Archives of Gynecology and Obstetrics最新文献

Purpose

Surgical-site infections (SSI) remain a major complication of cesarean delivery, increasing maternal morbidity, hospital stay, and healthcare costs. Although several risk factors have been identified, the role of maternal nutritional status—particularly the Prognostic Nutritional Index (PNI)—has not been well established in obstetric surgery.

Methods

This retrospective case–control study included 190 women who underwent cesarean delivery at a tertiary referral center between 2020 and 2025. Preoperative PNI was calculated using serum albumin and lymphocyte counts obtained within 24 h before surgery. SSI was defined according to the Centers for Disease Control and Prevention criteria. Demographic, perioperative, and postoperative variables were analyzed to evaluate associations between PNI, SSI status, infection severity, and inflammatory markers.

Results

A total of 190 women were analyzed, including 98 cases with surgical-site infection (SSI) and 92 controls without SSI. Although preoperative PNI was not an independent predictor of SSI, lower PNI values were associated with more severe infections requiring broad-spectrum antibiotics, longer hospitalization, and higher postoperative C-reactive protein (CRP) levels. A significant inverse correlation was observed between PNI and CRP among patients with SSI (r =  − 0.338, p = 0.001). Longer operative duration and smoking were also associated with an increased risk of SSI.

Conclusion

Although PNI was not an independent predictor of SSI, lower values were strongly associated with infection severity and systemic inflammation. Integrating PNI into preoperative risk assessment—alongside body mass index and operative time—may help identify high-risk women. Larger prospective studies are warranted to confirm these findings and to evaluate the impact of nutritional optimization on postoperative outcomes.

Background

To investigate whether cured tuberculosis affect assisted reproductive outcomes in infertile women.

Methods

A retrospective cohort study was conducted to compare assisted reproductive outcomes between cured tuberculosis and non-tuberculosis patients. The primary outcomes were compared between fresh and frozen–thaw cycles in cured tuberculosis, as well as among different types of tuberculosis in fresh cycles.

Results

In IVF and total patients, the pregnancy (P > 0.05) and live birth rates (P > 0.05) did not differ between tuberculosis and control group. However, in ICSI protocol, the pregnancy (P = 0.027) and live birth rates (P = 0.027) in tuberculosis group were lower than those in controls. The pregnancy rate in fresh cycles for tuberculosis patients was lower than in frozen–thaw cycles in each protocol (P = 0.001). The live birth rate in fresh cycles was higher than in frozen–thaw cycles in IVF patients (P = 0.008) and total patients (P = 0.015), while the live birth rate in fresh cycles was lower in ICSI patients (P = 0.011). The pregnancy rate (P > 0.05) and live birth rate (P > 0.05) among cured patients diagnosed with pelvic tuberculosis demonstrated no significant difference compared to those with pulmonary tuberculosis or other types.

Conclusions

Cured tuberculosis did not affect the primary assisted reproductive outcomes in infertile patients using IVF protocol. However, in ICSI protocol, tuberculosis had a negative impact on pregnancy outcomes in infertile women, even with treatment. The final pregnancy outcome of frozen–thaw cycle was worse than that of fresh cycle in IVF protocol but opposite in ICSI. If treated, there were no variations in pregnancy outcomes among different tuberculosis types.

Objective

The main aim of this study is to evaluate value of maternal serum nuclear factor erythroid 2-related factor 2 (NRF2) levels at 12 weeks in predicting the development of gestational diabetes mellitus (GDM) at 24–28 weeks' gestation. Other maternal variables were also evaluated, and their relationships with NRF2 levels were assessed.

Methods

We conducted a single-center prospective cohort study including 1,270 pregnant women who attended their first-trimester antenatal visit between October 2021 and October 2023. At 12 weeks, fasting serum was collected to measure NRF2 by enzyme-linked immunosorbent assay (ELISA) and thyroid hormones on an automated chemiluminescent platform. Clinical data included age, pre-pregnancy body mass index (BMI), blood pressure, fasting plasma glucose, and lipids. All participants underwent a 75-g oral glucose tolerance test (OGTT) at 24−28 weeks and GDM was diagnosed by International Association of the Diabetes and Pregnancy Study Groups (IADPSG) criteria. Spearman correlation assessed associations between NRF2 and clinical variables. Receiver operating characteristic (ROC) analysis evaluated discrimination. Multivariable logistic regression identified independent predictors.

Results

Among 1,270 pregnant women enrolled, 177 (13.9%) developed GDM. At 12 weeks of gestation, women who later developed GDM had significantly lower serum NRF2 levels than those without GDM. They also showed higher fasting plasma glucose (FPG), mean arterial pressure (MAP), as well as lower free thyroxine (FT4) levels. ROC analysis demonstrated that serum NRF2 at 12 weeks had good predictive value for GDM, with an area under the curve (AUC) of 0.779 (95% confidence interval [CI] 0.745–0.812, p < 0.001). A combined model incorporating NRF2 with MAP, FT4, and FPG showed improved discrimination, with an AUC of 0.882 (95% CI 0.856−0.908, p < 0.001). In multivariable logistic regression, lower serum NRF2 (odds ratio [OR] = 0.948, 95% CI 0.938–0.957, p < 0.001), higher MAP (OR = 1.182, 95% CI 1.132–1.233, p < 0.001), higher FPG (OR = 3.911, 95% CI 2.471−6.190, p < 0.001), and lower FT4 (OR = 0.639, 95% CI 0.562–0.726, p < 0.001) were identified as independent predictors of GDM, whereas other baseline parameters were not significant.

Conclusion

Decreased maternal serum NFR2 levels at 12 weeks’ gestation appear to be associated with an increased risk of developing GDM later in pregnancy. Combining this marker, together with FT4, MAP and potentially FPG may serve as a useful first-trimester screening for the risk to develop GDM.

Purpose

As artificial intelligence (AI) models evolve into their next generations, their application in specialized medical fields requires rigorous validation. While large language models (LLMs) have shown promise in general medicine, their reliability in complex gynecological clinical reasoning remains under-explored. This pilot study aimed to comparatively assess the knowledge retention, safety, and reasoning limitations of advanced AI chatbots in gynecology using a constrained zero-shot multiple-choice question (MCQ) format.

Methods

A total of 70 text-based MCQs covering seven core gynecological modules were adapted from "USMLE Step 1 Sample Test Questions". The questions were administered to four advanced AI models: ChatGPT-5, Gemini-3, DeepSeek-V3.2, and Claude-4.5-Opus. To simulate a rapid-retrieval clinical scenario, models were tested under "zero-shot" conditions with a constrained prompt prohibiting reasoning steps. We performed both quantitative statistical analysis (Kruskal–Wallis, Cochran’s Q) and qualitative error analysis to identify specific failure modes.

Results

Contrary to expectations for advanced models, overall accuracy was unsatisfactory: Gemini-3 (32.86%), DeepSeek-V3.2 (30.00%), ChatGPT-5 (25.71%), and Claude-4.5-Opus (21.43%). Significant performance disparities were observed across modules. Notably, ChatGPT-5 scored 0.00% in Infertility, while DeepSeek-V3.2 reached 70.00% in Common Benign Conditions. Qualitative analysis revealed three critical failure patterns: (1) semantic association bias (confusing high-probability diseases with symptom-specific diagnoses), (2) spatial anatomy confusion, and (3) genetic logic reversal. No significant correlation was found between item difficulty and accuracy (p > 0.05).

Conclusion

Under constrained non-reasoning prompts, even next-generation AI chatbots demonstrate unsatisfactory performance in gynecology. The qualitative analysis suggests that models often rely on probabilistic keyword matching rather than physiological simulation, leading to theoretically dangerous clinical errors (e.g., misdiagnosing adrenal enzymes). While potential exists, current reliability is insufficient for unsupervised use in gynecological education. These findings highlight the critical need for "chain-of-thought" prompting and human expert oversight.

Background

Fetal growth restriction (FGR) is a major contributor to perinatal morbidity and mortality. While guidelines address timing of delivery, the optimal mode—induction of labor (IOL) versus planned cesarean delivery (CD)—remains uncertain.

Objective

To evaluate the association between planned mode of delivery and neonatal outcomes in pregnancies complicated by late onset FGR (LOFGR).

Study Design

We conducted a retrospective cohort study at a tertiary Canadian center (2017–2022). Singleton pregnancies with LOFGR (> 34 weeks’ gestation), defined by Society for Maternal–Fetal Medicine (SMFM) criteria, were eligible if the last ultrasound was within 14 days of delivery. Exclusions included spontaneous labor, delivery < 34 weeks, and contraindications to labor. Planned mode of delivery (IOL vs CD) was the exposure. Outcomes were classified as severe (perinatal death, 5-min Apgar < 4, umbilical arterial pH < 7.05, base deficit ≥ 12 mmol/L, hypoxic-ischemic encephalopathy/therapeutic hypothermia, grade III–IV intraventricular hemorrhage, necrotizing enterocolitis, sepsis, or invasive ventilation > 24 h) or moderate (NICU stay > 72 h, Apgar 4–6, pH 7.05–7.10, non-invasive respiratory support > 6–12 h, transient tachypnea, or brief resuscitation). Multivariable logistic regression adjusted for confounders. A prespecified subgroup applied the ISUOG criteria.

Results

Of 12,270 deliveries, 1,143 (9.3%) met SMFM criteria for LOFGR; 869 were eligible (192 planned CD, 677 IOL). Severe outcomes and moderate outcomes were more frequent after CD (23.4% vs 16.7%; p = 0.03 and 42.2% vs 31.2%; p < 0.01, respectively). IOL was associated with lower adjusted risk of severe outcomes (aOR 0.35; 95% CI 0.19–0.67) and moderate outcomes (aOR 0.43; 95% CI 0.24–0.76). Results were consistent using ISUOG criteria (aOR 0.33; 95% CI 0.17–0.62 and aOR 0.44; 95% CI 0.25–0.79, respectively) About 20% of induced patients required intrapartum CD.

Conclusions

IOL was associated with reduced severe and moderate neonatal morbidity compared with planned CD. IOL represents a safe alternative when intrapartum surveillance and timely operative delivery are available.

Background

Preterm birth is the leading cause of perinatal morbidity and mortality worldwide, with about 15 million preterm births occurring each year. Accurate prediction of preterm birth is essential for the implementation of effective interventions.

Aim

This study evaluates the combined predictive value of uterine electromyography (EMG) and transvaginal ultrasound (TVUS) cervical parameters for preterm birth in high-risk singleton pregnancies, assessing its superiority over individual methods.

Methods

A retrospective, propensity score-matched cohort study was conducted to collect data of pregnant women at high risk of preterm birth who gave birth in our hospital (Jan 2024–Apr 2025). After 1:1 matching, 180 participants were assigned to spontaneous preterm birth (sPTB, < 37 weeks, n = 90) or term birth (≥ 37 weeks, n = 90) groups. At 20–24 weeks, all underwent TVUS [cervical length (CL) and elastic strain rate (ESR)] and EMG [peak frequency (PK) and propagation velocity (PV)]. The primary aim was to compare the AUC of a combined model (CL + ESR + EMG–PV) versus CL alone. Secondary outcomes included group differences and each model’s sensitivity, specificity, and NPV. Logistic regression was used for multivariable analysis, and the DeLong test compared AUCs.

Results

After PSM, the baseline data of the two groups were balanced (P > 0.05). The sPTB group had shorter CL and higher ESR, EMG–PK, and EMG–PV (all P < 0.001). Multivariate logistic regression identified CL (OR = 0.85), ESR (OR = 1.32), and EMG–PV (OR = 1.45) as independent predictors. The combined model (CL + ESR + EMG–PV) yielded a significantly higher AUC (0.93, 95%CI 0.89–0.97) than CL alone (0.85, 95%CI 0.79–0.91; DeLong test, P = 0.002). The combined model also demonstrated superior sensitivity (88.9%), specificity (91.1%), and NPV (94.4%) compared with individual models.

Conclusions

Integrating TVUS cervical assessment with EMG monitoring significantly enhances preterm birth prediction in high-risk singletons, offering improved risk stratification despite limitations from its single-center retrospective design requiring future validation.

Objective

To identify trajectories of maternal–fetal attachment (MFA) in IVF mothers during the first trimester of pregnancy, and to explore categorical predictors of subgroups demonstrating MFA trajectories.

Methods

A prospective observational study was conducted on 495 IVF mothers. Self-report questionnaires assessing MFA, anxiety, depression, social support, and marital satisfaction were completed at three time points: 3–4 weeks (T1), 7–8 weeks (T2), and 12–14 weeks (T3) of gestation. Repeated measurement data were analyzed and growth trajectories identified.

Results

Three MFA trajectories were delineated in IVF mothers during the first trimester: low-attachment/slow-elevation group (30.9%), moderate-attachment/steady-elevation group (56.2%), and high-attachment/rapid-elevation group (12.9%). Secondary infertility (OR = 0.575, 95% CI: 0.368–0.901, p = 0.013) and higher marital satisfaction (OR = 0.941, 95% CI: 0.915–0.968, p < 0.001) were predictors for not being in the low-attachment/slow-elevation group.

Conclusion

Among IVF mothers with ongoing pregnancies, a subset exhibited suboptimal development of MFA during the first trimester. Marital satisfaction and infertility type may help identify those at potential risk.

Background

Platelet-rich plasma (PRP) is used as an adjunctive treatment in various medical fields, including orthopaedics, plastic surgery, dermatology, and wound healing. However, PRP application in reproductive medicine, particularly for fertility restoration, remains controversial, specifically for women with diminished ovarian reserve (DOR), premature ovarian insufficiency (POI), and pre-menopause. This systematic review aimed to comprehensively evaluate the existing literature on autologous PRP use for ovarian rejuvenation, with a focus on refining and standardising PRP protocols and identifying appropriate candidate populations.

Methods

A comprehensive literature search was conducted on the autologous PRP use using Cochrane, Embase, PubMed, MEDLINE, NCBI, and Google Scholar databases. Studies on PRP application in reproductive medicine were identified and classified into two categories: DOR and POI. The PRP preparation protocol was assessed, and the results were compared. This narrative systematic review, carried out in compliance with PRISMA 2020 guidelines, combined results, although meta-analysis was not feasible due to significant variation in study design, PRP preparation, and outcome measures.

Results

Several studies reported restoration of menstruation following intraovarian PRP administration in women with POI. In women with DOR, studies reported changes in ovarian reserve markers, including increases in serum anti-Müllerian hormone (AMH) levels and reductions in follicle-stimulating hormone (FSH) levels; however, these changes were not consistently associated with higher antral follicle counts or oocyte yield. Instead, sseveral studies reported higher proportions of mature oocytes and pregnancy and live birth outcomes, particularly following repeated PRP treatment cycles in women with long-standing ovarian dysfunction.

Conclusions

Intraovarian PRP may have a potential role in activating dormant follicles in a selected subset of women with poor ovarian response. However, the current evidence remains limited and heterogeneous, supporting its use only within structured clinical or research protocols until higher-quality, standardised studies are available.

Objective

This analysis aimed to characterize exceptional responder with long-term PARP inhibitor therapy (ExR-LT) in platinum-sensitive recurrent ovarian cancer.

Methods

This analysis included ExR-LT. ExR-LTs are defined as patients that received a continuous maintenance therapy for recurrent ovarian cancer with olaparib or niraparib for at least 5 years and showed an exceptional response. Exceptional response was defined as progression-free survival (PFS) of at least 5 years. This analysis has a retrospective and descriptive character.

Results

23 patients were included. The median duration of PARPi therapy was 7.1 years (range 5.3; 10.5). The longest treatment duration was reached in the BRCA1 mutation (BRCA1m) cohort with a mean duration of 8 years (range 5.3; 10.5 years). The majority of patients (16 patients, 69.7%) reported adverse events (AE) during PARPi therapy. 12 patients (52.2%) had mild AE (CTCAE 1 or 2), 4 patients (17.4%) reported more severe AE (CTCAE 3). 14 patients needed a dose reduction due to treatment-related AE (60.1%). The most common indications for dose reduction were anemia (17.4%), headache and limb pain (17.4%), and fatigue (13%). Four patients (17.4%) required an interruption of PARPi therapy. Ten patients received a dose reduction within the first 6 months and two patients after one and 1.3 years of PARPi therapy. No dose adjustments were necessary between 1.5 and 4 years. After 4 years, 3 patients (13%) received a late dose reduction. 8.6% had another cancer diagnosed before, 4.3% simultaneously, and 13% after the ovarian cancer diagnosis.

Conclusion

ExR-LTs present with heterogenic clinical and genetic characteristics. Clinical management is complex because of a high rate of AE and need of dose reductions at various points in time. Close monitoring for AE, recurrences and secondary malignancies must be carried out throughout the entire time of treatment.