The quality and utility of randomised controlled trials in in vitro fertilisation: an analysis of the published literature

Abstract

Study question: What is the quantity, quality and utility of research on in vitro fertilisation (IVF) and intracytoplasmic sperm injection (ICSI) in the last decade? Summary answer: Current IVF/ICSI research is of moderate quality, and can be improved by widening inclusion criteria for patient selection, and reporting live-birth as a cumulative measure. What is known already: Assisted reproductive technology (ART) is the cornerstone of fertility treatment worldwide for male, female and unexplained infertility. In the last few decades, ART and IVF techniques and protocols have changed considerably, guided by research in the field. Many researchers and journals worldwide continue to publish on ART and IVF yearly, to varying qualities and levels of evidence. Our study aims to analyse the published IVF literature over the last decade and comment on its quality, utility and its feasibility in translating the results into clinical practice. Study design, size, duration: We conducted a meta-epidemiological study based on a PubMed search of IVF randomised controlled trials (RCTs) published from 2010 to the present. All articles returned from the initial search were screened by two researchers for relevance, and any conflicts were resolved by a third researcher. We used newly-developed checklists to score usefulness of research, analysing parameters including participant baseline characteristics reported, participant inclusion/exclusion criteria, main study outcomes reported, and whether sample sizes were statistically significant. Participants/materials, setting, methods: We selected full text Englishlanguage articles, presenting primary data. We included studies comparing either two modalities of IVF, comparing IVF to another ART intervention, or those evaluating the effectiveness of IVF adjunct therapies. Studies focusing on embryo transfer (ET) and ICSI were included, as well as studies focusing on treating complications of IVF such as ovarian hyperstimulation syndrome (OHSS). Studies focusing on intrauterine insemination (IUI), psychosocial outcomes of IVF and lifestyle modifications were excluded. Main results and the role of chance: Our initial search returned 5,577 articles. Subsequently, we applied filters for publication date (01/01/2010 to 23/08/2021), Human species, and article types (clinical trials, controlled clinical trials, and randomised clinical trials). The remaining 1,290 articles were manually screened for relevance and adherence to inclusion criteria, leaving 631 RCTs. For this abstract, we analysed the 163 RCTs published in the years 2010, 2015 and 2020, and found that the annual number of publications remained relatively stable (53, 56 and 54, respectively). Regarding participants' baseline characteristics, 80/163 (49.1%) of studies stated the duration of infertility; only 3/163 (1.8%) of studies stated recurrent miscarriage rates. In terms of study outcomes, 78/163 (47.9%) of studies reported live birth rate (LBR) as an outcome, however only 13/163 (8.0%) of studies reported cumulative LBR, with the remaining studies reporting LBR per first cycle or first transfer. When analysing the participant inclusion, 63/163 (38.7%) of studies had number of cycles as a criterion, with 13/63 (20.6%) of those only including participants in their first cycle. Additionally, 57/163 (35.0%) of studies stated BMI as an inclusion criterion, with 36/57 (63.2%) of those only including participants with BMI of 30 or less. Limitations, reasons for caution: Due to practicality and feasibility reasons, we were unable to analyse articles not in English, thus overlooking research published in many parts of the world. Additionally, some parameters we scored are subjective in nature. Moreover, journals have discrepancies in their submission requirements, which may also affect our analysis outcomes. Wider implications of the findings: IVF research has many areas for improvement. Of interest are the inclusion criteria for participants: a significant proportion of studies required patients to have a normal BMI, and only included women undergoing their first IVF cycle. A widening of these criteria would increase the applicability of research to clinical practice.