In this study, our primary purpose was to gain a better understanding of the effectiveness of EDPS for managing pediatric asthma by evaluating whether it reduces the risk of repeat emergency department visits at one-year post-discharge in a real-world setting. We also sought to gain insight into the frequency of EDPS implementation at our tertiary care pediatric center. While we found that EDPS seemed to be associated with an 18% lower risk of emergency department re-visits, this association was not statistically significant. Therefore, it appears that EDPS at discharge may not be sufficient for controlling pediatric asthma and preventing repeat asthma ED visits. Although we found a high implementation of EDPS with 87% of physicians providing this to their patients and conforming to Canadian best practice guidelines for asthma management, current EDPS strategies are not enough. These recommendations should be revisited in order to reduce the observed incidence of ED re-visits in nearly 1 out of 4 patients in our study cohort.

Our findings are similar to those from a study in the United States, which likewise, did not find overall, that EDPS led to a reduction in hospital re-admission rates at 12 weeks; however, among all potential risk factors studied, strong education was considered the most essential factor to contribute to lower risk of hospital readmission in children [16]. This indicates that there is a need for stronger education and knowledge dissemination to parents and their children. Prior studies indicate that parents and caregivers desire to receive asthma management education through diverse media including videos, written material, and verbal face to face instructions with follow-up phone calls [17]. Thus, it may be that a more tailored and comprehensive approach, that could involve the diverse use of media may be needed to enhance the quality of asthma education and management provided for children and caregivers after an ED visit. As this may not always be available in the busy ED setting, increased referral to asthma specialists or educators may be needed for some patients, particularly those who may be at higher risk for ED re-visits.

Another aim of this study was to better understand the determinants of EDPS implementation and additional patient level risk factors linked with a higher risk of ED repeat visits. Given that higher PRAM scores were linked with an increased risk of ED revisits in this, and in previous studies [14, 15] stronger and targeted EDPS for this group of patients may be warranted. That is, even though we found that patients with higher PRAM scores are more likely to receive EDPS, they may benefit further from enhanced provision of these strategies as the current implementation approach is insufficient and does not result in best asthma control. The prevailing literature also supports our findings that previous ED visits increase the risk of hospital visits in the future [14, 18] yet it is important to note we did not find that patients with a previous history of ED visits were more likely to receive EDPS from their attending ED physicians. This suggests that ED physicians should be more mindful that patients with a previous history of ED visits need to be better targeted in order to optimize their receipt of EDPS. In addition to this, previous studies are also in agreement with our findings that having a primary care provider is associated with an increased risk of ED visits [19]. This may be due to confounding by indication, whereby patients with more severe asthma are more likely to seek out help from a PCP [24]. A prior study has suggested that patients with a PCP may be less likely to receive EDPS as ED physicians may assume that their PCP will manage their condition [20]. Yet up to 66% of patients do not seek follow-up care from their primary care provider [21] after an asthma ED visit, indicating that simply having a PCP is not protective against ED re-visits. Past research has found that second hand smoke significantly increases the risk of ED visits by 3.5 times and of increases the risk of hospitalization by 2.8 times in children with higher concentrations of second-hand smoke exposure based on serum cotinine measurements (> 3.0 ng/ml) relative to their counterparts, respectively [22]. However, we did not find this association in our study. This apparent lack of association in our current study may reflect the inconsistent recording of smoke exposure in a typical ED encounter and limitations of our data sources.

When considering other patient characteristics in an overall risk assessment, there are some additional risk factors that have previously been considered, including allergies and patient age. In the current study, we did not find an association between allergic history and ED re-visits, though we have previously found a higher risk for asthma hospitalizations among patients with comorbid peanut allergy at our centre, as also seen in other prior studies [23,24,25].This discrepancy could be accounted for by differences in study inclusion criteria; in the current study we included all children with any asthma ED visit, but excluded those who were already followed by an asthma specialist. This likely resulted in excluding a higher proportion of children with more severe asthma where a peanut allergy-association is more prominent, as evidenced by overall lower allergy prevalence in the current cohort (38.8%) compared to our previous study (62.7%). This cohort composition difference may have diminished the effect of allergies on ED re-visit risk. Future studies including larger sample sizes are needed to clarify this association.

Finally in this study, we did not find an association between young age and ED revisits; whether or not age is a potential independent patient level risk factor for ED visits is unclear per mixed findings in the literature. While some previous research has found an association between younger age and higher risk of repeat ED visits [26], other studies including ours have not found support for this [14, 19, 21, 26], indicating that age may not be as useful or of highest priority when assessing individual patient risk. Specifically, we did not find that age was a predictor of repeat asthma ED visits (OR = 0.96; p-value > 0.05; 95% CI includes 1). The study by Giangioppo et al. [14] examined 3300 paediatric ED asthma visits in Canada also did not find age to be a significant predictor of repeat asthma ED visits. As a result, we decided not to undertake further subgroup analyses stratified by age (i.e. comparing teenagers with younger children).

Our study has several strengths and limitations. Our chart review and data abstraction were performed directly by two of our authors, allowing us to perform a thorough search to extract relevant data. Our use of an objective diagnosis of asthma that was based on discharge diagnosis recorded in administrative data allowed reliable capture of our study population. However, we relied on ED physician documentation as our primary data source, which may not reflect true practices (i.e., in the case where a pertinent positive is documented, but a pertinent negative is not). We did encounter some missing data, such as PRAM at triage, which may have skewed our findings. In addition, our study was performed in a single paediatric tertiary care centre, and our population, has reasonably good access to health care, which may limit the generalizability of these findings to smaller community or rural EDs. Additionally, previous studies have shown that referral to a specialist improves asthma related morbidity [14], but we could not include this in our adjusted analyses due to the small percentage of patients referred to respirologists. Furthermore, we only considered variables that were readily available and documented in our ED records at the hospital, though our previous systematic review looking at predictors of future asthma hospitalization identified 28 risk-related variables, with only a small proportion included in the present study [27].

Finally, it should be noted that a major limitation of this study is that the proportion of children who did not receive EDPS is relatively small when compared with the number of children who received EDPS. However, our study was designed to be a real-life effectiveness study, which is likely the only feasible approach for evaluating the effectiveness of current EDPS, as a randomized controlled trial, by limiting access to EDPS for a large proportion of children would be unethical. As a result, our findings reflect real life trends and outcomes in an uncontrolled setting. Additionally, the severity of asthma differed between groups whereby 18.9% of patients in the EDPS group had mild asthma and 48% who did not receive EDPS also had mild asthma. This proportional difference may reflect the lower likelihood of physicians providing EDPS to a child presenting with mild asthma. Moreover, children with mild asthma presentations were less likely to return with a subsequent ED visits which may partially explain why there was not a significant difference in ED repeat visits. Though we undertook a sensitivity analysis to better understand the effectiveness of EDPS in children with milder versus more severe asthma presentations, we were limited by our small sample size and this approach may not be sufficient to be conclusive. Future studies with larger sample sizes in other settings may be needed to strengthen the evidence base. Although our study results should be interpreted with some caution given the unbalanced study sample, our findings are congruent with trends in repeat asthma ED visits seen at both at CHEO and across Canada over the past ten years, where the Canadian Institutes of Health Research [9] have shown that the rates of repeat asthma ED visits in children have not changed over the past decade. Therefore, despite some limitations, our study adds to the growing evidence-base which suggests that EDPS in Canada may not be enough and additional approaches to prevent repeat asthma ED visits should be explored.

Since this study was carried out at CHEO, we have now implemented several improvements and have developed a targeted approach for offering specialized care for children at higher risk for repeat asthma ED visits. A comprehensive asthma program was launched in 2018 at CHEO which involved extended nurse and staff training, interviews examining current practice barriers and needs, the use of an electronic medical record (EMR) to streamline referrals for comprehensive asthma education directly from the ED, and provision of virtual asthma education for eligible patients [28]. Evaluation of this new program is currently under way.