Background: Missing outcomes and noncompliance are complications that most randomized controlled trials suffer after randomization and during a treatment. This problems ends up creating groups that can not be compared, leading to a risk of bias. One method used to estimate the effect of the intervention in randomized samples, regardless of whether they have finished treatment, it's called intention to treat analysis (ITT). Therefore, as assisted reproduction has been gaining prominence and consequently, the search for new treatments has increased, this article will report the number of randomized clinical trials with that theme that has performed the intention to treat analysis.

Objectives: To evaluate the quality of description of ITT analysis in RCT about female infertility.

Methods: We performed a search strategy in Medline via Pubmed to identify RCTs published about female infertility. The identified studies were selected by two researchers using Rayyan including RCTs that mentioned ITT analysis.

Results: 68 RCTs were included. Of these, 68% clearly stated the analysis of all randomized participants (CI 95%, 56% to 78%). A total of 34% of included RCTs mentioned the use of “per protocol” analysis (CI 95%, 24% to 46%) with only one detailing if “as treated” or “naive per protocol”. Four percent of included RCTs reported the use of a modified ITT analysis.

Conclusion: There is a relevant proportion of RCTs that do not report properly information regarding intention to treat analysis.