A systematic review and quality assessment of systematic reviews of randomised trials of interventions for preventing and treating preterm birth

Abstract

The aim of this paper is to identify reviews of interventions for preventing and treating preterm birth so that these could be appraised and the findings from good quality reviews highlighted. Reviews, rather than individual studies, are the basis for this systematic review because of the proliferation of reviews and the benefits of a single, consistent appraisal and assessment of evidence from these reviews rather than further attempts to find and appraise the many individual studies in the literature. Our systematic review consists of a description of five interventions for preventing and treating preterm birth; antibiotics, cervical cerclage, bed rest, progesterone, and tocolytic therapy, for which at least one relevant review was found. The scope and quality of the identified reviews are described, and their conclusions and the strength of these conclusions discussed. Potentially eligible reviews were sought primarily through searches of the electronic databases MEDLINE (1966-2008), EMBASE (1980-2008), CINHAL (1982-2008), Science Citation Index (1970-2008) and The Cochrane Library (Issue 1, 2008). Thirty-seven reviews were identified of which 22 were included in this systematic review of reviews. This shows that antibiotics may significantly delay, but might not prevent, preterm birth for women with preterm prelabour rupture of membranes; there is insufficient evidence to show the absolute efficacy of cerclage and bed rest in preventing preterm birth; the use of progesterone appears promising; and the possible benefits of certain tocolytics, such as beta-mimetics, need to be reliably measured against the possible adverse effects when used in preventing preterm birth.