The impact of operative hysteroscopy on pregnancy outcomes in patients with suspected uterine cavity lesions prior to the first frozen-thawed embryo transfer: a retrospective propensity-score matching cohort study.

From July 1, 2021, to May 31, 2024, a total of 1,731 embryo transfer cycles were conducted at this reproductive center, of which 879 cycles satisfied the inclusion criteria. Among these patients, 526 cases with suspected uterine cavity lesions underwent operative hysteroscopy prior to their first FET, whereas 353 cases without significant intrauterine abnormalities did not undergo such surgery. The patient selection process is illustrated in Fig.  1 . Fig. 1 Flowchart of the patients selection process Flowchart of the patients selection process We observed significant differences in baseline clinical characteristics between the two groups, prompting the use of PSM to balance these discrepancies. Following matching, 286 cycles were included in each group for analysis. Upon comparison, no significant differences were found in age, BMI, history of pregnancy, number of miscarriages, fertilization method, attributes of the transferred embryos, and the number of embryos transferred between the two groups ( P  > 0.05). However, notable differences were observed in delivery method, endometrial thickness on the day of luteal transformation, and frozen embryo transfer protocol ( P  < 0.05). The hysteroscopy group exhibited a higher proportion of cesarean section history compared to the control group (68.75% vs. 42.11%, P  < 0.05). Furthermore, the endometrial thickness on the day of luteal transformation was significantly thinner in the hysteroscopy group compared to the control group (10.02 ± 2.04 mm vs. 10.44 ± 1.73 mm, P  < 0.05). In the FET protocol, HRT, GnRH-a–HRT, and NC accounted for 50.70%, 44.41%, and 4.89% in the hysteroscopy group, respectively, compared to 47.55%, 36.36%, and 16.08% in the control group. The baseline characteristics of patients before and after PSM in each group are summarized in Table  1 . Table 1 Baseline characteristics between two groups before and after PSM Variables Before PSM After PSM Hysteroscopy Control P value Hysteroscopy Control P value No. of cycles 526 353 286 286 Age (year) 33.00 (30.00,35.00) 32.00 (29.00,34.00)  < 0.001 34.00 (32.00,36.00) 36.00 (33.75,38.00) 0.059 BMI (Kg/m 2 ) 21.60 (19.90,23.60) 21.80 (19.80,24.20) 0.393 21.89 ± 1.95 22.79 ± 2.29 0.061 History of pregnancy (%) 44.30 (233/526) 30.02 (106/353)  < 0.001 39.16 (112/286) 33.22 (95/286) 0.139 Delivery method (%)  Vaginal delivery (VD) 39.91 (93/233) 59.43 (63/106) 0.002 a 31.25 (35/112) 57.89 (55/95)  < 0.001  Cesarean section (CS) 58.37 (136/233) 40.57 (43/106) 66.96 (75/112) 42.11 (40/95)  VD and CS 1.72 (4/233) 0 (0/106) 1.79 (2/112) 0.00 (0/95) No. of miscarriages 1.00 (0.00,2.00) 0.00 (0.00,1.00)  < 0.001 1.00 (0.00,2.00) 0.00 (0.00,2.00) 0.215 Endometrial thickness (mm) 9.90 ± 1.99 10.47 ± 1.74  < 0.001 10.02 ± 2.04 10.44 ± 1.73 0.008 FET protocol (%)  HRT 50.95 (268/526) 45.33 (160/353)  < 0.001 50.70 (145/286) 47.55 (136/286)  < 0.001  GnRH-a–HRT 45.44 (239/526) 35.98 (127/353) 44.41 (127/286) 36.36 (104/286)  NC 36.12 (19/526) 18.69 (66/353) 4.89 (14/286) 16.08 (46/286) Fertilization (%)  IVF 32.51 (171/526) 46.17 (163/353)  < 0.001 38.81 (111/286) 34.62 (99/286) 0.298  ICSI 62.49 (355/526) 53.82 (193/353) 61.19 (175/286) 65.38 (187/286) Embryo attributes (%)  Cleavage embryo (D3) 16.92 (89/526) 41.08 (145/353)  < 0.001 29.72 (85/286) 30.07 (86/286) 0.661  Blastocyst embryo (D5) 72.62 (382/526) 56.65 (200/353) 66.08 (189/286) 67.13 (192/286)  D3 + D5 10.46 (55/526) 2.26 (8/353) 4.2 (12/286) 2.80 (8/282) No. of embryos transferred 1.00 (1.00,1.00) 1.00 (1.00,1.00) 0.808 1.00 (1.00,1.00) 1.00 (1.00,1.00) 0.29 ‘a’ indicates that at least one cell has a theoretical frequency T < 5，for which Fisher’s exact probability method is use Baseline characteristics between two groups before and after PSM ‘a’ indicates that at least one cell has a theoretical frequency T < 5，for which Fisher’s exact probability method is use During the follow-up period of this study, the hysteroscopy group exhibited higher rates of biochemical pregnancy, clinical pregnancy, and live birth compared to the control group before matching baseline clinical data. Specifically, biochemical pregnancy rate for the two groups were 73.38% and 59.49% (RR 1.23, 95% CI 1.12–1.36), CPR were 68.06% and 53.54% (RR 1.27, 95% CI 1.14–1.42), and LBR were 53.80% and 43.34% (RR 1.24, 95% CI 1.07–1.42), respectively, with all differences being statistically significant ( P   0.05). After using the propensity score to match the clinical baseline data, the hysteroscopy group still demonstrated significantly higher rates of biochemical pregnancy (70.63% vs. 60.49%, RR 1.17,95%CI 1.04–1.32), clinical pregnancy (65.03% vs. 54.90%, RR 1.18, 95% CI 1.03–1.36), and live birth (52.79% vs. 44.40%, RR 1.19, 95% CI 1.00–1.41) compared to the control group, with all differences being statistically significant ( P   0.05). A summary of the reproductive outcomes is presented in Table  2 . Table 2 Pregnancy outcomes between two groups before and after PSM Variables Before PSM After PSM Hysteroscopy Control P value RR (95% CI) Hysteroscopy Control P value RR (95% CI) No. of cycles 526 353 286 286 Biochemical pregnancy rate(%) 73.38 (386/526) 59.49 (210/353)  < 0.001 1.23 (1.12–1.36) 70.63 (202/286) 60.49 (173/286) 0.011 1.17 (1.04–1.32) CPR (%) 68.06 (358/526) 53.54 (189/353)  < 0.001 1.27 (1.14–1.42) 65.03 (186/286) 54.90 (157/286) 0.012 1.18 (1.03–1.36) LBR (%) 53.80 (283/526) 43.34 (153/353) 0.003 1.24 (1.07–1.42) 52.79 (151/286) 44.40 (127/286) 0.045 1.19 (1.00–1.41) Miscarriage rate (%) 13.69 (49/358) 11.64 (22/189) 0.498 1.17 (0.73–1.88) 12.37 (23/186) 12.10 (19/157) 0.941 1.02 (0.58–1.80) Ectopic pregnancy rate (%) 0.84 (3/358) 1.59 (3/189) 0.424 a 0.53 (0.11–2.59) 1.07 (2/186) 0.64 (1/157) 0.664 a 1.68 (0.15–18.44) Multiple pregnancies rate (%) 5.59 (20/358) 8.46 (16/189) 0.144 0.63 (0.33–1.18) 5.38 (10/186) 6.37 (10/157) 0.696 0.84 (0.36–1.98) Preterm birth rate (%) 11.30 (32/283) 12.42 (19/153) 0.498 1.18 (0.73–1.88) 12.58 (19/151) 11.02 (14/127) 0.689 1.14 (0.60–2.18) ‘a’ indicates that at least one cell has a theoretical frequency T < 5，for which Fisher’s exact probability method is used Pregnancy outcomes between two groups before and after PSM ‘a’ indicates that at least one cell has a theoretical frequency T < 5，for which Fisher’s exact probability method is used We conducted a binary logistic regression analysis on clinical data following PSM. Models 1, 2, and 3 consistently indicated that operative hysteroscopy prior to FET significantly enhanced both CPR and LBR. Specifically, in Model 1, CPR was 52.8%, and LBR was 40% (OR 1.528, 95% CI 1.091–2.14; OR 1.4, 95% CI 1.008–1.946, P  < 0.05). In Model 2, these rates increased to 55.7% for clinical pregnancies and 45.7% for live births (OR 1.557, 95% CI 1.09–2.225; OR 1.457, 95% CI 1.029–2.064, P  < 0.05). Finally, in Model 3, the rates were 54.3% and 44.2%, respectively (OR 1.543, 95% CI 1.070–2.227; OR 1.442, 95% CI 1.011–2.056, P < 0.05), refer to Table  3 . Table 3 Logistic regression analysis of pregnancy outcomes Variables Clinical pregnancy rate Live birth rate OR 95% CI P value OR 95% CI P value Model 1 1.528 1.091–2.14 0.014 1.4 1.008–1.946 0.045 Model 2 1.557 1.090–2.225 0.015 1.457 1.029–2.064 0.034 Model 3 1.543 1.070–2.227 0.02 1.442 1.011–2.056 0.044 Model 1: Crude model Model 2: Adjusted for delivery method, FET protocol, endometrial thickness on the day of luteal transformation Model 3: Adjusted for age, body mass index, delivery method, number of miscarriages, FET protocol, fertilization, embryo attributes, number of embryos transferred and endometrial thickness on the day of luteal transformation Logistic regression analysis of pregnancy outcomes Model 1: Crude model Model 2: Adjusted for delivery method, FET protocol, endometrial thickness on the day of luteal transformation Model 3: Adjusted for age, body mass index, delivery method, number of miscarriages, FET protocol, fertilization, embryo attributes, number of embryos transferred and endometrial thickness on the day of luteal transformation In addition, the full-factor regression model revealed that, apart from operative hysteroscopy, patients with a history of vaginal delivery exhibited a significantly higher clinical pregnancy rate compared to those without such a history. Furthermore, compared to the HRT protocol, the GnRH-a–HRT transfer cycle protocol, the selection of blastocyst transfer over cleavage-stage embryos, and the number of embryos transferred all positively influenced the clinical pregnancy rate (Fig.  2 ). Meanwhile, factors associated with LBR included operative hysteroscopy, a history of vaginal delivery, and the number of embryos transferred (Fig.  3 ). Fig. 2 Multivariate analysis of clinical pregnancy rate after PSM Fig. 3 Multivariate analysis of live birth rate after PSM Multivariate analysis of clinical pregnancy rate after PSM Multivariate analysis of live birth rate after PSM In this study, 89.16% (469/526) of patients and 91.25% (261/286) of patients had uterine cavity lesions before and after PSM, respectively, and only about 10% of patients had normal uterine cavity morphology and normal endometrial structure. Uterine cavity lesions encompassed a variety of conditions, including intrauterine adhesions (IUA), endometrial polyps (EP), chronic endometritis (CE), cesarean section scar diverticulum (CSD), uterine septum or arcuate uterus, and submucosal uterine fibroids. The incidence of mild intrauterine adhesions was notably high both before and after PSM, recorded at 20.91% (110/526) and 24.13% (69/286), respectively. In addition, 23.95% (126/526) and 24.48% (70/286) of patients presented with two or more pathologies. For further details, refer to Figs. 4 and 5 . Fig. 4 Upset diagram of the results of uterine lesions found by hysteroscopy before PSM Fig. 5 Upset diagram of the results of uterine lesions found by hysteroscopy after PSM Upset diagram of the results of uterine lesions found by hysteroscopy before PSM Upset diagram of the results of uterine lesions found by hysteroscopy after PSM After conducting PSM, the study subjects were divided into three subgroups according to the severity of uterine cavity lesions: mild uterine cavity lesions group, moderate uterine cavity lesions group, and severe uterine cavity lesions group. The mild lesions group comprised EP, mild CE, mild IUA, and submucosal fibroids, with patients undergoing only one hysteroscopic surgery (165 cycles). The moderate lesions group included moderate IUA, uterine septum, and severe CE, necessitating two hysteroscopic surgeries (75 cycles). The severe lesions group consisted of severe IUA and CSD (46 cycles). In comparison with the control group, the mild lesions group demonstrated significantly higher rates of biochemical pregnancy, clinical pregnancy, and live birth, with statistically significant differences observed (70.30% vs. 60.49%, OR 1.16, 95% CI 1.01–1.33; 67.27% vs. 54.90%, OR 1.23, 95% CI 1.06–1.42; 58.18% vs. 44.41%, OR 1.31, 95% CI 1.09–1.57, P  < 0.05, respectively). Although the biochemical pregnancy rate, CPR, and LBR in the moderate lesions group were higher than those in the control group, the differences were not statistically significant (70.67% vs. 60.49%, OR 1.17, 95% CI 0.98–1.39; 62.67% vs. 54.90%, OR 1.14, 95% CI 0.93–1.40; 50.67% vs. 44.41%, OR 1.14, 95% CI 0.88–1.48, respectively, P  > 0.05).Despite the severe lesions group having higher rates of biochemical pregnancy (71.74% vs. 60.49%, OR 1.19, 95% CI 0.97–1.45) and clinical pregnancy (58.70% vs. 54.90%, OR 1.07, 95% CI 0.82–1.39) compared to the control group, live birth rate (36.96% vs. 44.41%, OR 0.83, 95% CI 0.56–1.24) was lower than that of the control group, and none of these differences were statistically significant ( P  > 0.05). The miscarriage rates among the three subgroups did not exhibit significant differences when compared to the control group, with rates of 13.51% vs. 12.10% (OR 1.12, 95% CI 0.59–2.10), 6.38% vs. 12.10% (OR 0.53, 95% CI 0.16–1.71), and 18.52% vs. 12.10% (OR 1.53, 95% CI 0.62–3.75), respectively. In the mild lesions group, no ectopic pregnancies were reported; however, there were 6 cases of multiple pregnancies and 13 cases of preterm births; the moderate lesions group had 1, 3, and 6 cases, respectively; and the severe lesions group had 1, 1, and 0 cases, respectively. Some of the information is presented in Table  4 Table 4 Subgroup analysis of pregnancy outcomes after PSM Variables No. of cycles Clinical pregnancy rate Live birth rate Rate (%) P value RR (95% CI) Rate (%) P value RR (95% CI) Control 286 54.9(157/286) Reference 44.41 (127/286) Reference Mild lesions 165 67.27 (111/165) 0.01 1.23 (1.06–1.42) 58.18 (96/165) 0.005 1.31 (1.09–1.57) Moderate lesions 75 62.67 (47/75) 0.227 1.14 (0.93–1.40) 50.67 (38/75) 0.333 1.14 (0.88–1.48) Severe lesions 46 58.7 (27/46) 0.63 1.07 (0.82–1.39) 36.96 (17/46) 0.344 0.83 (0.56–1.24) Variables No. of cycles Biochemical pregnancy rate Miscarriage rate Rate (%) P value RR (95% CI) Rate (%) P value RR (95% CI) Control 286 60.49 (173/286) Reference 12.1 (19/157) Reference Mild lesions 165 70.3 (116/165) 0.036 1.16 (1.01–1.33) 13.51 (15/111) 0.73 1.12 (0.59–2.10) Moderate lesions 75 70.67 (53/75) 0.105 1.17 (0.98–1.39) 6.38 (3/47) 0.267 0.53 (0.16–1.71) Severe lesions 46 71.74 (33/46) 0.144 1.19 (0.97–1.45) 18.52 (5/27) 0.359 a 1.53 (0.62–3.75) Subgroup analysis of pregnancy outcomes after PSM

This study was a retrospective PSM cohort study involving patients who underwent in vitro fertilization–embryo transfer at the Reproductive Medicine Center of the Second Affiliated Hospital of Naval Medical University between January 1, 2021, and May 31, 2024. The follow-up period extended until December 31, 2024. The inclusion criteria were patients undergoing their first FET within the study period. The exclusion criteria included: (i) age ≥ 40 years; (ii) body mass index (BMI) ≥ 28; (iii) patients undergoing fresh cycle embryo transfer; (iv) patients undergoing their second or subsequent frozen embryo transfer cycle at the center; (v) patients with severe intrauterine adhesions resulting in uterine cavity occlusion and presenting with amenorrhea; (vi) patients with uterine anomalies, such as unicornuate uterus, bicornuate uterus, or didelphic uterus; (vii) patients with severe adenomyosis, uterine fibroids affecting the morphology of the uterine cavity, or severe endometriosis impacting the pelvic environment; (viii) patients with untreated hydrosalpinx; and (ix) patients with diagnosed immune system disorders or thrombophilia. Patients were categorized into two groups based on whether they underwent operative hysteroscopy prior to their first FET: the hysteroscopy group and the control group. The hysteroscopy group comprised patients with suspected uterine cavity lesions who received hysteroscopic surgery, whereas the control group consisted of patients without suspected uterine cavity lesions who did not undergo the procedure. This study received approval from the Medical Ethics Committee of the Second Affiliated Hospital of Naval Medical University (approval number: [2024SL138]), and clinical data were collected with informed consent obtained from all patients. The indications for hysteroscopic surgery prior to FET in infertile patients were as follows: (i) vaginal ultrasound reveals endometrial hyperechogenicity or uneven endometrial echogenicity; (ii) a history of failed transplants at other institutions; (iii) the presence of uterine fluid accumulation during fresh cycles; (iv) 3D-TVS evaluation indicates an irregular uterine cavity shape or disruption of the endometrial line, suggesting intrauterine adhesions; (v) 3D-TVS evaluation suggests an internal depression at the midline of the fundus, indicative of a uterine septum; and (vi) patients who meet the criteria for diverticulum management of cesarean section incisions at my reproductive center [ 20 ]. The exclusion criteria included: (i) patients exhibiting acute inflammation of the reproductive tract and (ii) severe hepatic or renal dysfunction, or systemic diseases that render them unsuitable for surgical intervention. Hysteroscopic procedures were conducted by two experienced reproductive surgeons of our reproductive center. The surgeries were conducted within 1 week after the cessation of the patient’s menstrual period. A 22Fr hysteroscope (Karl Storz, Germany) was utilized, with normal saline serving as the distension medium, and the distension pressure was maintained between 120 and 130 mmHg. Following the administration of intravenous anesthesia, the surgical area was disinfected and draped. The cervix was gradually dilated to 6.5 mm using cervical dilator, and the hysteroscope was inserted to thoroughly examine the uterine cavity and cervix to identify any intrauterine lesions. If no intrauterine lesions were detected, the endometrium was gently scraped using 5Fr dissecting forceps, concluding the procedure. Conversely, if intrauterine lesions were identified, appropriate surgical interventions were performed based on the type of lesion. For endometrial polyps, either 5Fr scissors or dissecting forceps were employed to excise the polyp from its base. In cases involving mini polyps, the endometrium was thoroughly cleaned with 5Fr dissecting forceps. In instances of intrauterine adhesions or uterine septum, 5Fr scissors were utilized to separate the adhesions or resect the septum. If a uterine scar diverticulum requiring treatment was identified, the procedure transitioned to a resectoscope. The cervix was dilated to 8.5 mm, and the resectoscope was inserted. Under ultrasound guidance, a loop electrode was employed to excise the lower and lateral flap tissues of the diverticulum, leveling the base of the diverticulum. Subsequently, a rollerball electrode was utilized to coagulate the endometrial tissue and hyperplastic vessels within the diverticulum [ 20 ]. After confirming the absence of significant bleeding, the hysteroscope was withdrawn. Patients with intrauterine adhesions or uterine septum had a uterine balloon placed postoperatively. For patients diagnosed with severe chronic endometritis or uterine septum, a second hysteroscopic surgery was performed within 1 week following the subsequent menstrual period. Conversely, patients exhibiting moderate to severe intrauterine adhesions underwent a second hysteroscopic procedure after a duration of two menstrual cycles. Patients with chronic endometritis were administered oral doxycycline for 2 weeks postoperatively. For patients with intrauterine adhesions or uterine septum, hormonal therapy was initiated on the second day after surgery. The specific medication regimen was individualized according to the patient’s menstrual cycle: initially, Femoston (Abbott Biologicals B.V. Netherlands) red tablets (2 mg) were prescribed for 3–7 days, followed by Femoston yellow tablets (2 mg) for 14 days. There are three endometrial preparation protocols for the FET cycle: natural cycle (NC), hormone replacement therapy (HRT), and HRT with GnRH-a pretreatment cycles (GnRH-a–HRT). Natural cycles were employed for women with regular menstrual cycles. In the HRT cycle, Femoston red tablets were taken orally from the second or third days of menstruation, with the dosage adjusted to 4–6 mg/day based on the thickness of the endometrium. After 10–14 days, a transvaginal ultrasound was conducted, and serum progesterone levels were measured. If the endometrial thickness reached or exceeded 7 mm, endometrial transformation was initiated with a daily dose of 90 mg of progesterone vaginal slow-release gel (Crinone, Merck Serono, Switzerland), along with a vaginal suppository and 6 mg/day of Femoston yellow tablets. For the GnRH-a–HRT cycle, a 3.75 mg intramuscular injection of gonadotrophin-releasing hormone agonist (Diphereline, Ipsen Biopharmaceuticals, France) was administered on the second or third day of the menstrual cycle, followed by hormone replacement as previously described after 28 days. Embryo thawing and transfer methods were implemented as our previous study [ 20 ]. Primary outcomes: clinical pregnancy rate (CPR) and live birth rate(LBR). Clinical pregnancy is defined as at least one intrauterine gestational sac detected by ultrasound 4–6 weeks after embryo transfer. CPR is calculated as the ratio of the number of clinical pregnancy cycles divided by the number of transferred cycles. Live birth is defined as the delivery of a live fetus after 28 complete weeks of gestation. LBR is calculated as the ratio of the number of live-born cycles till December 31, 2024 divided by the number of transferred cycles. Secondary outcomes: biochemical pregnancy rate, miscarriage rate, ectopic pregnancy rate, preterm birth rate and multiple pregnancy rate. Biochemical pregnancy is defined as serum β-hCG detection greater than 5U/L 14 days after embryo transfer. Biochemical pregnancy rate is calculated as the ratio of the number of biochemical pregnancy cycles divided by the number of transferred cycles. Miscarriage rate is calculated as the number of miscarriages before 28 weeks of gestation divided by the number of clinical pregnancy cycles. The ectopic pregnancy rate is defined as the number of cycles with gestational sac implantation outside the uterine cavity divided by the number of clinical pregnancy cycles. Preterm birth rate is calculated as the number of deliveries between 28 weeks and less than 37 weeks of gestation divided by the number of live birth cycles. Multiple pregnancy rate is defined as the ratio of the number of cycles in which two or more fetuses are conceived simultaneously in a single pregnancy to the number of all clinical pregnancy cycles. Data were processed using SPSS 27.0 statistical software. Quantitative data that conformed to normal distribution were represented by mean ± standard deviation (x ± s), independent sample t test was used for intergroup comparison, median (25th percentile, 75th percentile) [M (Q1, Q3)] was used for non-normal distribution, and the Mann–Whitney U rank-sum test was used for intergroup comparison. Categorical data were presented as percentages (numbers), with comparisons between groups using the χ2 test. When the theoretical frequency T was less than 5, Fisher’s exact test was used. PSM was used to adjust for confounders between the hysteroscopy and control groups, minimized the baseline disparities between the two groups [ 21 ]. Variables with significant differences in baseline characteristics were included in the propensity score: age, history of childbirth, number of miscarriages, fertilization method, embryo attributes; Considering the delivery method as a possible cause for patients to undergo hysteroscopy, and the endometrial thickness on the day of luteal transformation and the FET protocol being related to the presence of uterine cavity lesions, these three baseline data were not included in the propensity score. We matched the two groups in a 1:1 ratio with a caliper width of 0.02. Three binary logistic regression models were used to assess the effects of hysteroscopic surgery on CPR and LBR, and the OR and 95% CI were calculated. Model 1 was unadjusted; Model 2 adjusted for delivery method, FET protocol, endometrial thickness on the day of luteal transformation; Model 3 adjusted for all confounding factors, including age, body mass index (BMI), delivery method, number of miscarriages, FET protocol, fertilization, embryo attributes, number of embryos transferred and endometrial thickness on the day of luteal transformation. Statistical differences were considered significant at  p  < 0.05.

This study represents the first retrospective PSM analysis investigating the impact of operative hysteroscopy on reproductive outcomes in patients with suspected uterine lesions prior to their first FET. Our research findings indicated that both before and after PSM, the operative hysteroscopy group exhibited higher rates of biochemical pregnancy, clinical pregnancy, and live birth compared to the control group. Furthermore, binary logistic regression analysis across multiple models demonstrated that operative hysteroscopy was significantly associated with improved CPR and LBR. These results suggested that the use of operative hysteroscopy for the treatment of suspected uterine cavity lesions can significantly enhance reproductive outcomes. Hysteroscopic procedures are classified into two categories: diagnostic hysteroscopy and operative hysteroscopy. Diagnostic hysteroscopy is primarily employed for diagnostic purposes and is typically conducted on an outpatient basis, often requiring no anesthesia or only local anesthesia. In contrast, operative hysteroscopy involves the removal of pathological lesions following a diagnostic assessment [ 22 ]. The role of hysteroscopic procedures within the context of assisted reproduction remains a topic of ongoing debate. Key areas of controversy include: the potential benefits of performing diagnostic hysteroscopy in patients without suspected intrauterine pathology on pregnancy outcomes; the effect of operative hysteroscopy on pregnancy outcomes in patients with intrauterine pathology; and the optimal timing for conducting hysteroscopy in patients who have experienced implantation failure. In the 2023 ESHRE good practice recommendations on reproductive medicine, hysteroscopy is not recommended for routine clinical use. However, hysteroscopic screening can be considered for patients with recurrent implantation failure [ 23 ]. The ESHRE recommendations reference a 2019 Cochrane review, which indicates that screening hysteroscopy before IVF may increase CPR (RR 1.32, 95% CI 1.20–1.45) and LBR (RR 1.26, 95% CI 1.11–1.43). However, sensitivity analyses based on trials with a low risk of bias revealed no increase in live birth rate following screening hysteroscopy (RR 0.99, 95% CI 0.82–1.18) [ 24 ]. A subsequent systematic review and meta-analysis corroborated these findings, noting that moderate-quality evidence suggests hysteroscopy before the first ART cycle improves CPR (RR 1.32, 95% CI 1.11–1.57, I 2  = 42%) but does not significantly affect LBR (RR 1.44, 95% CI 0.83–2.48) [ 13 ]. Conversely, a more recent systematic review and meta-analysis reported that outpatient hysteroscopy before ART is associated with increased LBR (RR 1.22, 95% CI 1.03–1.45) and CPR (RR 1.27, 95% CI 1.10–1.47) [ 25 ]. Although these studies employ varying terms to describe hysteroscopic procedures, such as “screening hysteroscopy” and “outpatient hysteroscopy,” the study populations consistently comprise patients without suspected intrauterine pathologies. The research aims are also consistent: to investigate the impact of hysteroscopic examination on pregnancy outcomes in patients without suspicious uterine cavity lesions before undergoing IVF. In contrast to the aforementioned study, the present research specifically targets patients with suspected uterine cavity lesions, investigating differences in pregnancy outcomes between two cohorts: those with suspected uterine cavity lesions who underwent operative hysteroscopy prior to their first FET and those without suspected uterine cavity lesions who did not undergo hysteroscopic intervention. To date, this is the first retrospective PSM analysis aimed at evaluating the impact of operative hysteroscopy on reproductive outcomes in patients with suspected uterine lesions before their first FET. Moreover, this study is the first to analyze the effects of operative hysteroscopy on pregnancy outcomes following frozen embryo transfer after the treatment of various uterine lesions, rather than focusing solely on single uterine lesions. This study revealed that the biochemical pregnancy rate, clinical pregnancy rate, and live birth rate in the hysteroscopy group were 73.38%, 68.06%, and 53.5%, respectively, all of which exceeded those in the control group (59.49%, 53.54%, and 43.34%). The differences were statistically significant ( P  < 0.05). Given the substantial disparities in baseline clinical characteristics between the two groups, PSM was employed to adjust for several baseline variables. However, due to potential differences in delivery mode, endometrial thickness on the day of luteal conversion, and embryo transfer protocols associated with the presence or absence of uterine cavity lesions, as well as sample size requirements, these three baseline clinical variables were not matched. Following PSM, the hysteroscopy group continued to exhibit a favorable effect on pregnancy outcomes. The biochemical pregnancy rate, clinical pregnancy rate, and live birth rate remained higher in the hysteroscopy group compared to the control group(70.63% vs. 60.49%,RR 1.17, 95% CI 1.04–1.32;65.03% vs. 54.90%, RR 1.18, 95% CI 1.03–1.36; 52.79% vs. 44.40%, RR 1.19, 95% CI 1.00–1.41), while no significant differences were observed between the two groups regarding miscarriage rate, ectopic pregnancy rate, multiple pregnancy rate, or preterm birth rate(12.37% vs. 12.10%, RR 1.02, 95% CI 0.58–1.80; 1.07% vs. 0.64%, RR 1.68, 95% CI 0.15–18.44; 5.38% vs. 6.37%, RR 0.84, 95% CI 0.36–1.98; 12.58% vs. 11.02%, RR1.14, 95% CI 0.60–2.18). Logistic regression analyses, both crude and multivariate, revealed that hysteroscopic surgery was associated with improved clinical pregnancy rate and live birth rate. Specifically, the multivariate model indicated that hysteroscopic surgery increased the clinical pregnancy rate by 54.3% and the live birth rate by 44.2%. Uterine cavity abnormalities and endometrial pathologies are among the leading causes of reduced fertility and implantation failure. These conditions can interfere with embryo implantation, consequently decreasing clinical pregnancy rates. To enhance LBR, improve CPR, and reduce the risk of early pregnancy loss, it is essential to maintain a morphologically normal uterine cavity and an endometrium of appropriate thickness [ 26 – 29 ]. In this study, only about 10% of the patients were free from uterine cavity lesions, while the remainder exhibited various types of uterine lesions, including IUA, EP, CE, uterine septum, CSD, and submucosal fibroids. Notably, over 20% of the patients presented with two or more lesions, which could adversely affect the outcomes of ART. However, operative hysteroscopy has been demonstrated to significantly improve pregnancy outcomes in these cases [ 18 , 20 , 30 – 35 ]. In light of the high cesarean section rate in China, the adverse effects of CSD on ART are well-documented [ 36 , 37 ]. Our team’s 2024 study indicated no statistical difference in clinical pregnancy and live birth rates between symptomatic CSD patients treated with hysteroscopy and asymptomatic CSD patients [ 20 ]. This study conducted both univariate and multivariate regression analyses on patients with suspected uterine lesions following surgery, revealing that hysteroscopic surgery significantly improved pregnancy outcomes. For further analysis, the hysteroscopy group was categorized into three subgroups based on the severity of uterine lesions: mild, moderate, and severe. Compared to the control group, the mild lesions subgroup demonstrated significant improvements in biochemical pregnancy, clinical pregnancy, and live birth rates; the moderate lesions subgroup exhibited a similar trend, albeit without statistical significance; and the severe lesions subgroup had higher biochemical and clinical pregnancy rates but lower live birth rates than the control group, also without statistical significance. The reasons for these findings can be summarized as follows: first, hysteroscopy, recognized as the gold standard for diagnosing uterine pathologies, provides high diagnostic accuracy. In contrast, patients in the control group were diagnosed using TVS or HSG, which are associated with certain misdiagnosis rates and are heavily reliant on the clinician’s expertise. This reliance can lead to undetected uterine abnormalities that may hinder embryo implantation and negatively impact the success rates of ART. Second, operative hysteroscopy not only effectively corrects uterine morphological abnormalities and increases uterine volume but also enhances endometrial receptivity by addressing uterine pathologies. Treating any identified uterine lesions may consequently improve ART outcomes. Furthermore, the standardization of hysteroscopic surgery and the concept of endometrial protection were closely related. In managing uterine diseases, except for cases of CSD, where electrical instruments were utilized, other conditions were treated with scissors and forceps. These cold instruments safeguarded the endometrium, minimized the risk of endometrial damage from electrical procedures, and reduced the incidence of intrauterine adhesions. Meanwhile, during the surgical procedure for treating uterine cavity lesions, the surgeon employed dissecting forceps to meticulously clean the endometrium. This approach, which integrates hysteroscopy with endometrial scratching, has been shown in some clinical studies to potentially improve pregnancy outcomes [ 38 ]. However, the study also indicated that even following hysteroscopic surgery, the extent of uterine pathology influenced pregnancy outcomes differently, as evidenced by the varying results among the three subgroups. In the multivariate regression analysis of CPR and LBR, this study revealed that, in addition to operative hysteroscopy, factors influencing CPR included a history of vaginal delivery, the GnRH-a–HRT transfer protocol, blastocyst transfer, and the number of embryos transferred. Factors affecting LBR included a history of vaginal delivery and the number of embryos transferred. These findings were consistent with previous research conclusions [ 16 , 39 – 43 ]. This indicated that the factors influencing assisted reproductive outcomes were multifaceted, and therefore, individualized treatment strategies should be selected for patients prior to embryo transfer. At the same time, clinicians should rigorously evaluate the uterine cavity to enable early diagnosis and effective treatment of potential uterine factors that may reduce the success rate of ART. This study possesses several strengths. First, despite being a real-world retrospective cohort study, we effectively mitigated the influence of confounding factors by matching baseline clinical data, thereby enhancing the reliability of intergroup comparisons. Second, all data were sourced from a single reproductive medical center, which featured a substantial sample size, and all hysteroscopic procedures were performed by two experienced specialists. This not only ensured the reliability of the case data but also improved diagnostic accuracy, significantly reducing selection bias. Most importantly, this is the first study to employ three logistic regression models to investigate the impact of operative hysteroscopy on the pregnancy outcomes of the first FET in patients with suspected uterine cavity lesions. The positive influence of hysteroscopic surgery on pregnancy outcomes provides evidence for clinical decision-making. This study has several limitations. First, as a retrospective cohort study based on real-world data, it lacks an ideal control group. Although we performed statistical adjustments for known confounding factors through multivariate regression analysis and propensity score matching, unmeasured confounders may still influence the results. Second, the final sample size was limited; therefore, the findings should be interpreted with caution. These results primarily reveal associations between variables rather than establishing definitive causal relationships. Future prospective studies or randomized controlled trials will be necessary to validate our findings.

In summary, this study demonstrated that for patients with suspected uterine cavity lesions performing hysteroscopic correction prior to the first FET significantly enhances CPR and LBR. Multivariable regression analysis demonstrated that hysteroscopic surgery was associated with increased clinical pregnancy and live birth rates. Therefore, it is recommended that hysteroscopic surgery be considered prior to the first FET for patients with suspected uterine cavity lesions to improve their pregnancy outcomes.

Infertility is defined as the inability of a couple to achieve a successful pregnancy after 12 months or more of regular unprotected sexual intercourse. Studies indicate that approximately 1 in 6 couples experience fertility issues [ 1 ]. Globally, the prevalence of infertility is increasing [ 2 ]. An epidemiological study conducted in China revealed a significant rise in the prevalence of infertility among couples of reproductive age, reaching as high as 25% [ 3 ]. This trend has led to a notable increase in the number of patients seeking ART for assistance. Despite continuous advancements in ART, implantation failure remains a significant clinical challenge. Uterine abnormalities are a common cause of implantation failure. Approximately 10% of women with reduced fertility exhibit uterine cavity abnormalities, and nearly half of women experiencing recurrent implantation failure may have uterine abnormalities [ 4 ]. Successful embryo implantation is a process that involves the synchronized development of the embryo and the endometrium, with numerous factors influencing this process. Both organic uterine lesions and inflammatory diseases of the endometrium can significantly reduce endometrial receptivity, leading to implantation failure [ 5 ]. Therefore, it is essential to exclude uterine cavity lesions prior to embryo transfer. Common diagnostic methods for assessing the uterine condition in infertility patients include transvaginal ultrasound scanning (TVS), hysterosalpingography (HSG), three-dimensional transvaginal ultrasound scanning (3D-TVS), and pelvic MRI. Although these examinations are relatively straightforward to perform, they each have certain limitations in accurately identifying intrauterine pathologies. While TVS demonstrates a sensitivity and specificity range of 84.5–98% [ 6 ], its effectiveness is operator-dependent. HSG, on the other hand, exhibits poor diagnostic performance with a sensitivity of only 21.6% and a false-negative rate of 78.4% [ 7 ]. Although 3D-TVS achieves an accuracy of 84.1%, hysteroscopy remains the gold standard as it allows for direct visualization, precise diagnosis, and simultaneous treatment of intrauterine lesions [ 8 ].As a common clinical procedure in assisted reproductive technologies, hysteroscopy provides the advantage of direct visualization of the uterine cavity, allowing for the assessment of lesion characteristics, such as nature, size, shape, location, and vascularity [ 4 ]. It is recognized as the gold standard for evaluating uterine cavity lesions and is currently the only method that enables direct observation of both physiological and pathological changes in the endometrium, along with the possibility of simultaneous targeted treatment [ 9 ]. Hysteroscopy fulfills both diagnostic and therapeutic roles for anatomical structural lesions or endometrial pathologies within the uterine cavity. With the continuous improvements, diversification, and miniaturization of hysteroscopic instruments [ 10 , 11 ], hysteroscopy has evolved into a well-tolerated procedure that can be performed without anesthesia [ 12 ]. Although the value of hysteroscopy in ART has not yet reached a consensus, the controversy primarily revolves around diagnostic hysteroscopy and office hysteroscopy, with most studies focusing on patients without suspected uterine cavity abnormalities [ 13 – 16 ]. Research on the application of operative hysteroscopy in ART, particularly prior to the first FET, is relatively limited and often restricted to single disease types [ 17 – 19 ]. This study employed a retrospective propensity-score matching (PSM) analysis for the first time to investigate the impact of operative hysteroscopy on pregnancy outcomes in patients with suspected uterine cavity lesions before their first FET. In addition, logistic binary regression was utilized for multi-model analysis to evaluate the influence of operative hysteroscopy on pregnancy outcomes.