体外受精（IVF）是一个充满失败的过程，大约60%的尝试都会以失败告终。为了提高成功率，生育诊所通常会推广各种被称为“IVF附加项目”的测试和程序。然而，6月23日发表在《柳叶刀·产科学、妇科学与女性健康》上的一项全面证据审查评估了157项随机试验，发现没有令人信服的证据表明这些附加项目中的大多数是有效的。此外，该审查还揭示了对已发表科学文献完整性的重大担忧，在被评估的试验中，近一半未能通过旨在检测数据操纵和欺诈的清单。

在被认为可靠的85项试验中，研究人员评估了十种常用的IVF附加项目。只有三种程序显示出有益的证据，尽管支持证据相对较弱。这些程序包括子宫内膜搔刮术（涉及故意干扰子宫内膜）、胚胎胶（将胚胎浸泡在透明质酸溶液中）以及生理性单精子显微注射（PICSI，旨在选择高质量的精子）。相反，另外三种附加项目——皮质类固醇、胚胎染色体异常基因检测和子宫内膜活检——显示对妊娠成功率没有可衡量的影响。

关于其余四种程序的数据（包括针灸、静脉注射源自鸡蛋或大豆的脂肪，以及子宫或卵巢注射富含血小板的血浆）过于稀少，无法做出明确的结论。审查强调，评估的试验大多规模较小，中位样本量仅为160名患者，这限制了它们检测微弱益处的能力。虽然研究人员指出某些干预措施可能仍会使特定的患者群体受益，但确定这些群体将需要更严格、更大规模的研究。此外，学术期刊必须实施更严格的审查程序，以防止生殖医学领域发表欺诈性或被操纵的数据。

In vitro fertilisation (IVF) is a process fraught with failure, as approximately 60% of attempts do not succeed. To improve outcomes, fertility clinics often promote a wide variety of tests and procedures known as "IVF add-ons." However, a comprehensive review of evidence published on June 23rd in the Lancet Obstetrics, Gynaecology & Women’s Health evaluated 157 randomised trials and found no convincing proof that the majority of these additions are effective. Furthermore, the review exposed significant concerns regarding the integrity of the published scientific literature, with nearly half of the analysed trials failing to pass a checklist designed to detect data manipulation and fraud.

For the 85 trials that were deemed reliable, researchers evaluated ten common IVF add-ons. Only three procedures demonstrated any evidence of benefit, though the supporting evidence was relatively weak. These included endometrial scratching, which involves disturbing the uterine lining; EmbryoGlue, where the embryo is dipped in a hyaluronic acid solution; and physiological intracytoplasmic sperm injection (PICSI), which aims to select high-quality sperm. Conversely, three other add-ons—corticosteroids, genetic testing of embryos for abnormal chromosomes, and uterine lining biopsies—showed no measurable effect on pregnancy success rates.

Data for the remaining four procedures, including acupuncture, intravenous fat infusions, and platelet-rich plasma injections, were too scarce to reach a definitive conclusion. The review highlighted that the evaluated trials were mostly small, with a median size of only 160 patients, which restricts their power to detect subtle benefits. While the researchers note that certain interventions might still benefit specific subsets of patients, identifying these individuals will require more rigorous, large-scale studies. Additionally, scientific journals must implement stricter vetting processes to prevent the publication of fraudulent or manipulated data in reproductive medicine.

Source: Do high-tech “add-ons” increase the chance that IVF will work?

Subtitle: A big new review suggests most such procedures don’t help

Dateline: 6月 25, 2026 09:02 上午