Objective: To assess multicentre, randomised, controlled trials (MC-RCTs) of systemic inflammatory response syndrome (SIRS) and sepsis conducted in Japan, published in Japanese and not available to English-language medical databases.
Design: Methodological review.
Subjects: All Japanese RCTs relevant to SIRS and sepsis.
Intervention: Identification of manuscripts using a Japanese electronic library. Critical analysis of methodology and reporting quality using a modified Methodological Quality Assessment Score and the CONSORT group check list.
Measurements and results: Three MC-RCTs were identified. In the first, 147 patients with septic shock were randomised to methylprednisolone (1000 mg i.v.) or placebo. In the second, 221 patients were randomised to 0.20 mg/kg per h or 0.004 mg/kg per h of sivelestat for acute lung injury with SIRS. In the third, 504 patients were randomised to immunoglobulin (5 g for 3 days) or to a control group. The average methodological quality score was higher than that of equivalent Western trials. The reporting quality (CONSORT checklist) was comparable to Western studies published during the same period.
Conclusions: Despite sound methodology and quality, the information obtained from relatively large Japanese critical care trials is not widely available to English-speaking investigators and therefore might be ignored in meta-analyses.