Laparoscopy versus laparotomy for the treatment of heterotopic pregnancy: a systematic review and meta-analysis Authors: Qifei Zhou, Kun Fu Published: 2025 ABSTRACT BackgroundHeterotopic pregnancy is a rare but potentially life-threatening condition for which both laparoscopy and laparotomy are treatment options. This meta-analysis aimed to compare the efficacy and safety of laparoscopy and laparotomy in the treatment of heterotopic pregnancies.MethodsPubMed, Web of Science, ScienceDirect, Wanfang, CNKI, and VIP databases were searched for articles published up to January 2025. Studies were included if they involved women with heterotopic pregnancies, used laparoscopy as the intervention and laparotomy as the comparator, and reported pregnancy-related outcomes. Effect estimates were presented as odds ratios (ORs) with 95% confidence intervals (CIs) and pooled analyses were performed using a random-effects model.ResultsTwenty-two studies involving 1,133 patients with heterotopic pregnancies were included in the final meta-analysis. No significant differences were found between laparoscopy and laparotomy regarding live birth rate (OR: 0.93; 95% CI: 0.59–1.46; P = 0.743), abortion rate (OR: 1.28; 95% CI: 0.84–1.97; P = 0.254), preterm birth rate (OR: 1.00; 95% CI: 0.66–1.53; P = 0.988), full-term birth rate (OR: 0.91; 95% CI: 0.65–1.27; P = 0.575), cesarean section (OR: 0.82; 95% CI: 0.41–1.61; P = 0.560), and vaginal delivery (OR: 1.12; 95% CI: 0.49–2.55; P = 0.788). Subgroup analysis indicated that laparoscopy may be associated with an increased abortion rate when only high-quality studies were pooled (OR: 1.98; 95% CI: 0.81–4.84; P = 0.135).ConclusionThis study found that intrauterine pregnancy outcomes in patients with heterotopic pregnancies in the studied population (all Chinese) were comparable between laparoscopy and laparotomy. However, the generalizability of these findings to other populations requires further investigation.Study registrationINPLASY202530131Supplementary InformationThe online version contains supplementary material available at 10.1186/s12884-025-08602-z. Introduction Heterotopic pregnancy refers to a condition in which an intrauterine pregnancy (IUP) coexists with an ectopic pregnancy (EP). Its incidence is approximately 1 in 30,000 in natural conceptions but increases to 1%–2% among patients undergoing assisted reproductive technology [1, 2]. This term specifically describes situations in which two or more embryos are implanted in different anatomical locations, with at least one embryo located within the uterine cavity. The word “heterotopic” derives from the Greek roots hetero (other) and topos (location), meaning “occurring in an abnormal position” [3]. Despite continuous advances in modern imaging and laboratory diagnostics, clinical management of heterotopic pregnancies remains challenging. The presence of an intrauterine gestational sac makes serum human chorionic gonadotropin (hCG) levels non-diagnostic, as they cannot indicate the presence of an extrauterine pregnancy. Although the identification of an intrauterine pregnancy via ultrasound generally rules out a single EP, ectopic sites in heterotopic pregnancies are often latent; approximately 50% of cases lack typical symptoms, which can lead to missed diagnoses [4, 5]. Therefore, multimodal imaging is essential for individuals who conceive through assisted reproductive technologies or for those in other high-risk populations. More than 90% of ectopic implantations occur in the fallopian tubes, but pregnancies at unusual sites (such as the cervix, cesarean scar, ovary, or abdominal cavity) have been reported [6–8]. These cases can result in uterine rupture, life-threatening hemorrhage, or hemorrhagic shock [9]. Careful and precise interventions are required to preserve intrauterine pregnancy. The primary goal of surgical treatment for heterotopic pregnancy is to remove the ectopic focus while preserving both the intrauterine pregnancy and the anatomical integrity of the uterus. In clinical decision making, it is crucial to consider the potential impact of different surgical methods on pregnancy outcomes. For patients wishing to continue their pregnancy, often owing to the high value of these pregnancies and the associated risks, surgical intervention must carefully balance hemostasis with embryo preservation. With the rise in minimally invasive techniques, laparoscopic surgery has become the mainstay of treatment for heterotopic pregnancies because of its advantages of reduced trauma and quicker recovery [10, 11]. However, academic debate continues over its effect on intrauterine pregnancy compared with traditional laparotomy. This study aimed to systematically evaluate the impact of laparoscopy versus laparotomy on intrauterine pregnancy outcomes in heterotopic pregnancies and to provide evidence-based guidance for clinical practice. Materials and methods Data sources, search strategy, and selection criteria This study strictly followed the standardized process outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [12]. This study was registered on the INPLASY platform (number: INPLASY202530131). Women with heterotopic pregnancy were included if laparoscopy was used as the intervention and laparotomy as the comparator, and if the study reported pregnancy-related outcomes. No restrictions were placed on the language, geographic region, or publication status of the publication. A dual strategy was adopted for the literature search. (1) Electronic searches included three major English databases: PubMed, Web of Science, and ScienceDirect, and three major Chinese databases: Wanfang Data, China National Knowledge Infrastructure (CNKI), and VIP. Boolean logic expressions were constructed using “heterotopic pregnancy” and “laparoscopic surgery” as core search terms (for the complete search strategy, please refer to Supplementary File 1). The search time frame was extended to January 2025 to maximize coverage. (2) Manual screening of the reference lists of the included studies was conducted using citation tracing to reduce the risk of missing literature due to limitations in database indexing. A two-person back-to-back screening was performed. After standardized training, two reviewers independently conducted the following tasks according to the PRISMA flowchart: (1) in the initial screening stage, studies that were clearly inconsistent with inclusion criteria were excluded based on titles and abstracts; (2) in the full-text review stage, standardized forms were used for eligibility assessment; (3) during data extraction, key indicators were cross-verified. In cases of disagreement, a three-level arbitration process was initiated: first, a double-blind discussion to reach consensus; second, if unresolved, referral to a third senior reviewer for adjudication; and third, final resolution during a consensus meeting. The inclusion criteria were as follows: (1) diagnosis of heterotopic pregnancy via imaging; (2) intervention involving laparoscopic removal of EP foci; (3) comparator group treated with laparotomy; (4) outcomes including intrauterine pregnancy indicators such as live birth rate, abortion rate, preterm birth rate, full-term birth rate, cesarean section, and vaginal delivery; and (5) study design classified as observational studies or clinical trials. Additional criteria included a minimum total sample size of ≥ 5 for single-center studies. In cases of multiple publications from the same research team, the study with the largest sample size or the most complete follow-up period was selected. Exclusion criteria were: (1) reviews/meta-analyses, commentaries, conference abstracts, or case reports with a sample size ≤ 4; (2) studies that did not report outcomes separately for laparoscopic and laparotomy groups or did not distinguish heterotopic pregnancy from simple ectopic pregnancy; (3) lack of explanation for surgical method selection or absence of postoperative pregnancy monitoring; and (4) single-arm studies reporting only laparoscopic or laparotomy procedures. Data collection and quality assessment A structured data extraction template was developed to collect the following information: first author; publication year; study design; region; sample size; average patient age; use of assisted reproductive technology; intervention and control groups; and outcome. Abortion was defined as termination of pregnancy before 28 weeks with a fetal weight of less than 1,000 g; preterm birth as delivery between 28 and 37 weeks; full-term delivery between 37 and 42 weeks; and live birth as the sum of preterm and full-term deliveries, excluding stillbirth and late abortion. It is important to note that this definition aligns with the clinical guidelines used in the included studies from China, but it differs from the current WHO standard. This definitional framework was consistently applied across the included studies, and as a meta-analysis, our analysis was bound by the definitions used in the primary literature. We acknowledge this as a potential limitation in harmonizing with international standards. Two authors independently assessed the quality of the included studies using the Newcastle-Ottawa Scale (NOS) under double-blind conditions. The NOS includes three domains: selection bias control (case representativeness, appropriateness of the control source, exposure confirmation, and clarity of the study starting point), comparability (adjustment for key confounders using statistical methods), and outcome assessment reliability (blinding, follow-up completeness, and objectivity of outcome criteria). Studies scoring 7–9 points were classified as high quality, 4–6 as medium quality, and ≤ 3 as having a high risk of bias [13]. Any disagreements during data extraction or quality assessment were resolved by a third reviewer with reference to the original study to ensure accuracy and consistency. Statistical analysis The treatment effects of laparoscopy versus laparotomy were considered as categorical outcomes, and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated before pooling. A random-effects model using the DerSimonian–Laird method was employed to incorporate clinical heterogeneity across studies [14]. Heterogeneity was assessed using the I2 statistic and Q test, with I2 ≥ 50.0% or P < 0.10 considered indicative of significant heterogeneity [15, 16]. Leave-one-out cross-validation was used to evaluate the influence of individual studies on the pooled estimates [17]. Pre-specified subgroup analyses were planned a priori to investigate the potential sources of heterogeneity. These included analyses of study quality (high vs. low, as defined by the NOS score) and patient conception method (assisted reproductive technology [ART] vs. non-ART or mixed conception). An additional subgroup analysis based on sample size was conducted post hoc. Differences between subgroups were tested using interaction P value [18]. The certainty of evidence for each primary outcome (live birth, abortion, preterm birth, full-term birth, cesarean section, and vaginal delivery) was assessed using the GRADE (Grading of Recommendations, Assessment, Development, and Evaluations) methodology [19]. According to GRADE, evidence from observational studies (such as all studies included in this meta-analysis) begins with low-certainty. Certainty can be downgraded based on the following five factors: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Conversely, the certainty can be upgraded for a large magnitude of effect, a dose-response gradient, or if all plausible confounding factors reduce the demonstrated effect. To assess potential publication bias, a funnel plot was constructed using Egger’s and Begg’s tests [20, 21]. If a funnel plot asymmetry was observed, the trim-and-fill method was used to adjust for publication bias [22]. All hypothesis tests were two-sided, and P < 0.05 was considered statistically significant. Data analysis was performed using Stata software (version 12.0; StataCorp LLC, College Station, Texas, USA). Results Literature search A total of 7,258 relevant studies were identified through the database search. After initial screening based on titles and abstracts, 7,185 studies were excluded, and the remaining 73 studies underwent a full-text evaluation. After detailed review, 50 additional studies were excluded from the analysis. Ultimately, 22 studies were included in the final analysis [23–44]. The detailed literature screening process is shown in Fig. 1. In addition, no new studies meeting the inclusion criteria were identified through manual reference list searches. Fig. 1PRISMA flowchart illustrating the study selection process PRISMA flowchart illustrating the study selection process Study characteristics The baseline characteristics of the included studies are presented in Table 1. All the studies had a retrospective design and were conducted in China. Collectively, the 22 studies included 1,133 patients with heterotopic pregnancies, with individual sample sizes ranging from 5 to 264 cases. Fifteen studies exclusively included patients who conceived through assisted reproductive technology, whereas the remaining Seven included patients who conceived via both assisted reproduction and natural conception. The results of the quality assessments are presented in Table 1. Five studies were assessed as high quality, and the remaining 17 were rated as moderate quality. Table 1Baseline characteristics of included studies and involved patientsStudyStudy designCountrySample sizeAge (years)ART proportion (%)InterventionControlNOSLin 2011 [23]RetroChina14 (11/3)31.6100.0%LaparoscopyLaparotomy6Tan 2013 [24]RetroChina31 (13/18)31.3100.0%LaparoscopyLaparotomy6Yu 2014 [25]RetroChina20 (6/14)31.6100.0%LaparoscopyLaparotomy7Nie 2015 [26]RetroChina81 (36/45)30.298.8%LaparoscopyLaparotomy6Wan 2016 [27]RetroChina30 (17/13)30.096.7%LaparoscopyLaparotomy7Song 2017 [28]RetroChina31 (23/8)32.1100.0%LaparoscopyLaparotomy6Sun 2017 [29]RetroChina5 (2/3)30.880.0%LaparoscopyLaparotomy5Ma 2017 [30]RetroChina39 (37/2)31.091.5LaparoscopyLaparotomy6Chang 2017 [31]RetroChina40 (20/20)31.2100.0%LaparoscopyLaparotomy5Liu 2018 [32]RetroChina46 (15/31)30.787.9LaparoscopyLaparotomy7Jiang 2018 [33]RetroChina10 (7/3)29.8100.0%LaparoscopyLaparotomy6Huang 2018 [34]RetroChina6 (5/1)30.5100.0%LaparoscopyLaparotomy5Gu 2019 [35]RetroChina18 (11/7)31.6100.0%LaparoscopyLaparotomy6Hu 2019 [36]RetroChina264 (145/119)30.8100.0%LaparoscopyLaparotomy5Li 2020 [37]RetroChina52 (36/16)30.9100.0%LaparoscopyLaparotomy6Xu 2020 [38]RetroChina29 (22/7)32.0100.0%LaparoscopyLaparotomy6Qiao 2021 [39]RetroChina66 (30/36)NA100.0%LaparoscopyLaparotomy6Chen 2022 [40]RetroChina96 (69/27)30.997.1%LaparoscopyLaparotomy6Chen 2022 [41]RetroChina86 (62/24)30.7100.0%LaparoscopyLaparotomy7Wang 2022 [42]RetroChina32 (20/12)30.2100.0%LaparoscopyLaparotomy6Zheng 2023 [43]RetroChina28 (24/4)31.390.6LaparoscopyLaparotomy5Chen 2023 [44]RetroChina109 (62/47)29.8100.0%LaparoscopyLaparotomy7 Table 1Baseline characteristics of included studies and involved patientsStudyStudy designCountrySample sizeAge (years)ART proportion (%)InterventionControlNOSLin 2011 [23]RetroChina14 (11/3)31.6100.0%LaparoscopyLaparotomy6Tan 2013 [24]RetroChina31 (13/18)31.3100.0%LaparoscopyLaparotomy6Yu 2014 [25]RetroChina20 (6/14)31.6100.0%LaparoscopyLaparotomy7Nie 2015 [26]RetroChina81 (36/45)30.298.8%LaparoscopyLaparotomy6Wan 2016 [27]RetroChina30 (17/13)30.096.7%LaparoscopyLaparotomy7Song 2017 [28]RetroChina31 (23/8)32.1100.0%LaparoscopyLaparotomy6Sun 2017 [29]RetroChina5 (2/3)30.880.0%LaparoscopyLaparotomy5Ma 2017 [30]RetroChina39 (37/2)31.091.5LaparoscopyLaparotomy6Chang 2017 [31]RetroChina40 (20/20)31.2100.0%LaparoscopyLaparotomy5Liu 2018 [32]RetroChina46 (15/31)30.787.9LaparoscopyLaparotomy7Jiang 2018 [33]RetroChina10 (7/3)29.8100.0%LaparoscopyLaparotomy6Huang 2018 [34]RetroChina6 (5/1)30.5100.0%LaparoscopyLaparotomy5Gu 2019 [35]RetroChina18 (11/7)31.6100.0%LaparoscopyLaparotomy6Hu 2019 [36]RetroChina264 (145/119)30.8100.0%LaparoscopyLaparotomy5Li 2020 [37]RetroChina52 (36/16)30.9100.0%LaparoscopyLaparotomy6Xu 2020 [38]RetroChina29 (22/7)32.0100.0%LaparoscopyLaparotomy6Qiao 2021 [39]RetroChina66 (30/36)NA100.0%LaparoscopyLaparotomy6Chen 2022 [40]RetroChina96 (69/27)30.997.1%LaparoscopyLaparotomy6Chen 2022 [41]RetroChina86 (62/24)30.7100.0%LaparoscopyLaparotomy7Wang 2022 [42]RetroChina32 (20/12)30.2100.0%LaparoscopyLaparotomy6Zheng 2023 [43]RetroChina28 (24/4)31.390.6LaparoscopyLaparotomy5Chen 2023 [44]RetroChina109 (62/47)29.8100.0%LaparoscopyLaparotomy7 Baseline characteristics of included studies and involved patients All included studies reported the effects of laparoscopy versus laparotomy on live birth rates. No significant difference was found between the two surgical methods (OR: 0.93; 95% CI: 0.59–1.46; P = 0.743; Fig. 2), and no significant heterogeneity was observed (I2 = 23.9%; P = 0.156). Sensitivity analysis indicated that the overall conclusion was stable and not influenced by any single study (Supplementary File 2). Subgroup analysis suggested that sample size and patient status may influence outcomes; however, no statistically significant differences were found across the subgroups (Table 2). No significant publication bias was detected (Egger’s test: P = 0.361; Begg’s test: P = 0.450; Supplementary File 3). Fig. 2Treatment effect of laparoscopy versus laparotomy on live birth rate Treatment effect of laparoscopy versus laparotomy on live birth rate Table 2Subgroup analyses for pregnancy outcomesOutcomesFactorsSubgroupsOR and 95%CIP valueI2 (%)Q statisticInteraction testLive birth rateSample size≥ 501.22 (0.59–2.49)0.59344.60.0940.091< 500.68 (0.40–1.18)0.1730.00.478Condition of patientsART0.75 (0.43–1.30)0.30017.30.2600.073Both1.39 (0.70–2.75)0.34918.70.292Study qualityHigh0.51 (0.21–1.24)0.1350.00.8230.126Moderate1.07 (0.63–1.81)0.81230.60.113Abortion rateSample size≥ 501.10 (0.47–2.54)0.82937.60.1410.464< 501.49 (0.84–2.64)0.1740.00.710Condition of patientsART1.45 (0.86–2.44)0.1661.70.4310.480Both0.99 (0.45–2.19)0.9776.50.375Study qualityHigh1.98 (0.81–4.84)0.1350.00.8230.300Moderate1.13 (0.66–1.92)0.65112.20.314Preterm birth rateSample size≥ 501.21 (0.74–1.97)0.4450.00.5510.125< 500.55 (0.23–1.32)0.1830.00.994Condition of patientsART1.02 (0.65–1.61)0.9200.00.9320.859Both0.86 (0.25–3.02)0.8150.00.573Study qualityHigh0.36 (0.07–1.90)0.2300.00.7260.213Moderate1.08 (0.69–1.67)0.7380.00.960Full-term birth rateSample size≥ 501.18 (0.56–2.47)0.66357.40.0950.648< 500.82 (0.47–1.43)0.4830.00.902Condition of patientsART0.80 (0.54–1.19)0.2730.00.7210.240Both1.26 (0.67–2.38)0.4770.00.569Study qualityHigh0.78 (0.33–1.88)0.5800.00.9310.714Moderate0.93 (0.65–1.34)0.7040.00.488 Table 2Subgroup analyses for pregnancy outcomesOutcomesFactorsSubgroupsOR and 95%CIP valueI2 (%)Q statisticInteraction testLive birth rateSample size≥ 501.22 (0.59–2.49)0.59344.60.0940.091< 500.68 (0.40–1.18)0.1730.00.478Condition of patientsART0.75 (0.43–1.30)0.30017.30.2600.073Both1.39 (0.70–2.75)0.34918.70.292Study qualityHigh0.51 (0.21–1.24)0.1350.00.8230.126Moderate1.07 (0.63–1.81)0.81230.60.113Abortion rateSample size≥ 501.10 (0.47–2.54)0.82937.60.1410.464< 501.49 (0.84–2.64)0.1740.00.710Condition of patientsART1.45 (0.86–2.44)0.1661.70.4310.480Both0.99 (0.45–2.19)0.9776.50.375Study qualityHigh1.98 (0.81–4.84)0.1350.00.8230.300Moderate1.13 (0.66–1.92)0.65112.20.314Preterm birth rateSample size≥ 501.21 (0.74–1.97)0.4450.00.5510.125< 500.55 (0.23–1.32)0.1830.00.994Condition of patientsART1.02 (0.65–1.61)0.9200.00.9320.859Both0.86 (0.25–3.02)0.8150.00.573Study qualityHigh0.36 (0.07–1.90)0.2300.00.7260.213Moderate1.08 (0.69–1.67)0.7380.00.960Full-term birth rateSample size≥ 501.18 (0.56–2.47)0.66357.40.0950.648< 500.82 (0.47–1.43)0.4830.00.902Condition of patientsART0.80 (0.54–1.19)0.2730.00.7210.240Both1.26 (0.67–2.38)0.4770.00.569Study qualityHigh0.78 (0.33–1.88)0.5800.00.9310.714Moderate0.93 (0.65–1.34)0.7040.00.488 Subgroup analyses for pregnancy outcomes Abortion rate Data on abortion rates were available for 20 studies. The combined results indicated no significant difference between laparoscopy and laparotomy (OR: 1.28; 95% CI: 0.84–1.97; P = 0.254; Fig. 3). The heterogeneity was minimal (I2 = 0.4%; P = 0.452). Sensitivity analysis confirmed the stability of the results, with no single study significantly affecting the conclusions (Supplementary File 2). In our pre-specified subgroup analysis by study quality, a non-significant trend towards an increased abortion rate with laparoscopy was observed in the high-quality studies (OR: 1.98; 95% CI: 0.81–4.84; P = 0.135; Table 2). It is important to note that this analysis was exploratory, and the result was not statistically significant. No significant publication bias was found (Egger’s test: P = 0.497; Begg’s test: P = 0.417; Supplementary File 3). Fig. 3Treatment effect of laparoscopy versus laparotomy on abortion rate Treatment effect of laparoscopy versus laparotomy on abortion rate Preterm birth rate Thirteen studies reported preterm birth rates. The results showed no association between laparoscopy and increased risk of preterm birth compared to laparotomy (OR: 1.00; 95% CI: 0.66–1.53; P = 0.988; Fig. 4). There was no evidence of heterogeneity (I2 = 0.0%; P = 0.944). The pooled conclusions remained consistent when the individual studies were removed (Supplementary File 2). Subgroup analyses revealed no significant differences in the preterm birth rates between the two surgical approaches (Table 2). Although Begg’s test showed no publication bias (P = 0.360), Egger’s test suggested possible bias (P = 0.072). However, this conclusion remains unchanged after correction using the trim-and-fill method (Supplementary File 3). Fig. 4Treatment effect of laparoscopy versus laparotomy on preterm birth rate Treatment effect of laparoscopy versus laparotomy on preterm birth rate Full-term birth rate Fourteen studies reported data on the full-term birth rates. The pooled results showed no significant difference between laparoscopy and laparotomy (OR: 0.91; 95% CI: 0.65–1.27; P = 0.575; Fig. 5). No heterogeneity was found among the studies (I2 = 0.0%, P = 0.715). The sensitivity analysis confirmed the robustness of the results, which were not affected by the exclusion of individual studies (Supplementary File 2). The subgroup analysis also revealed no statistically significant differences across the subgroups (Table 2). No publication bias was identified (Egger’s test: P = 0.308; Begg’s test: P = 0.511; Supplementary File 3). Fig. 5Treatment effect of laparoscopy versus laparotomy on full-term birth rate Treatment effect of laparoscopy versus laparotomy on full-term birth rate Cesarean section and vaginal delivery Seven studies reported on cesarean sections, and four reported on vaginal delivery (Fig. 6). No significant differences were found between laparoscopy and laparotomy for either cesarean section (OR: 0.82; 95% CI: 0.41–1.61; P = 0.560) or vaginal delivery (OR: 1.12; 95% CI: 0.49–2.55; P = 0.788). No heterogeneity was observed for cesarean sections (I2 = 0.0%, P = 0.456) or vaginal deliveries (I2 = 0.0%, P = 0.801). Fig. 6Treatment effect of laparoscopy versus laparotomy on the incidences of caesarean section and vaginal delivery Treatment effect of laparoscopy versus laparotomy on the incidences of caesarean section and vaginal delivery Summary of findings and certainty of evidence (GRADE) The results for all primary outcomes and the GRADE assessment of the certainty of evidence are presented in Table 3. For all outcomes, the evidence was assessed as having low certainty. The primary reasons for downgrading the evidence were as follows: (1) Study design (risk of bias): All evidence originated from non-randomized observational studies, which are inherently susceptible to confounding and selection biases. This led to a downgrade by one level; and (2) Imprecision: For most outcomes, the 95% CI included the possibility of both no effect and an appreciable benefit or harm. This imprecision in the effect estimate led to a further downgrading to one level. Table 3Summary of findings and GRADE assessment: laparoscopy vs. Laparotomy for heterotopic pregnancyOutcomesOR and 95%CINumber of studiesCertainty of evidenceCommentsLive birth rate0.93 (0.59–1.46)22Lowa, bEvidence suggests little to no differenceAbortion rate1.28 (0.84–1.97)20Lowa, bEvidence suggests little to no differencePreterm birth rate1.00 (0.66–1.53)13Lowa, bEvidence suggests little to no differenceFull-term birth rate0.91 (0.65–1.27)14Lowa, bEvidence suggests little to no differenceCaesarean section0.82 (0.41–1.61)7Lowa, bEvidence suggests little to no differenceVaginal delivery1.12 (0.49–2.55)4Lowa, bEvidence suggests little to no differenceaDowngraded one level for serious risk of bias: All evidence comes from retrospective observational studiesbDowngraded one level for serious imprecision: Confidence intervals are wide and include both potential benefit and harm Table 3Summary of findings and GRADE assessment: laparoscopy vs. Laparotomy for heterotopic pregnancyOutcomesOR and 95%CINumber of studiesCertainty of evidenceCommentsLive birth rate0.93 (0.59–1.46)22Lowa, bEvidence suggests little to no differenceAbortion rate1.28 (0.84–1.97)20Lowa, bEvidence suggests little to no differencePreterm birth rate1.00 (0.66–1.53)13Lowa, bEvidence suggests little to no differenceFull-term birth rate0.91 (0.65–1.27)14Lowa, bEvidence suggests little to no differenceCaesarean section0.82 (0.41–1.61)7Lowa, bEvidence suggests little to no differenceVaginal delivery1.12 (0.49–2.55)4Lowa, bEvidence suggests little to no differenceaDowngraded one level for serious risk of bias: All evidence comes from retrospective observational studiesbDowngraded one level for serious imprecision: Confidence intervals are wide and include both potential benefit and harm Summary of findings and GRADE assessment: laparoscopy vs. Laparotomy for heterotopic pregnancy aDowngraded one level for serious risk of bias: All evidence comes from retrospective observational studies bDowngraded one level for serious imprecision: Confidence intervals are wide and include both potential benefit and harm Discussion Summary of findings This study provided a comprehensive comparison of pregnancy-related outcomes between laparoscopy and laparotomy in women with heterotopic pregnancies. Despite the rarity of this condition, our systematic review and meta-analysis successfully synthesized the available evidence, including 22 retrospective studies involving 1,133 patients. The findings indicated no statistically significant differences between laparoscopic surgery and laparotomy in terms of live birth rate, abortion rate, preterm birth rate, full-term birth rate, cesarean section, and vaginal delivery. Sensitivity analysis showed that the combined effect size was robust. After the exclusion of individual studies, the pooled conclusions remained stable. Subgroup analysis indicated that in high-quality studies, laparoscopic surgery might be associated with an increased risk of abortion compared to laparotomy. According to the GRADE framework, the overall certainty of evidence for all primary outcomes was assessed as low primarily because of the observational nature of the included studies and imprecision in the effect estimates. Therefore, our results should be interpreted with caution. Methodological considerations A dual literature search strategy that combines electronic searches across multiple databases and manual screening of reference lists was employed to minimize the risk of literature omission. However, despite a comprehensive search, the fact that all included studies were conducted in China may limit the generalizability of the findings to other regions. The two-person back-to-back screening mechanism combined with a three-level arbitration process enhanced the reliability of the study selection. The Newcastle-Ottawa scale was used to standardize the quality assessment of the included studies. Nevertheless, as most studies were of moderate quality, potential limitations in study design and execution may have affected the validity of the results. Pregnancy-related outcomes Our meta-analysis showed no significant difference in live birth rate between the laparoscopy and laparotomy groups. The absence of significant heterogeneity across studies and stable results from the sensitivity analysis suggest the robustness of the conclusion. However, the subgroup analysis indicated that sample size and patient status might influence the comparative impact of the two surgical methods, although no statistically significant differences were found. These factors may act as potential confounders. Future studies with larger and more diverse samples are needed to further explore this relationship. Regarding abortion rate, the pooled results also showed no significant differences between the two surgical methods. Similar to the live birth rate, no heterogeneity was detected, and the conclusion was stable in the sensitivity analysis. As part of our pre-specified investigation of heterogeneity, we observed a trend in the subgroup analysis of high-quality studies suggesting a potential increase in abortion rates with laparoscopy, although this was not statistically significant. This exploratory finding generates the hypothesis that any potential risk associated with laparoscopy (e.g., related to pneumoperitoneum) might be more accurately detected in better-designed studies; however, it must be interpreted with extreme caution and requires confirmation in future research. The surgical approach did not affect the risk of preterm delivery. The pathogenesis of preterm birth is often associated with subclinical uterine infection or inflammation. The fact that neither laparoscopy nor laparotomy increased this risk implies that both procedures, when performed with careful tissue handling and adherence to aseptic techniques, do not precipitate a sufficiently significant inflammatory or infectious insult to initiate the pathway to preterm labor later in gestation. This is a reassuring finding, as it indicates that the initial surgical trauma does not predispose the uterus to subsequent dysfunction. The lack of heterogeneity and consistency of the results in the sensitivity analysis support the reliability of this conclusion. However, a potential publication bias should be considered when interpreting these results. Analysis of the full-term birth rate also revealed no significant differences between laparoscopy and laparotomy, with no observed heterogeneity and stable pooled conclusions. Subgroup analysis confirmed consistent findings across all subgroups, strengthening the conclusion that both procedures had similar effects on full-term birth rates. Regarding cesarean section and vaginal delivery, the limited number of studies reporting these outcomes may have reduced the precision of the results. Nonetheless, the absence of significant differences and heterogeneity suggests that these two surgical methods do not substantially influence delivery modes. Broadening the context of reproductive outcomes While our analysis focused on the comparative impact of the two surgical approaches, it is crucial to recognize that overall reproductive success in women with heterotopic pregnancy is multifactorial. Our finding that the surgical method itself did not significantly alter key pregnancy outcomes is reassuring. This suggests that, within the selected patient population for each procedure, surgical intervention is not the primary determinant of the fate of intrauterine pregnancy. This aligns with the broader understanding that patient-specific factors often have a stronger influence on reproductive success. For instance, a history of cesarean section—a common antecedent in many gynecological patients—has been independently associated with lower implantation rates, ongoing pregnancy rates, and live birth rates following embryo transfer compared to a history of prior vaginal delivery [45, 46]. This underscores the fact that baseline patient characteristics, including uterine integrity and surgical history, constitute a critical background against which the effects of any new intervention, such as surgery for heterotopic pregnancy, must be evaluated. Our results should, therefore, be interpreted as indicating that laparoscopy does not introduce additional harm to intrauterine pregnancy compared to laparotomy within the context of these pre-existing patient factors. Implications for clinical practice Based on our findings, laparoscopy can be considered a viable alternative to laparotomy for women with heterotopic pregnancy as it does not significantly affect pregnancy-related outcomes such as live birth rate, abortion rate, preterm birth rate, full-term birth rate, cesarean section, or vaginal delivery. Although the overall clinical outcomes of the two procedures were comparable, laparoscopic surgery offers notable benefits, including minimal invasiveness, faster postoperative recovery, and better short-term quality-of-life scores, particularly in regional tertiary medical centers [47, 48]. The favorable safety profile of minimally invasive techniques is corroborated by evidence from other gynecological oncology surgeries. A meta-analysis demonstrated that sentinel lymph node biopsy via laparoscopy or robotics was associated with significantly fewer surgical complications than systematic lymphadenectomy without compromising oncological safety [49]. This principle of reduced morbidity with minimally invasive precision surgery likely extends to the management of heterotopic pregnancies and supports the choice of laparoscopy in eligible patients. Furthermore, the safety and feasibility of such minimally invasive sentinel node mapping have been demonstrated even in older patient populations [50], reinforcing the idea that age alone should not preclude a patient from benefiting from a less-invasive approach when their functional status is adequate. Despite its advantages in visualizing anatomical structures and surgical duration similar to that of laparotomy, laparoscopic surgery involves more complex preoperative preparation and provides less operational freedom, necessitating strict adherence to surgical indications [51]. For patients with stable hemodynamics and pelvic hematoceles less than 500 ml, laparoscopic surgery is preferred [52]. During the operation, key measures should be taken: maintaining pneumoperitoneal pressure at ≤ 12 mmHg; continuously monitoring end-tidal carbon dioxide levels; achieving precise hemostasis using bipolar electrocoagulation; irrigating with warm normal saline at 37 °C; and using non-traumatic grasping forceps to gently handle tissues—all of which may improve the likelihood of continued pregnancy after surgery [53]. Patient perspective: pain and tolerability In addition to objective surgical outcomes, the patient’s subjective experience is a critical component of high-quality care. Our study, by its design, could not capture data on postoperative pain or patient satisfaction. However, insights from diagnostic gynecological procedures highlight the importance of this dimension. Research on outpatient hysteroscopy has shown that patient pain levels can be significantly influenced by underlying pathology, such as tubal obstruction [54]. This underscores the fact that patient factors and specific clinical scenarios directly impact the procedural experience. Furthermore, the good tolerability and diagnostic accuracy of minimally invasive outpatient procedures for evaluating tubal patency [55] set a precedent for patient-centered care in gynecology. Translating this to the surgical management of heterotopic pregnancy, the well-documented advantages of laparoscopy, including reduced postoperative pain, smaller incisions, and quicker return to normal activities, are not merely convenient but are fundamental to improving the overall patient experience and reducing the burden of treatment. Therefore, when pregnancy outcomes are equivalent, as suggested by our findings, the choice of a less painful and more patient-friendly surgical approach such as laparoscopy is strongly justified from a holistic clinical perspective. Interpretation of heterogeneity and generalizability Although our meta-analysis demonstrated a low statistical heterogeneity for all primary outcomes, this finding should be interpreted with caution. In our study, the consistently low I2 value may largely reflect the narrow clinical and demographic spectrum of the included studies, as all studies were conducted in a single country (China). This geographical homogeneity implies potential similarities in surgical protocols, patient genetics, baseline health characteristics, and perioperative care, which may not be representative of other populations. Consequently, the apparent consistency of our results may be specific to this context and should not be extrapolated to assume that the same effects would be observed in more genetically and clinically diverse populations. True clinical heterogeneity across global practice settings may be masked by the methodological homogeneity of the available literature. Therefore, the main limitation highlighted by our low heterogeneity is not internal inconsistency but rather a significant constraint on the generalizability of our findings. Strengths and limitations This study is based on multiple retrospective studies. The large sample size enabled a quantitative comparison of the efficacy of laparoscopic and open abdominal surgery in the treatment of heterotopic pregnancy, making our results more reliable than those of individual studies. In addition, we conducted exploratory sensitivity and subgroup analyses. However, several limitations remain. (1) All included studies were retrospective, making the results subject to selection and confounding bias. Most critically, our analysis was at a high risk of confounding by indication. In clinical practice, the choice between laparoscopy and laparotomy is not always randomized. Surgeons are more likely to select laparotomy for patients who are hemodynamically unstable, have a large volume of hemoperitoneum, present with a ruptured ectopic pregnancy, or have complex implantation sites. These patients inherently have more severe clinical status, which can lead to worse pregnancy outcomes. Conversely, laparoscopy is typically reserved for stable patients with less complex presentations. Therefore, any observed difference in outcomes could be profoundly influenced by baseline differences in patient acuity and disease severity rather than by the surgical approach itself. This unmeasured confounding represents the most significant threat to the validity of our comparative results; (2) differences in patient characteristics and surgical procedures between laparoscopic and open surgery could affect pregnancy outcomes; and (3) our analysis focused on pregnancy and neonatal outcomes; consequently, we could not systematically compare important maternal short-term outcomes, such as postoperative pain, length of hospital stay, cosmetic results, and recovery time. Although these are well-documented advantages of laparoscopy, the lack of comparative data on these parameters in our study represents a significant limitation, as they are crucial for comprehensive clinical decision-making; (4) The analysis was based on published studies, making it vulnerable to publication bias. This also limited the ability to perform more extensive exploratory analyses. (5) All the included studies were conducted in China, and the generalizability of the conclusions was therefore limited. Conclusion This meta-analysis provides valuable insights into pregnancy-related outcomes after laparoscopy and laparotomy for heterotopic pregnancies. These findings, supported by low-certainty evidence, indicate that pregnancy outcomes are comparable between laparoscopy and laparotomy. Future high-quality prospective studies are needed to provide high-certainty evidence and explore their impact on important short-term maternal outcomes. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.