Feng Qian, Li Wanlin, Crispin James, Longobardi Salvatore, D'Hooghe Thomas, Mol Ben W, Li Wentao
Department of Obstetrics and Gynaecology, Monash University, Clayton, VIC, Australia.
Global Clinical Development Fertility, Research and Development, Merck, Darmstadt, Germany.
Hum Reprod Open. 2025 Jan 24;2025(1):hoaf004. doi: 10.1093/hropen/hoaf004. eCollection 2025.
What are the trial characteristics, geographic distribution, and selected methodological issues of randomized controlled trials (RCTs) in infertility published from 2012 to 2023?
Of the 1425 infertility RCTs, over two-thirds focused on IVF, nearly two-fifths did not use pregnancy or live birth as the primary outcome, a third lacked a primary outcome, a half were unregistered, and just over half were conducted in China (22%), Iran (20%), or Egypt (10%).
RCTs are the main source of evidence on the effectiveness of interventions. Knowledge about RCTs in infertility from the recent past will help to pinpoint research gaps and prioritize the future research agenda. Here, we aim to present a descriptive analysis of trial characteristics, geographic distribution, and selected methodological issues in infertility trials published in the last decade.
This is a systematic review. We systematically searched Embase, Medline, and Cochrane Central for RCTs in infertility from January 2012 to August 2023. RCTs involving subfertile women and women who reported pregnancy endpoints were eligible, while conference abstracts or secondary analyses were not. We did not limit our search based on the language of the articles.
PARTICIPANTS/MATERIALS SETTING METHODS: The full articles were text-mined and manually extracted for the description of trials' characteristics (e.g. sample size, blinding method, types of intervention), the country where the patients were recruited, and methodological issues (trial registrations and specification of primary outcomes). We extracted funding statements from Dimensions, a literature database chosen for its comprehensive and robust metadata. Gross domestic product (GDP) data were obtained from the United Nations' official website. The accuracy of extracted data was validated in a random sample of 50 articles, and false positivity and false negativity were all at or below 8%. We used descriptive statistics, including frequencies and percentages to illustrate the overall and temporal trends.
Among 8757 records, we found 1425 eligible RCTs, with a median sample size of 140, and 33.3% had a sample size <100. Most (69.6%) of the trials focused on IVF, with the rest focusing on ovulation induction (12.4%), intrauterine insemination (10.6%), surgeries (4.8%), or other interventions (2.6%). Regarding the geographic distribution, China (n = 310), Iran (n = 284), and Egypt (n = 138) contributed to 51% of the RCTs, followed by Turkey (n = 82), India (n = 71), and the USA (n = 69); mainland Europe produced 343 trials. Ranked by publications of trials per trillion GDP, Greece had the most papers with 4.6, followed by Iraq at 3.9, and Iran at 2.5. Regarding trial registration, 47.8% of trials were unregistered, the proportion of studies that were unregistered halved from 70.0% in 2012 to 34.6% in 2022. Of all RCTs, 37.6% had primary outcomes unspecified; the proportion of trials specifying primary outcomes increased from 49.5% in 2012 to 61.4% in 2022. The proportion of trials which declared receiving no funding was 76.9%.
We primarily used text mining for data extraction. Despite optimizing the algorithm to identify all outcome definitions and manually curating the extracted data, there were inaccuracies in data extraction; however, the false positivity and false negativity of data extraction were all at or below 8%. Also, we focused on trials reporting pregnancy outcomes, as these are of primary interest to patients and carry significant implications on clinical practice. However, we acknowledge that early-stage trials with only upstream endpoints also play an important role and should be considered when evaluating the full spectrum of infertility trials. Finally, we only included published RCTs and hence, our results cannot be extrapolated to unpublished RCTs.
The domination of RCTs on IVF calls for a reconsideration of other topics to be studied and a realignment of research priorities. The imbalanced geographic distribution of infertility trials raises questions about the generalizability of study results and equity in the distribution of healthcare resources. The prevalence of trials without registration or primary outcomes specified highlights the imperative to improve trial design and reporting quality. Encouragingly, the improving trial registrations suggest the enforcement of trial registrations from the journals is effective.
STUDY FUNDING/COMPETING INTERESTS: B.W.M. is supported by an NHMRC Investigator grant (GNT1176437). W.T.L. is supported by an NHMRC Investigator grant (GTN2016729). W.L.L. reports receiving a PhD scholarship from the China Scholarship Council. Q.F. reports receiving a PhD scholarship from Merck. B.W.M. reports receiving consultancy fees, travel support, and research funding from Merck; consultancy fees from Organon and Norgine; and stock ownership in ObsEva. T.D.H and S.L. are employees of Merck. W.T.L., W.L.L., and J.C. report no conflicts of interest.
PROSPERO CRD42024498624.
2012年至2023年发表的不孕症随机对照试验（RCT）的试验特征、地理分布及选定的方法学问题有哪些？
在1425项不孕症RCT中，超过三分之二聚焦于体外受精（IVF），近五分之二未将妊娠或活产作为主要结局，三分之一缺乏主要结局，一半未注册，略超一半在中国（22%）、伊朗（20%）或埃及（10%）开展。
RCT是干预措施有效性证据的主要来源。了解近期不孕症RCT情况有助于明确研究差距并确定未来研究议程的优先级。在此，我们旨在对过去十年发表的不孕症试验的试验特征、地理分布及选定的方法学问题进行描述性分析。
研究设计、规模、持续时间：这是一项系统评价。我们系统检索了Embase、Medline和Cochrane Central，以查找2012年1月至2023年8月期间的不孕症RCT。涉及亚生育力女性和报告妊娠结局的女性的RCT符合条件，而会议摘要或二次分析不符合。我们未基于文章语言限制检索。
参与者/材料、设置、方法：对全文进行文本挖掘并手动提取，以描述试验特征（如样本量、盲法、干预类型）、患者招募国家以及方法学问题（试验注册和主要结局的明确）。我们从Dimensions（一个因其全面且强大的元数据而被选用的文献数据库）中提取资助声明。国内生产总值（GDP）数据从联合国官方网站获取。在50篇文章的随机样本中验证了提取数据的准确性，假阳性和假阴性均在8%或以下。我们使用描述性统计，包括频率和百分比来说明总体和时间趋势。
在8757条记录中，我们发现1425项符合条件的RCT，样本量中位数为140，33.3%的样本量<100。大多数试验（69.6%）聚焦于IVF，其余聚焦于促排卵（12.4%）、宫内人工授精（10.6%）、手术（4.8%）或其他干预（2.6%）。关于地理分布，中国（n = 310）、伊朗（n = 284）和埃及（n = 138）占RCT的51%，其次是土耳其（n = 82）、印度（n = 71）和美国（n = 69）；欧洲大陆有343项试验。按每万亿GDP的试验发表量排名，希腊最多，为4.6篇，其次是伊拉克，为3.9篇，伊朗为2.5篇。关于试验注册，47.8%的试验未注册，未注册研究的比例从2012年的70.0%降至2022年的34.6%。在所有RCT中，37.6%未明确主要结局；明确主要结局的试验比例从2012年的49.5%增至2022年的61.4%。声明未获得资助的试验比例为76.9%。
局限性、谨慎理由：我们主要使用文本挖掘进行数据提取。尽管优化了算法以识别所有结局定义并手动整理提取的数据，但数据提取仍存在不准确之处；然而，数据提取的假阳性和假阴性均在8%或以下。此外，我们关注报告妊娠结局的试验，因为这些是患者的主要关注点且对临床实践有重大影响。然而，我们承认仅具有上游终点的早期试验也起着重要作用，在评估不孕症试验的全貌时应予以考虑。最后，我们仅纳入已发表的RCT，因此，我们的结果不能外推至未发表的RCT。
RCT对IVF的主导要求重新考虑其他待研究的主题并重新调整研究重点。不孕症试验地理分布不均衡引发了关于研究结果的可推广性以及医疗资源分配公平性的问题。未注册或未明确主要结局的试验普遍存在凸显了改善试验设计和报告质量的紧迫性。令人鼓舞的是，试验注册情况的改善表明期刊对试验注册的执行是有效的。
研究资金/利益冲突：B.W.M. 得到澳大利亚国家卫生与医学研究委员会（NHMRC）研究员资助（GNT1176437）。W.T.L. 得到NHMRC研究员资助（GTN2016729）。W.L.L. 报告获得中国国家留学基金管理委员会的博士奖学金。Q.F. 报告获得默克公司的博士奖学金。B.W.M. 报告从默克公司获得咨询费、差旅支持和研究资金；从欧加农和诺金公司获得咨询费；以及在ObsEva公司拥有股份。T.D.H和S.L. 是默克公司的员工。W.T.L.、W.L.L. 和J.C. 报告无利益冲突。
PROSPERO CRD420且24498624 。
