Background: Molluscum contagiosum is a benign skin disorder of viral origin, often seen in children. In our Cochrane review ‘Interventions for cutaneous molluscum contagiosum’, we included all interventions studied in randomised controlled trials (RCTs). One of the treatments that was included was imiquimod cream, a drug that enhances the immune response. In the most recent version of the review (updated in 2009), two small studies were included that had imiquimod in one of the treatment arms. Only 41 patients had been given the treatment. The study comparing imiquimod cream to its vehicle (n = 23) found a risk ratio (RR) of 4.62 for our primary outcome, complete clearance of lesions four weeks after end of treatment, with a huge confidence interval: 95% CI 0.25 to 86.72, including unity. In 2014, we were made aware of the existence of three vehicle-controlled imiquimod trials, completed in 2005 by 3M Pharmaceuticals. The only publication that was available, was an executive summary produced by the FDA.
Objectives: We aimed to obtain the reports of these unpublished trials and combine the results with the study we previously included.
Methods: Several attempts were made to obtain the trial reports: contacting the Cochrane Skin Group; principal investigators of the trials; and the current proprietor of the drug.
Results: The current proprietor kindly provided us with reports of the three studies. All these were performed in pediatric patients, with a total of 827 patients, 532 of whom were assigned to imiquimod cream. The two largest studies provided data for the primary outcome of the review. Adding these two studies changed the effect to RR 0.89, and the confidence interval to 0.68 to 1.16, a reduction in size of more than 99%. The company has no plans for publishing these studies.
Conclusions: Exclusion of unpublished trials may grossly distort our view of reality.