Nomogram to predict cumulative live birth rate following in vitro fertilization/intracytoplasmic sperm injection cycles in patients with endometriosis

In this study, a total of 1457 individuals were identified as eligible from January 2017 to August 2022. Among the patients included, the majority were females aged < 35, corresponding to 84.69% of the sample (Supplementary Table S2). There were 223 individuals aged ≥ 35, making up 15.31% of the total. Among them, a total of 811 individuals achieved successful CLB (55.66%). The model was constructed using a training group comprising 1019 patients and subsequently tested on a separate validation group consisting of 438 patients. Table 1 summarizes the patient characteristics for both groups. Baseline data revealed no significant differences concerning general demographics, strategies for ART and CLB rate between the training and validation groups. In the training cohort, there were 573 cases of CLB (56.23%), while in the validation cohort, there were 238 cases of CLB (54.34%). In the univariate logistic regression analysis conducted within the training group, various factors, including female age, male age, female BMI, DIE, coexisting adenomyosis, GnRH-a treatment after operation, duration of infertility, level of carbohydrate antigen 125 (CA125), follicle-stimulating hormone (FSH), and AMH, AFC on both sides, COS protocol, duration of gonadotropin administration, Estradiol (E 2 ) levels before HCG trigger, Progesterone (P) levels before HCG trigger, number of retrieved oocytes, number of fertilization, and the total number of high-quality embryos, were considered potential predictors of CLB (Supplementary Table S3). Table 1 Characteristics of the training and the validation cohorts Variable Total ( n  = 1457) Training cohort ( n  = 1019) Validation cohort ( n  = 438) Statistic P Female age, n (%) χ 2  = 0.232 0.630  < 35 1234 (84.69) 860 (84.40) 374 (85.39)  ≥ 35 223 (15.31) 159 (15.60) 64 (14.61) Male age, Mean ± SD 32.05 ± 4.38 32.02 ± 4.36 32.13 ± 4.42 t = 0.459 0.647 Female BMI (kg/m 2 ), n (%) - 0.962  < 18.5 187 (12.83) 132 (12.95) 55 (12.56)  18.5–23.9 1065 (73.1) 744 (73.01) 321 (73.29)  24.0–27.9 191 (13.11) 134 (13.15) 57 (13.01)  ≥ 28.0 14 (0.96) 9 (0.88) 5 (1.14) Male BMI (kg/m 2 ), n (%) χ 2  = 2.541 0.468  < 18.5 47 (3.23) 37 (3.63) 10 (2.28)  18.5–23.9 702 (48.18) 483 (47.40) 219 (50.00)  24.0–27.9 529 (36.31) 370 (36.31) 159 (36.30)  ≥ 28.0 179 (12.29) 129 (12.66) 50 (11.42) Education, n (%) - 0.510  < Secondary school 5 (0.34) 2 (0.20) 3 (0.68)  Secondary school 563 (38.64) 394 (38.67) 169 (38.58)  College graduate 726 (49.83) 511 (50.15) 215 (49.09)  Post-graduate 163 (11.19) 112 (10.99) 51 (11.64) Smoking, n (%) χ 2  = 0.000 1.000  No 1452 (99.66) 1016 (99.71) 436 (99.54)  Yes 5 (0.34) 3 (0.29) 2 (0.46) Gravidity, n (%) χ 2  = 1.016 0.602  0 998 (68.5) 697 (68.40) 301 (68.72)  1 313 (21.48) 215 (21.10) 98 (22.37)  ≥ 2 146 (10.02) 107 (10.50) 39 (8.90) Parity, n (%) - 0.642  0 1309 (89.84) 917 (89.99) 392 (89.50)  1 139 (9.54) 97 (9.52) 42 (9.59)  2 9 (0.62) 5 (0.49) 4 (0.91) Endometriosis recurrence, n (%) - 0.703  No 1051 (72.13) 729 (71.54) 322 (73.52)  Yes 219 (15.03) 154 (15.11) 65 (14.84)  Never surgery 173 (11.87) 127 (12.46) 46 (10.50)  Aspiration 14 (0.96) 9 (0.88) 5 (1.14) Times of endometriosis surgery, n (%) - 0.563  0 187 (12.83) 136 (13.35) 51 (11.64)  1 1265 (86.82) 880 (86.36) 385 (87.90)  2 5 (0.34) 3 (0.29) 2 (0.46) ASRM Stage, n (%) χ 2  = 1.035 0.596  1–2 416 (28.55) 293 (28.75) 123 (28.08)  3–4 854 (58.61) 590 (57.90) 264 (60.27)  unknown 187 (12.83) 136 (13.35) 51 (11.64) DIE, n (%) χ 2  = 0.269 0.604  No 1406 (96.5) 985 (96.66) 421 (96.12)  Yes 51 (3.5) 34 (3.34) 17 (3.88) Adenomyosis, n (%) χ 2  = 0.029 0.865  No 1307 (89.7) 915 (89.79) 392 (89.50)  Yes 150 (10.3) 104 (10.21) 46 (10.50) GnRH-a treatment after operation, n (%) χ 2  = 0.806 0.668  None 1046 (71.79) 738 (72.42) 308 (70.32)  1-3 months 233 (15.99) 161 (15.80) 72 (16.44)  4-6 months 178 (12.22) 120 (11.78) 58 (13.24) Infertility type, n (%) χ 2  = 3.951 0.139  Primary 939 (64.45) 654 (64.18) 285 (65.07)  Secondary 401 (27.52) 274 (26.89) 127 (29.00)  Not reach 1 year 117 (8.03) 91 (8.93) 26 (5.94) Infertility duration (years), Mean ± SD 2.92 ± 2.53 2.90 ± 2.58 2.98 ± 2.40 t = 0.577 0.564 HCY (umol/L), Mean ± SD 9.14 ± 2.72 9.19 ± 2.89 9.03 ± 2.29 t = -1.013 0.311 CA125 (U/ml), M (Q 1 , Q 3 ) 22.00 (14.10—37.30) 21.60 (14.20—36.35) 23.40 (13.83—39.03) Z = 0.770 0.441 FSH (IU/L), Mean ± SD 7.42 ± 3.17 7.45 ± 3.31 7.33 ± 2.80 t = -0.683 0.495 LH (IU/L), Mean ± SD 5.26 ± 2.99 5.35 ± 3.24 5.06 ± 2.29 t = -1.698 0.090 E 2 (pmol/l), Mean ± SD 127.29 ± 77.86 128.65 ± 77.97 124.13 ± 77.60 t = -1.017 0.309 P (nmol/l), Mean ± SD 1.19 ± 0.68 1.21 ± 0.69 1.16 ± 0.63 t = -1.421 0.155 AMH (ng/ml), M (Q 1 , Q 3 ) 2.53 (1.49—3.96) 2.52 (1.48—3.89) 2.56 (1.56—4.09) Z = 0.848 0.397 AFC, Mean ± SD 9.93 ± 4.80 10.01 ± 4.89 9.74 ± 4.61 t = -0.965 0.335 COS protocol, n (%) χ 2  = 3.070 0.800  GnRH-a prolonged protocol 398 (27.32) 281 (27.58) 117 (26.71)  GnRH-a long protocol 388 (26.63) 269 (26.40) 119 (27.17)  GnRH-a short protocol 38 (2.61) 26 (2.55) 12 (2.74)  LPS/PPOS 54 (3.71) 43 (4.22) 11 (2.51)  GnRH antagonist 464 (31.85) 321 (31.50) 143 (32.65)  Mild stimulation 103 (7.07) 70 (6.87) 33 (7.53)  Natural cycle 12 (0.82) 9 (0.88) 3 (0.68) Total dosage of Gn (IU), Mean ± SD 2287.49 ± 849.16 2295.12 ± 849.29 2269.74 ± 849.56 t = -0.523 0.601 Duration of Gn stimulation (days), Mean ± SD 10.35 ± 2.60 10.35 ± 2.60 10.35 ± 2.59 t = 0.028 0.977 E 2 on HCG trigger day (pmol/l), M (Q 1 , Q 3 ) 8561.00 (5410.00—13,827.00) 8563.50 (5393.25—13,766.00) 8558.00 (5449.00—13,933.00) Z = 0.233 0.816 P (nmol/l) on HCG trigger day, M (Q 1 , Q 3 ) 2.42 (1.70—3.48) 2.45 (1.73—3.46) 2.38 (1.59—3.56) Z = 0.910 0.363 Sperm source, n (%) χ 2  = 1.195 0.274  Husband 1432 (98.28) 1004 (98.53) 428 (97.72)  Donor 25 (1.72) 15 (1.47) 10 (2.28) Thickness of endometrium on trigger day (cm), Mean ± SD 1.10 ± 0.27 1.09 ± 0.27 1.11 ± 0.27 t = 1.249 0.212 Infertility with tubal factor, n (%) χ 2  = 0.727 0.394  0 899 (61.7) 636 (62.41) 263 (60.05)  1 558 (38.3) 383 (37.59) 175 (39.95) Infertility with male factor, n (%) χ 2  = 1.004 0.316  0 1132 (77.69) 799 (78.41) 333 (76.03)  1 325 (22.31) 220 (21.59) 105 (23.97) Number of oocytes retrieved, M (Q 1 , Q 3 ) 8.00 (5.00—13.00) 8.00 (5.00—13.00) 8.00 (5.00—12.00) Z = 0.534 0.594 Numer of Fertilization, M (Q 1 , Q 3 ) 5.00 (2.00—8.00) 5.00 (2.00—8.00) 4.50 (2.00—7.00) Z = 0.292 0.771 Number of high-quality embryos, Mean ± SD 2.15 ± 2.02 2.21 ± 2.05 2.01 ± 1.92 t = -1.749 0.081 CLB, n (%) χ 2  = 0.445 0.505  No 646 (44.34) 446 (43.77) 200 (45.66)  Yes 811 (55.66) 573 (56.23) 238 (54.34) Note: CLB cumulative live birth, BMI body mass index, ASRM American Society for Reproductive Medicine, DIE deep infiltrating endometriosis, GnRH-a gonadotrophin releasing hormone analogue, HCY homocysteine, CA125 carbohydrate antigen 125, FSH follicle-stimulating hormone, LH luteinizing hormone, E 2 estradiol, P progesterone, AMH anti-mullerian hormone, AFC antral follicle count, COS controlled ovulation stimulation, Gn gonadotropins, LPS luteal phase ovarian stimulation, PPOS progestin-primed-ovarian-stimulation, HCG human chorionic gonadotropin Characteristics of the training and the validation cohorts Note: CLB cumulative live birth, BMI body mass index, ASRM American Society for Reproductive Medicine, DIE deep infiltrating endometriosis, GnRH-a gonadotrophin releasing hormone analogue, HCY homocysteine, CA125 carbohydrate antigen 125, FSH follicle-stimulating hormone, LH luteinizing hormone, E 2 estradiol, P progesterone, AMH anti-mullerian hormone, AFC antral follicle count, COS controlled ovulation stimulation, Gn gonadotropins, LPS luteal phase ovarian stimulation, PPOS progestin-primed-ovarian-stimulation, HCG human chorionic gonadotropin A multiple logistic regression analysis was performed, incorporating the statistically significant variables from the univariate analysis in a stepwise fashion. Ultimately, adenomyosis, duration of infertility, number of fertilization, and number of high-quality embryos were included in the predictive model. Overall, a higher number of fertilized oocytes (OR = 1.19, 95% CI: 1.13–1.25) and a high number of high-quality embryos (OR = 1.65, 95% CI: 1.47–1.86) were associated with increased odds of achieving CLB, as detailed in Table  2 . Conversely, individuals with adenomyosis were less likely to achieve CLB (OR = 0.51, 95% CI: 0.31–0.82). Longer duration of infertility was associated with reduced odds of CLB rate (OR = 0.93, 95% CI: 0.88–0.99). We developed an individualized nomogram prediction model for CLB rate after IVF/ICSI in patients with endometriosis, as illustrated in Fig.  2 . When evaluating an individual's risk of CLB rate, the points assigned to each variable are summed to obtain a total point, which corresponds to a specific risk level. Table 2 Establishment of the predictive nomogram Variables Beta S. E Z P OR (95%CI) Intercept -1.27 0.163 -7.740 <.001 Female age -0.36 0.22 -1.68 0.094 0.69 (0.45 - 1.06) Adenomyosis  No 1.00 (Reference)  Yes -0.68 0.25 -2.75 0.006 0.51 (0.31 - 0.82) Duration of infertility -0.07 0.03 -2.33 0.020 0.93 (0.88 - 0.99) Number of fertilization 0.17 0.03 6.55 <.001 1.19 (1.13 - 1.25) Number of high-quality embryos 0.50 0.06 8.50 <.001 1.65 (1.47 - 1.86) OR Odds Ratio, CI Confidence Interval, S.E Standard Error Fig. 2 Nomogram to predict cumulative live birth rate after IVF/ICSI in patients with endometriosis Establishment of the predictive nomogram OR Odds Ratio, CI Confidence Interval, S.E Standard Error Nomogram to predict cumulative live birth rate after IVF/ICSI in patients with endometriosis The area under the curve (AUC) of the training model was found to be 0.823, with a 95% CI of 0.798–0.849. The receiver operating characteristic (ROC) curve for the training model can be seen in Fig.  3 A. The optimal cutoff point was determined based on the Youden index, resulting in a sensitivity of 0.702 (95% CI: 0.659–0.744), specificity of 0.812 (95% CI: 0.779–0.844), negative predictive value (NPV) of 0.778 (95% CI: 0.744–0.811), and positive predictive value (PPV) of 0.743 (95% CI: 0.702–0.785). The AUC of the ROC curve in the validation set was determined to be 0.773 (95% CI: 0.729–0.818), suggesting a fair level of performance (Fig.  3 B). Fig. 3 ROC curves in training group ( A ) and validation group ( B ) ROC curves in training group ( A ) and validation group ( B ) When applying the model to subgroups within the test population, the AUC for individuals from different age groups, BMI, and numbers of high-quality embryos indicated fair performance (Table  3 ). Table 3 Applications of the model in subgroups of the validation cohort Variables n Cut off AUC (95%CI) Accuracy (95%CI) Sensitivity (95%CI) Specificity (95%CI) NPV (95%CI) PPV (95%CI) Age  <35 374 0.472 0.761 (0.712-0.811) 0.714 (0.665-0.759) 0.599 (0.524-0.673) 0.807 (0.753-0.861) 0.714 (0.656-0.772) 0.714 (0.639-0.789)  ≥35 64 0.472 0.835 (0.731-0.938) 0.750 (0.626-0.850) 0.818 (0.687-0.950) 0.677 (0.513-0.842) 0.778 (0.621-0.935) 0.730 (0.587-0.873) Female BMI  <18.5 55 0.472 0.795 (0.673-0.917) 0.691 (0.552-0.809) 0.565 (0.363-0.768) 0.781 (0.638-0.924) 0.714 (0.565-0.864) 0.650 (0.441-0.859)  18.5-23.9 321 0.472 0.778 (0.726-0.830) 0.726 (0.674-0.774) 0.623 (0.545-0.702) 0.811 (0.753-0.869) 0.721 (0.658-0.783) 0.734 (0.656-0.812)  24.0-27.9 57 0.472 0.706 (0.566-0.845) 0.684 (0.548-0.801) 0.704 (0.531-0.876) 0.667 (0.498-0.835) 0.714 (0.547-0.882) 0.655 (0.482-0.828)  ≥28.0 5 - - - - - - - Number of high-quality embryos  0 98 0.472 0.739 (0.611-0.866) 0.806 (0.714-0.879) 0.939 (0.887-0.991) 0.125 (<0.001-0.287) 0.286 (<0.001-0.620) 0.846 (0.772-0.920)  1 108 0.472 0.725 (0.629-0.820) 0.639 (0.541-0.729) 0.804 (0.690-0.919) 0.516 (0.392-0.641) 0.780 (0.654-0.907) 0.552 (0.433-0.671)  ≥2 232 0.472 0.658 (0.578-0.737) 0.707 (0.644-0.765) 0.194 (0.103-0.286) 0.938 (0.900-0.975) 0.721 (0.660-0.782) 0.583 (0.386-0.781) AUC area under curve, BMI body mass index, NPV negative predictive value, PPV positive predictive value Applications of the model in subgroups of the validation cohort AUC area under curve, BMI body mass index, NPV negative predictive value, PPV positive predictive value

From January 2017 to August 2022, women diagnosed with endometriosis in IVF/ICSI cycles were identified at Women's Hospital, Zhejiang University School of Medicine. The diagnosis was confirmed surgically or based on imaging of ultrasound by two experienced sonographers in all patients [ 13 , 14 ]. The exclusion criteria were: (1) patients with chronic hypertension, diabetes, or heart disease; (2) treatment with donated oocytes; (3) women suffered from malignant tumors; (4) couple members with chromosome abnormality and (5) patients lost to follow-up. The protocol was approved by the Ethical Review Board of Women's Hospital, Zhejiang University School of Medicine. Figure  1 demonstrates the flow chart of the study. Fig. 1 The flow chart of the study The flow chart of the study Available information on patient's characteristics included maternal age, gravidity, parity, the duration of infertility, maternal education, previous cesarean delivery, pre-pregnant body mass index (BMI), homocysteine (Hcy), basal hormone level (measured on days 2 to 4 of the menstrual cycle), anti-Mullerian hormone (AMH), antral follicle count (AFC) and the characteristic of IVF/ICSI. Clinical data on endometriosis was retrieved from medical records as follows: operation history of ovary endometriosis cyst, American Society for Reproductive Medicine (ASRM) Stage, Gonadotrophin releasing hormone analogue (GnRH-a) treatment after operation, existence of deep infiltrating endometriosis (DIE) or adenomyosis, and relapse condition. All data were retrospectively collected on a computerized database or by telephone interview. The missing values were imputed by multiple imputations, and no significant impact was observed with a multiple imputation sensitivity analysis (Supplementary Table S1). The primary outcome was the CLB rate, which was defined as at least one live birth arising from a single ART cycle in the transfer of a fresh embryo or later frozen embryos. Controlled ovarian stimulation (COS) was performed according to the usual methods. Oocytes were collected via transvaginal follicular aspiration at 36 h after administration. Luteal support was provided after egg retrieval. Embryos or blastocysts with the highest scores by two experienced embryologists were selected for embryo transfer (ET). As to frozen-thawed embryos, the patients receive hormonal replacement therapy, ovulation or natural cycle for endometrial preparation. Maternal serum human chorionic gonadotropin (HCG) was measured 2 weeks later to confirm biological pregnancy. Subsequently, clinical pregnancy was detected by ultrasound 5 weeks after ET. During pregnancy, prenatal screening was performed. All statistical analysis was performed with SPSS (Version 22.0, USA) and R software (Version 4.2.0, USA). Continuous variables following a normal distribution were expressed as ‘Mean ± SD,’ and the comparison was performed using an independent t-test. Skewed continuous data were presented as ‘Median (Q1, Q3),’ and the comparison was conducted using an independent samples Mann–Whitney U test. Categorical data were presented as ‘n (%),’ and statistical analysis was carried out using either a chi-square test or Fisher's exact test when appropriate. The relationships between maternal factors and CLB rate were examined using multivariable logistic regression analysis. Differences of P  < 0.05 were considered statistically significant. The cohort was randomly split into the training set ( n  = 1019, 70%) and the validation set ( n  = 438, 30%) in constructing and testing the nomogram. The randomization process was conducted using a computer-generated random number table to ensure an unbiased allocation of patients. Univariate and multifactorial logistic regression analyses were conducted to determine significant predictive factors influencing the CLB rate. A nomogram was conducted to predict the CLB rate using the package “rms” via R software. In order to mitigate selection bias, age was integrated mandatorily into the predictive model, given that the study population primarily consisted of females aged < 35. In the validation set, the predictive model's performance in estimating the CLB rate was evaluated to validate the model's predictive capability.

The management of endometriosis aims to alleviate symptoms, prevent disease progression, and improve fertility if desired. It is well known that women with endometriosis experience diminished fecundity and fertility [ 15 ]. IVF or ICSI may be recommended when infertility is a concern. The use of ART has been proven to offer a higher fertilization rate and chance of pregnancy for patients with endometriosis [ 16 ]. These assisted reproductive techniques offer hope to many couples struggling with infertility due to endometriosis. However, the success of IVF/ICSI in endometriosis patients varies widely, and predicting the likelihood of achieving a live birth remains a crucial clinical challenge. CLB rate is an assessment measure for the effectiveness of ART treatment, both for couples dealing with infertility and for clinical practitioners. More and more researchers tend to utilize this comprehensive and precise indicator. However, there is currently no reported predictive model for CLB rate in infertility cases associated with endometriosis. In our study, we employed CLB rate as a more comprehensive parameter, while considering additional variables that might influence pregnancy outcomes, including the presence of concurrent adenomyosis, the COS protocol, the number of retrieved oocytes, the number of fertilized oocytes, and the total count of high-quality embryos, which was not previously addressed in prior researches. Independent predictive factors are identified, and a prediction model for CLB rate is established based on age, duration of infertility, adenomyosis, number of fertilization, and number of high-quality embryos. The calculator of the CLB rate can be used by patients, clinicians, and researchers, which is easy and straightforward in practice. It is meant to help fertility doctors make informed decisions and provide adequate counseling. Adenomyosis is characterized by the abnormal presence of endometrial epithelial cells and stromal fibroblasts within the myometrium of the uterus [ 17 ]. This condition is associated with symptoms such as dysmenorrhea, abnormal uterine bleeding, and uterine enlargement, which tend to worsen progressively over time. Impaired fertility is a typical clinical manifestation among patients with adenomyosis, with an incidence rate ranging from 20% to 29.7% in the infertility population [ 15 ]. Numerous studies had indicated a lower live birth rate among patients with adenomyosis, compared to those without adenomyosis [ 18 ]. In our study population, the CLB rate in endometriosis patients with concurrent adenomyosis decreased by 49%. This finding aligned with the trends observed in other research studies. This phenomenon can be attributed to two main factors. On the one hand, patients with adenomyosis often face difficulties in achieving pregnancy. Research found that the likelihood of conception was reduced by 31% to 43% among infertile patients with adenomyosis [ 19 , 20 ]. On the other hand, those with uterine adenomyosis encountered challenges in sustaining successful pregnancies leading to live births. Wang et. al [ 21 ] highlighted a higher incidence of miscarriage in pregnant women with adenomyosis, commonly occurring after the 12th week of gestation. This might be attributed to the fibrotic changes within the uterine myometrium, which affected the mechanical and physiological expansion of the uterus. The risk of mid-term miscarriages was elevated in the population of adenomyosis [ 22 ]. When assessing pregnancy or live birth outcomes, prognostic factors for infertility often involve patient age [ 23 ]. It is well-established that age is a critical predictive factor for ovarian responsiveness and oocyte quality [ 24 ]. As age increases, reproductive potential gradually declines. Among couples actively attempting pregnancy, the incidence of infertility before the age of 45 increases exponentially with advancing age [ 25 ]. However, our findings suggested that female age was not statistically significant in the multivariate logistic regression. Nevertheless, due to its clinical relevance and to mitigate selection bias, female age was forced into the prediction model. Upon conducting stratified analyses, we observed that in the population aged ≥ 35 years, the predictive model had an AUC (95% CI) of 0.835 (0.731–0.938), while in the < 35 years age group, the AUC was 0.761 (0.712–0.811). These results indicated the applicability of the model across all age groups, with enhanced predictive capability among patients aged ≥ 35 years. Patients with AFC of less than 10 had a CLB rate that was half as high as those with AFC of ≥ 10 [ 26 ], implying the likelihood of more retrieved oocytes, a greater opportunity for fertilization and subsequently more viable embryos. Our study similarly found that a higher number of fertilized oocytes was associated with an increased CLB rate. Dobson and colleagues suggested that the pregnancy rate for single low-quality ET was 16.3%, while for single high-quality ET, it was 37.6%, among patients undergoing IVF [ 27 ]. Our research also identified the total number of high-quality embryos as an independent influencing factor on the CLB rate. Therefore, in IVF/ICSI, efforts should be made to maximize the number of fertilization, as well as high-quality embryos. Previous studies indicated that a longer duration of infertility was associated with a lower live birth rate [ 28 , 29 ], which aligned with our research findings. For couples with infertility lasting more than five years, even ICSI as a fertilization method failed to significantly improve reproductive outcomes and fertilization rate [ 30 ]. It was worth noting that the prolonged duration of infertility also impacted male fertility, as longer infertility duration was linked to lower sperm concentrations and a higher incidence of obstructive azoospermia [ 31 ]. Therefore, it is essential to emphasize the importance of public awareness and education regarding infertility. Early assessment and intervention can improve the chances of successful fertility treatment and ultimately achieving live births. Huang et.al [ 11 ]employed multifactorial logistic regression analysis and identified postoperative GnRH-a treatment, ASRM stage, and the number of high-quality transferred embryos as independent influencing factors on pregnancy outcomes. Conversely, another study found a similar pregnancy rate between 121 patients with ASRM stage 1–2 and 58 patients with ASRM stage 3–4 (66.9% vs. 65.5%) [ 32 ]. In our present study, ASRM staging did not emerge as an independent influencing factor for CLB. This discrepancy might arise from the significant heterogeneity in symptomatology and disease activity among different patients, suggesting that considering ASRM staging alone was not insufficient. Furthermore, postoperative GnRH-a usage showed no correlation with CLB, which might be attributable to the potential adverse impact of pre-IVF GnRH-a treatment on ovarian reserve, particularly in women of advanced age. There are certain limitations to be acknowledged. Our study data were derived from a single center, with results that should be tested in multicentric studies. Furthermore, due to its retrospective design, a selection bias could not be ruled out completely. Thirdly, the prediction model was established from the limited range of data documented in the hospital information system. Therefore, it was possible to ignore some important factors relevant to the study. Collectively, the first individualized nomogram of CLB rate in women with endometriosis in IVF/ICSI cycles was established. The model accurately predicts the CLB rate for endometriosis patients’ guidance, which could be helpful for reproductive endocrinologists in terms of clinical decision-making. Multi-center prospective studies should be conducted to further validate the efficacy of the nomogram.

Endometriosis is a common gynecological disease, characterized by the growth of functional endometrium outside the uterus. It is a common cause of dysmenorrhea, pelvic pain, dyspareunia and infertility [ 1 ]. Statistics show that 30–50% of endometriosis is accompanied by infertility [ 2 ]. Endometriosis-associated infertility can be induced by many factors, including disrupted ovarian function, pelvic cavity adhesion and chronic inflammation, as well as altered pelvic anatomy [ 3 , 4 ]. These patients and their partners have to suffer great psychological pressure brought about by fertility issues, which remains a great challenge. Either surgical excision of endometriosis or assisted reproductive technology (ART) will work for pregnancy [ 5 ]. ART has been widely applied to manage endometriosis associated with infertility. Early diagnosis and proactive treatment of ART seem to successfully counteract the adverse effects of endometriosis on pregnancy and live birth rates, which were comparable to tubal infertility [ 6 ]. Feichtinger et. al [ 7 ] demonstrate that endometriosis made no difference to pregnancy rate and CLB rates. However, Others take a different view. It has been observed that patients with endometriosis result in a lower pregnancy rate and CLB rate after IVF [ 8 , 9 ]. So far, research on IVF/ICSI outcomes in patients with endometriosis is still limited. For now, it is unpredictable whether the couple could become successfully pregnant and deliver a baby. Success depends on various uncertainties. Although some scoring systems have been established to evaluate the outcome after IVF/ICSI of individual patients with endometriosis, most of them are aimed at the pregnancy rate [ 10 – 12 ]. In addition, limited studies are applicable to patients associated with adenomyosis. The study aims to develop and validate a nomogram to predict the CLB rate undergoing IVF/ICSI in women with endometriosis. The model will be conducive to clinical counselling from patients and clinical management of fertility doctors.

Supplementary Material 1. Supplementary Material 1.