AND PARTICIPANTS

Machine learning models trained on retrospective electronic health record data were evaluated in a decision analytical model study conducted at the ED of the Hospital for Sick Children Toronto, Canada. Data were collected on all patients aged 0 to 18 years presenting to the ED from July 1, 2018, to June 30, 2019 (77 219 total patient visits).

EXPOSURE

Machine learning models were trained to predict the need for urinary dipstick testing, electrocardiogram, abdominal ultrasonography, testicular ultrasonography, bilirubin level testing, and forearm radiographs.

MAIN OUTCOMES AND MEASURES

Models were evaluated using area under the receiver operator curve, true-positive rate, false-positive rate, and positive predictive values. Model decision thresholds were determined to limit the total number of false-positive results and achieve high positive predictive values. The time difference between patient triage completion and test ordering was assessed for each use of MLMD. Error rates were analyzed to assess model bias. In addition, model explainability was determined using Shapley Additive Explanations values.

RESULTS

There was a total of 42 238 boys (54.7%) included in model development; mean (SD) age of the children was 5.4 (4.8) years. Models obtained high area under the receiver operator curve (0.89-0.99) and positive predictive values (0.77-0.94) across each of the use cases. The proposed implementation of MLMDs would streamline care for 22.3% of all patient visits and make test results available earlier by 165 minutes (weighted mean) per affected patient. Model explainability for each MLMD demonstrated clinically relevant features having the most influence on model predictions. Models also performed with minimal to no sex bias.

CONCLUSIONS AND RELEVANCE

The findings of this study suggest the potential for clinical automation using MLMDs. When integrated into clinical workflows, MLMDs may have the potential to autonomously order common ED tests early in a patient's visit with explainability provided to patients and clinicians.

The Practice of Obtaining a Chest Radiograph in Pediatric Patients Presenting With Their First Episode of Wheezing: A Survey of Resident Physicians

A chest radiograph (CXR) is not routinely indicated in children presenting with their first episode of wheezing; however, it continues to be overused. A survey was distributed electronically to determine what trainees are taught and their current practice of obtaining a CXR in children presenting with their first episode of wheezing and the factors that influence this practice. Of the 1513 trainees who completed surveys, 35.3% (535/1513) reported that they were taught that pediatric patients presenting with their first episode of wheezing should be evaluated with a CXR. In all, 22.01% (333/1513) indicated that they would always obtain a CXR in these patients, and 13.75% (208/1513) would always obtain a CXR under a certain age (4 weeks to 12 years, median of 2 years). Our study identifies a target audience that would benefit from education to decrease the overuse of CXRs in children.

Implementation of Ontario's emergency department asthma care pathway for adults: determinants of uptake

OBJECTIVE

The Ontario Government funded the development and implementation of a standardized adult emergency department (ED) asthma care pathway (EDACP). We aimed to describe baseline patterns of ED use by adults for asthma in Ontario, Canada, and determine site characteristics associated with the EDACP implementation workshop attendance and subsequent pathway implementation.

METHODS

All Ontario EDs were offered EDACP implementation workshops by the Lung Assocation-Ontario between 2008 and 2011, and were surveyed regarding site implementation status as of October, 2013. Survey data were linked by site to Ontario's administrative health databases. Logistic regression models investigated the association between site and patient characteristics and: a) workshop attendance; b) pathway implementation.

RESULTS

In the 2 years prior to EDACP implementation, there were 41 143 asthma visits to 167 sites by adults (62.3% female). Asthma-related return visits within 72 h varied by hospital type (teaching 2.1%, community 2.8%, small 4.0%; p < 0.05). Implementation workshops were attended by staff from 122 sites (72.6%). Implementation status was known for 108 sites and varied by hospital type (p < 0.001), but not workshop attendance (p = 0.11). By 2013, 47% of all hospitals were using or planning to use the EDACP. Uptake was more likely in community hospitals.

CONCLUSIONS

Ontario adult asthma ED visitors are more often women. Asthma-related return visits are uncommon, but significantly higher in small community hospitals. This provincial QI initiative reached almost 75% of Ontario EDs, and achieved almost 50% implementation rate within 2 years. Factors other than workshop attendance, such as hospital size, were associated with EDACP implementation.

Successful implementation of evidence-based guidelines in a regional emergency department for children presenting with acute asthma

OBJECTIVE

To evaluate the adherence to asthma evidence-based guidelines before and after a quality improvement process.

DESIGN

A controlled trial was conducted at two regional hospitals (intervention and control hospitals). We performed a retrospective pre-intervention audit, followed by a post-intervention audit 1 year after the implementation of evidence-based guidelines.

SETTING

Emergency departments of two neighbouring hospitals serving regional and rural North West Tasmania.

PARTICIPANTS

All children (<18 years) with acute presentation of doctor-diagnosed asthma.

INTERVENTIONS

Implementation of evidence-based guidelines using the National Asthma Council of Australia and the Global Initiative of Asthma guidelines, at the intervention hospital and care as usual at the control hospital.

MAIN OUTCOME MEASURE(S)

The main outcome measures were the compliance to evidence-based guidelines, pre- and postintervention at the intervention hospital, compared to the control hospital. The specific outcomes measure included the clinical presentation, management, referral to asthma and allergy clinic, and hospitilisation.

RESULTS

Significantly improved adherence to evidence-based guidelines were noted post-intervention at the intervention hospital, that is severity recorded (21.4%-45.7%, P < 0.001), triggers identified (13.5%-45.3%, P < 0.001), spirometry usage (3.8%-25.8%, P = 0.03) and written action plans (29.7%-58.3%, P < 0.001). There was however no effect on hospitilisation (23.3%-29.8%, P = 0.48). At the control hospital, however, no significant improved adherence to evidence-based guidelines were noted.

CONCLUSIONS

Evidence-based implementation led to improved adherence to evidence-based guidelines across an expanded list of domains in a regional setting.

Differentiating non-responders from responders in children with moderate and severe asthma exacerbations

Objective: Our goal was to assess factors associated with non-response to treatment in children presenting to the Emergency Department (ED) with moderate and severe asthma exacerbations. Methods: A retrospective chart review was completed from 9/2014 to 2/2017 for patients with a discharge diagnosis of asthma exacerbation. The Modified Pulmonary Index Score (MPIS) was used to quantify illness acuity. The rate of change of MPIS per hour was calculated, and differentiated responders from non-responders. After examining a histogram of ΔMPIS/h, a threshold of ΔMPIS/h > 0 was used to define response for duration of ED stay. Children included were >2 years and had initial MPIS > 10. Results: Eight hundred and fifty-two children were included. There were 178 (21%) non-responders and 674 (79%) responders. Non-responders were significantly older (7.0 ± 4.0 versus 5.6 ± 3.2 years; p < 0.001), but there were no differences in gender, race, ethnicity or insurance status. There was also no statistical difference in time to first albuterol treatment (50 ± 41 versus 43 ± 40 min; p = 0.05), or in time to corticosteroid (95 ± 75 versus 79 ± 64 min; p = 0.06). Non-responders were significantly more likely to arrive by ambulance (OR 2.2; 95% CI 1.6-3.2), to be admitted to the hospital (OR 2.7; 95% CI 1.8-4.0), and to be admitted to the ICU (OR 5.0; 95% CI 3.1-8.1). Conclusions: One in five children with exacerbations did not respond to treatment. These children were older and more likely to be admitted. Non-measured factors, possibly genetic, may contribute to response to treatment.

Association of Guideline-concordant Acute Asthma Care in the Emergency Department With Shorter Hospital Length of Stay: A Multicenter Observational Study

OBJECTIVES

The objectives were to determine whether guideline-concordant emergency department (ED) management of acute asthma is associated with a shorter hospital length of stay (LOS) among patients hospitalized for asthma.

METHODS

A multicenter chart review study of patients aged 2-54 years who were hospitalized for acute asthma at one of the 25 U.S. hospitals during 2012-2013. Based on level A recommendations from national asthma guidelines, we derived four process measures of ED treatment before hospitalization: inhaled β-agonists, inhaled anticholinergic agents, systemic corticosteroids, and lack of methylxanthines. The outcome measure was hospital LOS.

RESULTS

Among 854 ED patients subsequently hospitalized for acute asthma, 532 patients (62%) received care perfectly concordant with the four process measures in the ED. Overall, the median hospital LOS was 2 days (interquartile range = 1-3 days). In the multivariable negative binomial model, patients who received perfectly concordant ED asthma care had a significantly shorter hospital LOS (-17%, 95% confidence interval [CI] = -27% to -5%, p = 0.006), compared to other patients. In the mediation analysis, the direct effect of guideline-concordant ED asthma care on hospital LOS was similar to that of primary analysis (-16%, 95% CI = -27% to -5%, p = 0.005). By contrast, the indirect effect mediated by quality of inpatient asthma care was not significant, indicating that the effect of ED asthma care on hospital LOS was mediated through pathways other than quality of inpatient care.

CONCLUSION

In this multicenter observational study, patients who received perfectly concordant asthma care in the ED had a shorter hospital LOS. Our findings encourage further adoption of guideline-recommended emergency asthma care to improve patient outcomes.

Emergency Department Revisits by Urban Immigrant Children in Canada: A Population-Based Cohort Study

OBJECTIVES

To examine the relationship between family immigrant status and unscheduled 7-day revisits to the emergency department (ED) and to test this relationship within subgroups of immigrants by visa class (family, economic, refugee), native tongue on landing in Canada, and region of origin.

STUDY DESIGN

Population-based cohort study that used multiple linked health administrative and demographic datasets of landed immigrant and nonimmigrant children (<18 years) in urban Ontario who visited an ED and were discharged between April 2003 and March 2010. Logistic regression was used to model the odds of 7-day ED revisits with family immigrant status, with adjustment for patient and ED characteristics.

RESULTS

Of 3,322,901 initial visits to the ED, 249,648 (7.5%) resulted in a 7-day revisit. There was no significant association of immigrant status with either ED revisits or poor revisit outcomes (greater acuity visit or need for admission) in the adjusted models. Within immigrants, the odds of revisit were not associated with immigrant classes or region of origin; however, immigrants whose native tongue was not English or French had a slightly greater odds of revisiting the ED (aOR 1.05; 95% CI 1.01, 1.09). Significant predictors of revisits included younger age, greater triage acuity score, greater predilection for using an ED, daytime shifts, and greater deprivation index.

CONCLUSIONS

Immigrant children are not more likely to have short-term revisits to the ED, but there may be barriers to care related to language fluency that need to be addressed. These findings may be relevant for improving translation services in EDs.

Knowledge translation studies in paediatric emergency medicine: A systematic review of the literature

AIM

Systematic review of knowledge translation studies focused on paediatric emergency care to describe and assess the interventions used in emergency department settings.

METHODS

Electronic databases were searched for knowledge translation studies conducted in the emergency department that included the care of children. Two researchers independently reviewed the studies.

RESULTS

From 1305 publications identified, 15 studies of varied design were included. Four were cluster-controlled trials, two patient-level randomised controlled trials, two interrupted time series, one descriptive study and six before and after intervention studies. Knowledge translation interventions were predominantly aimed at the treating clinician, with some targeting the organisation. Studies assessed effectiveness of interventions over 6-12 months in before and after studies, and 3-28 months in cluster or patient level controlled trials. Changes in clinical practice were variable, with studies on single disease and single treatments in a single site showing greater improvement.

CONCLUSIONS

Evidence for effective methods to translate knowledge into practice in paediatric emergency medicine is fairly limited. More optimal study designs with more explicit descriptions of interventions are needed to facilitate other groups to effectively apply these procedures in their own setting.

Have expert guidelines made a difference in asthma outcomes?

PURPOSE OF REVIEW

To analyze and summarize research from 2011 to 2012 that examines the relationship of guideline implementation and asthma outcomes.

RECENT FINDINGS

Evidence of an unmet need for better asthma management was reported in two large survey studies from the US and Europe. Interventional studies of guideline implementation were often limited by lack of uptake of the intervention (i.e. educational program, computer-assisted assessment). Even studies in which there was uptake to the intervention, asthma outcomes often did not improve. Certain interventions (specific electronic asthma management tools, provider education workshops, community-wide interventional programs, and parental educational programs) were associated with improved asthma outcomes. Observational studies, likewise, revealed that evidence of guideline implementation did not necessarily translate into improved asthma outcomes.

SUMMARY

Asthma guideline implementation studies are frequently associated with a limited impact on asthma outcomes. Understanding the gaps between guideline recommendations and translation to clinical practice remains an important opportunity to improve asthma outcomes.

Variation and trends in ED use of radiographs for asthma, bronchiolitis, and croup in children

OBJECTIVES

The objectives were (1) to determine trends in radiograph use in emergency department (ED) care of children with asthma, bronchiolitis, and croup; and (2) to examine the association of patient and hospital factors with variation in radiograph use.

METHODS

A retrospective, cross-sectional study of National Hospital Ambulatory Medical Care Survey data between 1995 and 2009 on radiograph use at ED visits in children aged 2 to 18 years with asthma, aged 3 months to 1 year with bronchiolitis, and aged 3 months to 6 years with croup. Odds ratios (ORs) were calculated and adjusted for all factors studied.

RESULTS

The use of radiographs for asthma increased significantly over time (OR: 1.06; 95% confidence interval [CI]: 1.03-1.09; P < .001 for trend) but were unchanged for bronchiolitis and croup. Pediatric-focused EDs had lower use for asthma (OR: 0.44; 95% CI: 0.29-0.68), bronchiolitis (OR: 0.37; 95% CI: 0.23-0.59), and croup (OR: 0.34; 95% CI: 0.17-0.68). Compared with the Northeast region, the Midwest and South had statistically higher use of radiographs for all 3 conditions. The Western region had higher use only for asthma (OR: 1.67; 95% CI: 1.07-2.60), and bronchiolitis (OR: 2.94; 95% CI: 1.48-5.87). No associations were seen for metropolitan statistical area or hospital ownership status.

CONCLUSIONS

The ED use of radiographs for children with asthma increased significantly from 1995 to 2009. Reversing this trend could result in substantial cost savings and reduced radiation. Pediatric-focused EDs used significantly fewer radiographs for asthma, bronchiolitis, and croup. The translation of practices from pediatric-focused EDs to all EDs could improve performance.