Dexamethasone Versus Prednisone for Asthma Treatment in the Pediatric Inpatient Population; a Feasibility Study

1. BACKGROUND AND RATIONALE 1.1 Background Information Summary of the problem Asthma is the most common childhood chronic disease, affecting over 15% of children between the ages of 4 and 11 years. In 2004, asthma contributed to 10% and 8% of all Canadian hospital admissions for children aged 0 to 4 years and 5 to 14 years, respectively. In Ontario, the burden of illness caused by asthma is enormous, accounting for one third of government expenditures for the general population. A key element for treating children with acute asthma exacerbations is the administration of corticosteroids. Corticosteroids (CS) reduce the need for hospitalization and the risk of relapse after initial treatment, and may also facilitate an earlier discharge from the hospital. While some hospitals already use a short course of dexamethasone for the inpatient asthma treatment, there is no evidence to support this practice in the inpatient population. Most current CS treatment regimens for the treatment of children hospitalized with an asthma exacerbation consist of a 5-day course of prednisone or prednisolone. However, prednisone and prednisolone are associated with poor taste and significant vomiting, as well as poor compliance with the treatment course. The lifetime prevalence of asthma in Canadian children has been estimated at 11% to 16% and asthma exacerbations are the leading cause of hospitalization in children. At the Children's Hospital of Eastern Ontario (CHEO), 215 children were admitted to the inpatient units for asthma in 2015, representing 5.4% of the yearly medical admissions. Current corticosteroid standard of care for inpatient treatment of asthma CS are one of the cornerstones of therapy for acute asthma. They should be administered as early as possible in the Emergency Department (ED). Treatment with CS reduces the need for hospitalization, decreases the risk of relapse post treatment, and leads to a shorter length of stay. Current corticosteroid treatment regimens for the treatment of children hospitalized with an asthma exacerbation typically consist of a 4-day course of prednisone or prednisolone in addition to the dose of oral CS received in the ED. Dexamethasone vs prednisone/prednisolone treatment A recent meta-analysis has shown that 2 days of once daily dexamethasone is a least as effective as 5 days of prednisone for preventing relapse in outpatient pediatric asthma. In addition, patients receiving dexamethasone were significantly less likely to experience vomiting in the ED and even after returning home. Preliminary cost-estimates suggest that dexamethasone may result in at least a $3500 cost savings per 100 patients compared to traditional treatment with prednisone or prednisolone. At CHEO, this would result in over $7000 in savings per year. Compliance with corticosteroid treatment Prednisone is only available in Canada as a tablet or compounded suspension, which limits use due to swallowing ability and accessibility. When compared to dexamethasone, prednisolone is associated with poor palatability, significant vomiting and poor compliance. Prednisone and prednisolone are compounded with similar recipes, and therefore have a similar taste. Compliance with 5 days of prednisone has been estimated in one study to be as low as 64% for pediatric asthma. As mentioned above, outpatient evidence shows that a 2-day course of dexamethasone is as effective as a longer course of prednisone and prednisolone, and better tolerated. However, as determined by a systematic review conducted by the principal investigator of this current proposal (PROSPERO 2016:CRD42016041766), studies in hospitalized patients are lacking. Promisingly, a retrospective cohort study in pediatric inpatients hospitalized with asthma has suggested that dexamethasone, when compared to prednisone/prednisolone, may result in a shorter length of hospital stay and reduced costs with no difference in number of transfers to intensive care or readmissions. Safety of dexamethasone Dexamethasone is a potent glucocorticoid with a long half-life, therefore concerns have been raised regarding its potential for adrenal suppression. However, a study of high-dose (~1.7 mg/kg) IM dexamethasone in acute asthma found no significant difference in adrenal function at day 14 between single-dose dexamethasone and 5 days of oral prednisone. The degree of adrenal suppression induced by dexamethasone is likely comparable to that observed with intravenous methylprednisolone or moderate to high doses of inhaled corticosteroids. Children in our proposed study will receive a cumulative dexamethasone dose of 1.2 mg/kg orally, significantly less than used in the aforementioned study. 1.2 Rationale Given the importance of CS in the treatment of asthma, the significant decrease in relapse associated with their use, and the high prevalence of childhood asthma, there is a need to determine whether a better tolerated and more palatable CS can be used as first line therapy. Although outpatient data on this issue exists, extrapolating treatment regimens from the outpatient population may not be appropriate as hospitalized children represent a sicker group of patients than those discharged from the emergency department. In order to determine whether dexamethasone is at least as effective of prednisone/prednisolone in the treatment of inpatient asthma, the investigators propose a feasibility study, as a first step in the development of a future multi-site trial.
2. STUDY OBJECTIVES

The investigators plan to determine the feasibility of a non-inferiority trial, comparing 2 days of dexamethasone to 4 days of inpatient prednisone/prednisolone for inpatient asthma treatment. Specifically, The investigators we will determine the feasibility of:

1. Enrolling patients upon admission to hospital, after they receive their first dose of corticosteroid (CS) in the ED
2. Asking patients and/or caregivers to complete a symptom diary weekly for 4 weeks
3. Reassessing patients 7 days after hospital admission day
4. Successfully completing phone follow up 4 weeks post hospital discharge
5. Collecting health utilization data post hospital discharge.

3 METHODOLOGY 3.1 Trial design A pragmatic randomized trial is proposed. Children presenting to the Children's Hospital of Eastern Ontario (CHEO) Emergency Department (ED) will receive CS (typically prednisone/prednisolone 2mg/kg) as per standard of care outlined in the ED pre-printed order and Nursing Medical Directive. . This practice differs from most other pediatric EDs across the country where children typically receive dexamethasone 0.3 mg/kg on presentation to the ED. As there is significant outpatient evidence showing that dexamethasone is non-inferior to prednisone for outpatient asthma treatment, it is possible that the CHEO ED will have switched over to administering dexamethasone 0.3 mg/kg to children presenting with acute asthma exacerbation by the time this study is implemented. Blinding of the participants and members of the health care team will not be feasible. Given that the investigators are aiming to test the effectiveness of their intervention across routine clinical practice, they elected to conduct a pragmatic trial. Since the palatability of the CS is likely to affect compliance, masking the taste of the CS to ensure blinding would decrease the ability to detect a difference in compliance due to taste. Moreover, given the length of therapy for prednisone/prednisolone is longer than that of dexamethasone, introducing placebo doses to ensure similar length of treatment would also potentially impact compliance and undermine the investigators' ability to detect a difference between groups. Investigators, data analysts, and research assistant completing patient assessment at the follow-up visit will be blinded to group assignment.

A research assistant will review the list of admissions for the previous 12 hours, on a twice daily basis, from Monday to Friday. The research assistant will then ask a member of the patient's health care team (bedside nurse, charge nurse, resident, staff physician) to ask the family members for permission to approach them about the study. The research assistant will then complete an eligibility screen, and maintain a phone follow-up ledger for consented patients. If the patient is eligible and the family gives their informed consent, they will be allocated to one of the study groups. Once written, informed consent has been obtained, the research assistant will communicate with the health care team to ensure the patient continues on the CS initiated in the ED, for the appropriate number of doses as determined by CS group assignment. A record will be kept of all screened patients, patient eligibility, allocation, and follow up to allow reporting according to CONSORT guidelines. Additional consent will be requested to access patients' health utilization data (readmissions, ED visits, visits to family physicians) through linkage with ICES data.

After enrollment, the patients will be randomly allocated to one of the two treatment groups. The randomization schedule will have been previously generated using a computer. Randomization will be blocked with randomly chosen block lengths of 4 or 6. Treatment assignments will be written on a piece of paper and concealed in sequentially numbered opaque envelopes kept in a secure locked location in the study research office. The PI and analyst will be blinded to the treatment intervention, but the research coordinator in charge of assessing eligibility and allocation and the patient's treating team (physicians and nurses) will not because of the nature of the pragmatic nature of the trial.

1. ANALYSIS 4.1 Sample size The investigators sought a sample size that would allow them to estimate their feasibility outcomes with reasonable precision. As such, they have set out to achieve a 15% margin of error (i.e. half-width of 95% confidence interval=0.15) for any proportions to be estimated (e.g. allocation success, retention success). At a hypothesized proportion of 50%, they determined that 43 patients will be needed to achieve this level of precision. This provides a most conservative estimate since any proportions other than 50% will require fewer patients. Factoring in an additional 15% of patients as allowance for attrition or incomplete data they have set a recruitment target of n=51 in total. With approximately 18 asthma admissions per month, assuming 70% are eligible and 50% of eligible patients consent, the investigators expect to meet this sample size requirement in 9 months.

4.2 Quantitative Data Analysis For all relevant feasibility outcomes, binary proportions (e.g. % success) and the associated 95% confidence interval will be estimated using the Wilson method. To help inform the design of main trial (i.e. expected effect sizes), the investigators will apply intention-to-treat principles and estimate the treatment effect (and 95% CI) by comparing outcomes (proposed for the main trial) between intervention groups using statistical techniques appropriate to the type and distribution of the various outcomes (e.g. continuous, binary, count data).

4.3 Qualitative Data Analysis Semi-structured interview at week 4 will be audio taped and transcribed verbatim. Analyses will be conducted using NVivo 9 software. Inductive analysis will be used to identify categories, patterns and themes.