LAUSR

Guideline recommendations about discharge of children with acute asthma treated with IV magnesium (IVMg) are disparate and inconclusive (see online supplemental material). The majority of children given IVMg in the emergency department (ED) are hospitalised, independent of asthma severity or degree of response to IVMg. The rationale for this practice is unknown but may be due to limited evidence whether children with a satisfactory response to IVMg can be safely discharged. We conducted this international survey of three paediatric emergency research networks in Canada (Paediatric Emergency Research Canada), Australia/New Zealand (Paediatric Research in Emergency Departments International Collaborative) and the UK/Ireland (Paediatric Emergency Research United Kingdom and Ireland) belonging to the international Paediatric Emergency Research Network to determine the proportion of paediatric ED physicians who agree there is adequate evidence that children with acute asthma refractory to the initial corticosteroid and bronchodilator therapy, and who attain a satisfactory and sustained response to IVMg, can be safely discharged home. The modified Dillman’s method was used for participant contact from March to June 2021. Using network membership and email lists, we invited participants by email to click on a link to complete a webbased, 25item, twopage Research Electronic Data Capture survey (online supplemental material). The first page asked screening and demographic questions; the second page contained the survey questions. Physicians not treating children and those in training were ineligible. Following best practices for survey studies, study authors performed item generation, reduction, pretesting and pilot testing. Physicians were asked to rate on a 4point Likert scale (‘strongly agree’ to ‘strongly disagree’) the extent to which they agreed that children remaining in marked respiratory distress after stabilisation therapy with bronchodilators and steroids, who then have a satisfactory and sustained response to IVMg (mild asthma: PRAM 2/12 points for ≥3 hours post IVMg) can be safely discharged home. To detect a proportion of 40% (based on authors’ consensus) of physicians agreeing (strongly agree or agree) that the evidence of IVMg benefit is adequate with an error margin of 5%, 369 participants were needed. We used the χ test with 95% CIs for differences in relevant proportions, which included post hoc betweencountry outcome comparisons. Surveys without responses to the actual survey questions were not analysed. A total of 886 physicians were invited to participate; 657 (74.1%) responded, of whom 36 were ineligible, and 35 replied only to the first page, leaving 586 surveys for analysis (Canada 131, UK/Ireland 391 and Australia/New Zealand 64). A total of 104/586 participants (104 (17.7%, 95% CI 14.7% to 21.1%)) agree that current evidence of IVMg benefit about safe discharge after a satisfactory and sustained response to IVMg is adequate, with a significant difference between Canada and UK/Ireland (difference 52% (95% CI 44% to 61%)) vs Australia/New Zealand (difference 48.6% (95% CI 38% to 60%)), with a p value of <0.001 for both. The decision to use IVMg prompts Canadian physicians to hospitalise less frequently than those in UK/Ireland (difference −52% (95% CI −62% to −43%)) and Australia/ New Zealand (difference −49% (−59% to −38%)), with a p value of <0.001 for both (table 1). Concern about inadequate evidence of benefit or discharge safety represents the main reason for routine hospitalisation after IVMg. This study has several limitations. Survey responses may not accurately reflect actual practice and may not be generalisable to other countries. This study was not a priori powered for betweencountry comparisons. A minority of paediatric ED physicians consider the current evidence regarding the safety of discharge after IVMg in refractory acute asthma adequate and the majority routinely hospitalised after IVMg, irrespective of clinical response. There are significant differences in hospitalisation after IVMg across networks. Studies are needed to determine if patients can be safely discharged after IVMg, and future guidelines should be based on such research.