This week in Obstetrics & Gynecology

This AudioScholar briefing is generated by artificial intelligence for healthcare professionals and trainees. It is not medical advice.

This week: doxycycline in early pregnancy and the risk of pregnancy loss, epidural analgesia and anal sphincter injury, and a new graded review of surgery for early cervical cancer.

Welcome to This Week in Obstetrics & Gynecology. We're covering 10 papers this week: 3 in depth, 4 in brief, 3 quick signals.

The first paper, in Annals of Internal Medicine, is a retrospective active-comparator cohort study led by Winterstein [1]. The team used United States commercial and Medicaid claims to study more than two hundred eighteen thousand pregnancies with a respiratory tract infection and about sixty-six hundred pregnancies with acne. First-trimester doxycycline fills were compared with amoxicillin, a macrolide or topical clindamycin, with matching on propensity score and gestational age. Among pregnancies with respiratory infection, doxycycline was associated with a roughly forty to fifty percent higher risk of pregnancy loss than amoxicillin in both databases, and a similar excess against macrolides. In the acne cohort, the excess was statistically significant in the commercial data but not in Medicaid. Where the associations were significant, the absolute risk difference was between seven and eleven percentage points. The main limitation is the claims design. Residual confounding and misclassification of exposure or outcome remain possible, and congenital malformations were not assessed. What the study adds is replication across two independent databases and two active comparators, which makes chance an unlikely explanation. The authors conclude that the association should inform how risks and benefits are weighed when doxycycline is prescribed to reproductive-aged women.

Next, a systematic review and meta-analysis in BJOG from Idzes and colleagues [2] asks whether epidural analgesia changes the risk of obstetric anal sphincter injury. The authors pooled 86 randomised and observational studies covering close to ten million women, using random-effects models and focusing on estimates adjusted for clinical confounders. In the pooled adjusted analysis, epidural use was associated with about fifteen percent lower odds of anal sphincter injury across the whole population. Among nulliparous women, the association was larger, at about twenty-three percent lower odds. The main limitation is that the evidence is largely observational. Women who receive an epidural may differ in ways that adjustment does not capture, and the authors state plainly that causality is not established. The abstract does not report results by mode of delivery or by episiotomy use. The abstract describes the prior evidence on this question as unclear. This pooled adjusted analysis points toward lower risk with epidural, most clearly in first births. The authors call for higher-quality studies to confirm whether the relationship is causal.

The third paper is a Society of Gynecologic Oncology evidence-based review in Gynecologic Oncology, led by Melamed [3], covering upfront surgery for cervical cancer. A panel addressed twelve clinical questions, synthesised the literature systematically and graded each recommendation for strength and evidence quality. For low-risk FIGO 2018 stage IA through IB1 disease, simple hysterectomy with nodal assessment is preferred. Fertility-sparing surgery is considered acceptable for selected low-risk stage IA1 through IB1 tumours, with conization or simple trachelectomy preferred over radical trachelectomy. Sentinel node biopsy may replace complete pelvic lymphadenectomy when an accepted algorithm is applied, with indocyanine green and near-infrared imaging the preferred mapping technique. When radical hysterectomy is indicated, the panel prefers the open approach outside a clinical trial. The panel does not recommend neoadjuvant chemotherapy followed by surgery in place of chemoradiation. It also does not recommend hysterectomy after definitive chemoradiation unless residual disease is documented. The limitation the authors name is uneven evidence quality, with several recommendations resting on observational data or expert consensus. What the review adds is a single graded synthesis of a decade of randomised and prospective data on extent of resection, nodal staging and surgical approach.

Now, four in brief.

In Obstetrics and Gynecology, Clapp and colleagues [4] used the Premier Healthcare Database to examine late preterm steroid use across 726 United States hospitals from 2015 to 2022, covering about three hundred ten thousand singleton late preterm births. Hospital betamethasone exposure rates ranged from zero to 92 percent. The median rate was about 15 percent in the lowest-adopting quartile of hospitals and about 54 percent in the highest. Urban hospitals had roughly double the odds of being high adopters, and Southern hospitals had about a third of the odds of Midwestern hospitals. The data are retrospective and descriptive, and the authors note that observable hospital characteristics leave much of the variation in adoption after the ALPS trial unexplained.

Two papers this week concern postmenopausal care. In Menopause, Okoror and colleagues [5] report a retrospective cohort of 1,082 women with postmenopausal bleeding on an urgent cancer pathway at a single United Kingdom district hospital. Endometrial thickness discriminated cancer or atypical hyperplasia very well in women not using hormone therapy, with an area under the curve of about 0.93. In women on continuous combined hormone therapy, discrimination was modest, at about 0.6. At a 4 millimetre threshold, negative predictive value exceeded 99 percent in both groups, while positive predictive value stayed low in hormone users. With only 32 cases of significant pathology from one centre, the subgroup estimates are imprecise.

The second, a network meta-analysis in BJOG by Randombage and Akurugodagama [6], compared treatments for genitourinary syndrome of menopause across 24 randomised trial families and almost five thousand women. On patient-reported symptom burden, local vaginal oestrogen was the only class that clearly outperformed placebo or sham, with a small-to-moderate effect. Prasterone, hyaluronic acid, ospemifene, moisturisers and CO2 laser all had imprecise estimates that did not exclude no effect, and CO2 laser sat closest to zero. Confidence was low for oestrogen and prasterone and very low for every other comparison, and eleven of the trial families carried high risk of bias. The authors conclude that laser and radiofrequency are not supported as equivalent first-line options.

In Obstetrics and Gynecology, Graham McNeil and colleagues [7] recontacted 817 women from two procedure studies who had undergone myomectomy, endometrial ablation, laparoscopic radiofrequency ablation or uterine artery embolization for fibroids, with four to ten and a half years of follow-up. About four in five women had no further uterine procedure at five years, and just under two thirds of women had none at ten years. Radiofrequency ablation carried about a sixty percent higher retreatment risk than myomectomy. Subsequent hysterectomy was least common after myomectomy and most common after endometrial ablation. Retreatment was self-reported and procedures were not randomised, so the between-procedure comparisons remain observational.

Three quick signals.

In Gynecologic Oncology, Zitter and colleagues [8] analysed about forty-six hundred patients with surgically treated stage III to IVA uterine carcinosarcoma in the National Cancer Database. Sequential chemoradiation was associated with about thirty percent lower mortality than chemotherapy alone, while concurrent chemoradiation was not. These are registry data open to selection bias, and the authors call for prospective trials.

In Fertility and Sterility, a narrative review by Cozzolino and colleagues [9] reports that the only randomised trial and meta-analyses restricted to frozen transfer showed no live birth benefit from gonadotropin-releasing hormone agonist pretreatment in adenomyosis. Letrozole matched the agonist in one trial, and the overall certainty of evidence is low.

In Ultrasound in Obstetrics and Gynecology, Lesieur and colleagues [10] report an international Delphi consensus of 48 experts that classified 35 of 48 fetal brain variants. Isolated small choroid plexus cysts were rated benign, and abnormal posterior fossa markers warranted full investigation. The classification has not been validated against postnatal outcomes.

Stepping back across this week's papers, most of the evidence rests on claims cohorts, registry analyses, meta-analyses built largely on observational studies, expert panels and narrative synthesis. Where randomised data do exist, they are small, imprecise and often at high risk of bias, and they tend to temper enthusiasm for newer options. Several of the clearest associations this week still await trials or validation studies that could establish whether they are causal.

If you only have time for one paper this week, make it the doxycycline cohort study [1]. It fills a long-standing gap in pregnancy safety data with consistent signals from two large databases, and it reopens the question of where doxycycline sits among antibiotic options for women in early pregnancy.

That's your roundup for This Week in Obstetrics & Gynecology. The full transcript and references are available on the episode page. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And a quick rating on Apple Podcasts or Spotify helps other physicians discover the show.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.