Morita 2017.

Methods	Study design: randomised, open study. Aim: to investigate whether pranlukast added to usual asthma therapy in Japanese children during the autumn reduces asthma exacerbations. The effects of age and sex on the efficacy of pranlukast were also evaluated. Study centres and method of recruitment: multiple clinical sites in Chiba, Japan. Study participants were recruited between July 2007 and August 2007 through advertising and from the clinical practices in Chiba, Japan. Dates of study: 15 September 2007 to 14 November 2007. Run‐in period: from recruitment until 15 September 2007. Duration of participation: 60 days in addition to run‐in period. Consent: the investigation was approved by the Research Ethics Board of Chiba Universiy, Chiba (approval number: 631). Written informed consent was obtained from the parents of all participants and child assent when appropriate. Power: no a priori calculation. Imputation of missing data, i.e. assumptions made for ITT analysis: 13.6% of children excluded after randomisation in the pranlukast group (2.8% placebo), but no imputation made.
Participants	Age (mean, range): 5.5 years (not reported but supplied by author), 1 to 14 years (divided into 2 age groups: 1 to 5 years and 6 to 14 years). Gender: 62.8% male. Asthma severity: 54.5% required inhaled corticosteroids. Diagnostic criteria: physician‐diagnosed asthma. Asthma was diagnosed by primary care doctors based on the Japanese paediatric guidelines for the treatment and management of bronchial asthma 2005. Number recruited: 204 Number randomised (intervention, control): 102, 102 Number completed (intervention, control): 59, 72 Number analysed (intervention, control): 51, 70 Withdrawals: 43 from intervention group and 30 from control group excluded before trial due to respiratory symptoms or insufficient diary recording by caregivers, or both, during the observation period. 8 from intervention group and 2 from control group excluded during the study period due to poor compliance or insufficient diary recording by caregivers, or both. Inclusion criteria: age 1 to 14 years old, physician‐diagnosed asthma needing a rescue inhaler in the last year, with a history of asthma exacerbations associated with apparent respiratory viral infections. Children who had been treated with LTRA were included after 14‐day washout period. Exclusion criteria: significant cardiorespiratory comorbidity; using regular oral corticosteroid; or had an asthma exacerbation in the month before treatment with pranlukast started. Children who had respiratory symptoms or problems with diary recording during observation, or both, were excluded from the study.
Interventions	Intervention: regular pranlukast, an LTRA. 7 mg/kg, twice daily, in addition to their usual asthma therapy. Comparison: usual therapy. Concomitant medication: intervention taken in addition to usual asthma therapy. No restriction, but children who had been treated with LTRA were included after a 14‐day washout period. Excluded medication: no restriction, but 14‐day washout of LTRA.
Outcomes	Primary outcome: total asthma score calculated during 8 weeks. Total asthma score was evaluated as follows: a blue sticker (score, 0) was applied on days when a child had no asthma symptoms; a green sticker (score, 1) indicated mild asthma symptoms; a yellow sticker (score, 2) indicated symptoms that were worse than usual or needed extra asthma medication, and an orange sticker (score, 3) was applied if a child's breathing symptoms required an unscheduled visit to a physician or treatment with oral corticosteroids. Secondary outcomes: days with worse asthma symptoms, number of colds, and days with fever. Days with worse asthma symptoms were defined as those with either an orange or a yellow sticker. A fever was defined as a temperature exceeding 38 °C. A “cold” was defined as the presence of more than 2 consecutive purple stickers indicating days with cold symptoms. At least 5 days with no cold symptoms were required before a subsequent new cold was identified. Time points measured: contemporaneous data collection at the end of 60 days. Primary outcome result: there were no significant differences between pranlukast and control group in total asthma score at 8 weeks (5.5 vs 7.8, P = 0.35), and in the days in which a child experienced a worsening of asthma symptoms (1.5 vs 1.8, P = 0.67) (data obtained through correspondence with the author). Secondary outcome results: higher number of colds in the control group compared to the pranlukast group (P = 0.06), and children taking pranlukast experienced fewer days with fever compared to the control group (P = 0.04). Adverse events: no children discontinued study medication due to adverse events.
Notes	Funding: not stated. Subgroups: Boys vs girls. 1 to 5 years vs 6 to 14 years. Boys aged 1 to 5 years had the lower total asthma score at 8 weeks (P = 0.002), and experienced fewer cold episodes (P = 0.007). In boys, pranlukast significantly reduced total asthma score among 1‐ to 5‐year‐olds (P = 0.010), but did not reduce it among 6‐ to 14‐year‐olds. In girls, pranlukast did not affect total asthma score among 1‐ to 5‐year‐olds, but increased total asthma score among 6‐ to 14‐year‐olds (P = 0.027). 60 cold episodes were reported in the pranlukast group and 107 cases in the control group. A significant reduction in the number of cold episodes was observed in 1‐ to 5‐year‐old boys who were treated with pranlukast (P < 0.001).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random assignment to either the pranlukast intervention group or the control group. Randomisation conducted according to sex and within the predefined age groups (1 to 5 years and 6 to 14 years).
Allocation concealment (selection bias)	Unclear risk	No description reported.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Study was of open‐label design. The authors recognised this as a limitation of the study.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Symptoms were reported subjectively by study participants. Participants and study observers were not blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	High rate of exclusions from pranlukast group after randomisation
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No baseline differences between groups. Comparisons of the baseline characteristics of the study groups were conducted using Chi² and Mann‐Whitney U‐tests.