Background: the exhaustive identification of controlled clinical trials (CCT) is a crucial step during systematic review development. Cochrane recommends implementing a multi-pronged approach including electronic and handsearching of different sources. Handsearching is an invaluable complement to electronic searches, since the latter may be limited by shortcomings in indexation. In addition, journals published in languages other than English are less likely to be indexed in the main literature databases. Handsearching, however, is not always feasible. Thus, the Iberoamerican Cochrane Centre leads an initiative that aims to handsearch CCTs in Spanish-language journals.
Objectives: we aimed to identify CCTs published in gynecology and obstetrics journals from the five countries in the Andean region (Colombia, Venezuela, Ecuador, Peru and Bolivia), as well as to describe their main characteristics, quality of reporting, and risk of bias.
Methods: we handsearched, following Cochrane guidelines, all eligible journals. This search was complemented with electronic searches in journals indexed in MEDLINE. We assessed the quality of reporting with CONSORT, while we assessed methodological quality with an adapted version of the Cochrane 'Risk of bias' tool. We sent the identified studies to CENTRAL (Cochrane Central Register of Controlled Trials), Cochrane’s global CCT repository.
Results: we identified seven eligible journals, which published a total of 25 CCTs. In general, the identified CCTs had a high risk of bias, mainly due to shortcomings in allocation concealment or blinding of participants, researchers or evaluators. Most study authors did not report prospective registration of protocols or sources of funding.
Conclusions: the number of CCTs published in gynecology and obstetrics journals from the Andean area is low. We identified weaknesses in the design and reporting of identified CCTs. We recommend adherence to CONSORT and other methodological guidelines when designing and reporting CCTs.
Patient or healthcare consumer involvement: none