Skip to main content
. 2018 Mar 8;2018(3):CD012393. doi: 10.1002/14651858.CD012393.pub2

Johnston 2007.

Methods Study design: randomised, double‐blind, placebo‐controlled trial.
Aim: to determine whether montelukast, added to usual asthma therapy, would reduce days with worse asthma symptoms and unscheduled physician visits of children during the September epidemic.
Study centres and method of recruitment: recruited through advertising and through clinical practices in Hamilton and Brantford, Canada.
Dates of study: 1 September 2005 to 15 October 2005.
Run‐in period: no run‐in period.
Duration of participation: 45 days.
Consent: approved by the research ethics board at St. Joseph's Healthcare Hamilton. Informed consent from parents and assent from appropriately aged children.
Power: a 40% reduction was expected in days with worse asthma symptoms in the montelukast group based upon results of a pilot study. Based upon 80% power and a 0.05 significance level, a sample‐size requirement of 88 per group was estimated. A 10% dropout rate was allowed for, so the final sample requirement was 97 per group.
Imputation of missing data, i.e. assumptions made for ITT analysis: all randomised children completed the study and were included in analysis.
Participants Age (mean, range): not reported, 2 to 14 years.
Gender: 65.0% male.
Asthma severity: not explicitly mentioned, but 90% required inhaled corticosteroids (likely moderate to severe).
Diagnostic criteria: physician‐diagnosed asthma.
Number recruited: 196
Number randomised (intervention, control): 98, 96
Number completed (intervention, control): 98, 96
Number analysed (intervention, control): 98, 96
Withdrawals: 100% completed, no withdrawals.
Inclusion criteria: 2 to 14 years old; physician‐diagnosed asthma needing a rescue inhaler in the last year; missing ≥ 1 day from school because of asthma in the last year or having significant limitation of normal activity; having a history of asthma exacerbations associated with apparent respiratory viral infections; ability to communicate in English.
Exclusion criteria: significant cardiorespiratory comorbidity; using an LTRA; using regular OCS medication; asthma exacerbation in the month before study inception.
Interventions Intervention: montelukast age‐specific dose from 1 September to 15 October.
Comparison: matched placebo.
Concomitant medication: usual therapy.
Excluded medication: already on montelukast.
Outcomes Primary outcome: percentage of days with worsening asthma symptoms during the intervention period (worsening symptoms defined as symptoms that were worse than usual or needed extra asthma medication, or requiring an unscheduled visit to a doctor or treatment with oral corticosteroids).
Secondary outcome: number of unscheduled care visits.
Time points measured: daily, then at the end of the study.
 Primary outcome result: the montelukast group experienced a 53% reduction in days with worse asthma symptoms compared with placebo (3.9% vs 8.3%, P = 0.02).
Secondary outcome results: the montelukast group experienced a 78% reduction in unscheduled physician visits for asthma (4 for montelukast vs 18 for placebo, P = 0.011).
Adverse events: minor adverse events occurred in 25 children in the montelukast group and in 35 children in the placebo group. 2 children discontinued study medication due to adverse events, 1 due to a personality change and 1 with change in appetite and increased tiredness; both children were taking placebo. The trial code was not broken, and symptom recording was continued. Another significant event was identified at the follow‐up interview after a child assigned to receive montelukast required emergency treatment for acute behaviour disorder.
Notes Funding: Merck Frosst Canada Ltd.
Subgroups: subgroup analyses were exploratory risk of asthma worsening intervention vs control:

  • regular ICS users OR 0.13 95% CI 0.03 to 0.51

  • no ICS use OR 0.14, 95% CI 0.04 to 0.53

  • intermittent ICS use OR 0.37, 95% CI 0.10 to –1.31

  • regular ICS/LABA use OR 0.44, 95% CI 0.11 to 1.75

  • intermittent ICS/LABA use OR 1.24, 95% CI 0.31 to 4.89

  • boys 2 to 5 years OR 0.03, 95% CI 0.01 to 0.21

  • boys 6 to 9 years OR 0.27, 95% CI 0.09 to 0.87

  • boys 10 to 14 years OR 0.81, 95% CI 0.24 to 2.77

  • girls 2 to 5 years OR 1.29, 95% CI 0.18 to 9.1

  • girls 6 to 9 years OR 0.68, 95% CI 0.13 to 3.45

  • girls 10 to 14 years OR 0.17, 95% CI 0.05 to 0.52
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Computer‐generated randomisation schedule. Randomly assigned in blocks of 4 according to gender and age.
Allocation concealment (selection bias) Low risk Randomisation schedule described as "concealed" and generated by an individual "not otherwise involved in the study". Mechanism of concealment described as based upon identical containers issued by third party (further information supplied by authors).
Blinding of participants and personnel (performance bias) 
 All outcomes Low risk Double‐blinded. Intervention drug and placebo prepared by Merck Frosst, no reason to suspect parent or child could identify intervention drug from placebo.
Blinding of outcome assessment (detection bias) 
 All outcomes Low risk Given the use of a placebo, unlikely that the assessors would have knowledge of participant group. Subjective participant‐reported parent‐assessed symptoms and questionnaire used to assess other outcomes; these could have been affected if blinding inadequate, but no reason to suspect placebo led to incomplete blinding. Physician validated unscheduled care.
Incomplete outcome data (attrition bias) 
 All outcomes Low risk Intention‐to‐treat primary analysis, 100% children completed the trial and returned 99.7% diary data. Adherence good in both groups (91.7% intervention vs 93.2% placebo).
Selective reporting (reporting bias) Unclear risk Unclear whether all prespecified outcomes included in the analysis
Other bias Low risk No baseline differences between groups, except more lifetime hospitalisations: 37.8% intervention vs 25.0% placebo