Trimester-specific safety of laparoscopic versus open abdominal surgery during pregnancy: systematic review and meta-analysis

Abstract

Background

Current guidelines endorse laparoscopy for non-obstetric abdominal surgery during pregnancy regardless of trimester but recent data suggest trimester-specific fetal risks. The aim of this study was to compare the maternal and fetal safety of laparoscopic versus open surgery, focusing on trimester-specific and pathology-stratified outcomes.

Methods

This systematic review and meta-analysis was conducted according to the PRISMA 2020 guidelines and was registered in PROSPERO, the international prospective register of systematic reviews (2026 CRD420261295995). PubMed, Embase, and Cochrane Library databases were searched for comparative studies (RCTs and cohort studies) of laparoscopic versus open surgery for acute appendicitis, gallstone disease, and adnexal masses. Random-effects meta-analysis synthesized data on fetal loss, preterm delivery, maternal complications, and length of hospital stay.

Results

In total, 22 studies comprising 28 160 pregnant women (15 786 of whom underwent laparoscopic surgery and 12 374 of whom underwent open surgery) were included. Laparoscopy was associated with a higher risk of fetal loss than open surgery (OR 2.02 (95% c.i. 1.40 to 2.92); P < 0.001). A first-trimester subgroup analysis showed a persistent trend toward higher fetal loss with laparoscopy (OR 1.35 (95% c.i. 0.84 to 2.19)). Laparoscopy reduced preterm delivery (OR 0.56 (95% c.i. 0.34 to 0.94); P = 0.020) and maternal complications (OR 0.45 (95% c.i. 0.30 to 0.68); P < 0.001). Trimester-specific analysis revealed a significantly elevated risk of composite adverse fetal outcomes for laparoscopy during the second trimester (OR 2.35 (95% c.i. 1.15 to 4.77); P = 0.020) and a similar trend in the third trimester.

Conclusion

Laparoscopy confers maternal benefits and reduces preterm delivery but is associated with higher fetal loss and elevated composite adverse outcomes in the second trimester.

This meta-analysis of 28 160 pregnant women reveals a critical clinical trade-off: while laparoscopy significantly mitigates the risk of preterm delivery and maternal complications, it is associated with a higher aggregate risk of fetal loss and elevated adverse outcomes during the second trimester. These divergent safety signals challenge current guidelines endorsing laparoscopy regardless of gestational age, advocating instead for a risk-stratified surgical strategy.

Introduction

Non-obstetric abdominal surgery constitutes a unique clinical challenge, estimated to affect up to 2% of all pregnancies^1,2. Acute appendicitis, gallstone disease, and adnexal masses represent the most frequent indications for intervention. Over the past three decades, the surgical paradigm has shifted decisively toward minimally invasive techniques. Current guidelines from the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) state that laparoscopy can be safely performed ‘during any trimester of pregnancy’, citing maternal benefits such as reduced narcotic requirements and a shorter length of hospital stay³. Similarly, the European Association for Endoscopic Surgery (EAES) rapid guideline recommends laparoscopy but explicitly notes the ‘very low certainty’ of existing evidence regarding fetal safety, particularly in advanced gestation⁴.

Indeed, the evidentiary basis underpinning these broad recommendations relies heavily on historical, single-centre case series that may not reflect contemporary practice or the full spectrum of surgical risks^3,4. This consensus is increasingly being scrutinized in the era of big data. Recent large-scale, population-based studies have reported divergent safety signals: while some demonstrate clear maternal recovery benefits, others indicate a potential association with adverse fetal outcomes, such as fetal loss, that were not evident in smaller cohorts^5,6. These conflicting findings postulate a mechanistic trade-off: the reduction in uterine manipulation via laparoscopy may prevent preterm labour, yet the physiological effects of pneumoperitoneum—such as fetal acidosis and maternal haemodynamic alterations—may pose distinct risks to fetal viability^7,8. Furthermore, previous systematic reviews have often aggregated heterogeneous pathologies, potentially obscuring disease-specific risks—such as the inflammatory impact of perforated appendicitis versus non-infectious adnexal pathology⁹.

Given this persistent clinical equipoise and the scarcity of high-quality evidence cited by international guidelines, a systematic review and meta-analysis of contemporary comparative studies was conducted. Unlike prior reviews, this study integrates data from recent large-scale national databases with pivotal multicentre cohorts, adhering to rigorous stratification by pathology and gestational age. The objective was to elucidate the trimester-specific safety profile of laparoscopy and to definitively examine the trade-offs between fetal loss and preterm delivery, thereby informing future updates to clinical guidelines.

Methods

Protocol and search strategy

This systematic review and meta-analysis was conducted according to the PRISMA 2020 guidelines¹⁰ (Table S1) and was registered in PROSPERO, the international prospective register of systematic reviews (2026 CRD420261295995). A comprehensive literature search was performed across PubMed, Embase, and Cochrane Library databases from inception to 10 January 2026. The search strategy employed a combination of Medical Subject Headings (MeSH) and Embase Subject Headings (Emtree) and free-text terms related to ‘pregnancy’, ‘laparoscopy’, ‘laparotomy’, ‘appendicitis’, ‘cholelithiasis’, and ‘adnexal masses’, without language restrictions. To ensure the inclusion of the most recent high-quality evidence, reference lists of relevant reviews and included articles were manually screened. The full search strings are detailed in Table S2.

Eligibility criteria and outcomes

Eligibility criteria were established in accordance with the Population, Intervention, Comparison, Outcomes (PICO) framework. Eligible studies comprised comparative analyses of pregnant women diagnosed with non-obstetric abdominal pathologies who underwent either laparoscopic or open surgical intervention. Inclusion was contingent upon the reporting of extractable raw data for at least one primary outcome, defined as fetal loss or preterm delivery, or secondary endpoints including maternal complications and length of hospital stay. Investigations comparing operative versus non-operative management or antepartum versus postpartum timing were excluded, as were single-arm case series lacking a control group. To preclude data duplication and ensure statistical independence, overlapping data sets derived from identical administrative registries were meticulously cross-referenced; in such instances, only the most comprehensive or recent analysis was retained. See Table S3.

Data extraction and quality assessment

Two independent investigators (S.W. and W.L.) extracted data using a standardized form. Data were extracted on study characteristics (authors, publication year, country, and design), patient demographics (maternal age and gestational age), and clinical outcomes (Table 1). Additional investigators (X.Z. and S.H.) reviewed data for accuracy. Gestational age was stratified according to the American College of Obstetricians and Gynecologists (ACOG) guidelines: first trimester as <14 weeks, second trimester as 14–27 weeks, and third trimester as ≥28 weeks³¹. For the trimester-specific analysis, data on ‘composite adverse fetal outcomes’ (spontaneous abortion, intrauterine fetal death, and preterm delivery) were extracted to account for the varying risks associated with gestational age. Methodological quality was critically appraised using the Newcastle–Ottawa scale (NOS) for observational studies, with scores of seven or more indicating high quality³². The single RCT was assessed using the Cochrane Risk of Bias 2 (RoB 2) tool³³. The certainty of the evidence was evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach³⁴.

Table 1.

Baseline characteristics

Study	Publication year	Country	Design	Pathology	Total number of pregnant women	Number of pregnant women who underwent laparoscopy	Number of pregnant women who underwent open surgery	Gestational age (weeks), laparoscopy/open surgery	Maternal age (years ), laparoscopy/open surgery
Sugai et al.5	2025	Japan	Retrospective cohort (database)	Acute appendicitis	1624	599	1025	19.0/22.0	30.0/31.0
Zhang et al.11	2024	China	Retrospective cohort	Adnexal masses	155	102	53	10.1/24.4	27.9/26.6
*Lee et al.6	2022	Korea	Retrospective cohort (database)	Mixed	2941	1504	1437	NR	38.0/40.0
Kozan et al.12	2020	Turkey	Retrospective cohort	Acute appendicitis	21	14	7	20.1 (mean)	29.1 (mean)
Rios-Diaz et al.13	2020	USA	Retrospective cohort (database)	Gallstone disease	6390	2440	3950	32.8/35.0	26.8/27.1
Tumati et al.14	2020	USA	Retrospective cohort (database)	Acute appendicitis	1006	547	459	NR	27.0/27.0
Gök et al.15	2018	Turkey	Retrospective cohort	Acute appendicitis	57	18	39	15.0/17.0	29.0/27.0
Karaman et al.16	2016	Turkey	Multicentre retrospective cohort	Acute appendicitis	48	12	36	22.4/25.7	27.1/28.8
Winter et al.17	2016	Australia	Multicentre retrospective cohort	Acute appendicitis	218	125	93	13.0/22.0	27.0/28.0
Cheng et al.18	2015	Taiwan	Retrospective cohort (database)	Acute appendicitis	781	128	653	NR	27.0/27.0
Cox et al.19	2015	USA	Retrospective cohort (database)	Acute appendicitis	1335	894	441	NR	27.7/28.2
Chen et al.20	2014	China	RCT	Adnexal masses	69	33	36	16.8/15.9	27.5/26.8
Peled et al.21	2014	Israel	Retrospective cohort	Acute appendicitis	85	26	59	14.6/19.3	29.2/27.6
Eom et al.22	2012	Korea	Retrospective cohort	Acute appendicitis	43	15	28	15.0/17.0	27.5/30.0
Koo et al.23	2012	Korea	Retrospective cohort	Adnexal masses	262	88	174	11.6/15.1	30.1/29.4
Sadot et al.24	2010	USA	Retrospective cohort	Acute appendicitis	65	48	17	18.1/24.3	29.8/28.8
Kirshtein et al.25	2009	Israel	Retrospective cohort	Acute appendicitis	42	23	19	12.0/16.2	29.8/26.8
Kuy et al.26	2009	USA	Retrospective cohort (database)	Gallstone disease	9714	8645	1069	NR	26.2/26.1
McGory et al.27	2007	USA	Retrospective cohort (database)	Acute appendicitis	3133	454	2679	NR	27.2 (mean)
Barone et al.28	1999	USA	Retrospective cohort (registry)	Gallstone disease	46	20	26	18.4/23.7	25.4/24.8
Cosenza et al.29	1999	USA	Retrospective cohort	Gallstone disease	32	12	20	20.5/21.0	29.0 (mean)
†Soriano et al.30	1999	Israel	Retrospective cohort	Adnexal masses	93	39	54	8.3/14.3	28.3/26.9

Values are mean or median unless otherwise indicated. * Lee et al.⁶ included both appendectomy and adnexal mass resection; they only reported the overall mean/median for the entire cohort, not stratified by surgical approach. † Soriano et al.³⁰ reported data for first-trimester laparoscopy versus first-trimester open surgery versus second-trimester open surgery; values are weighted averages for the open surgery group. T1–T3, trimester 1 to trimester 3; NR, not reported (data not available in the original article, typically in large administrative database studies like the National Inpatient Sample (NIS) or the Statewide Planning and Research Cooperative System (SPARCS).

Statistical analysis

Meta-analysis was performed using R software (version 4.3.0; R Foundation for Statistical Computing, Vienna, Austria) with the meta and forestploter packages³⁵. For dichotomous outcomes (fetal loss, preterm delivery, and maternal complications), ORs with 95% confidence intervals were calculated. For continuous outcomes (length of hospital stay), mean differences (MDs) were utilized. Given the anticipated clinical and methodological diversity among studies, a random-effects model (DerSimonian-Laird method) was employed for all analyses to provide conservative estimates³⁶. To address zero-event studies, the Haldane-Anscombe correction was applied (adding 0.5 to each cell) to facilitate OR calculation; however, ‘double-zero’ studies (no events in either arm) were excluded from forest plots to prevent statistical artefacts³⁷. Heterogeneity was quantified using the I² statistic and Cochran’s Q test, where I² >50 indicated substantial heterogeneity³⁸. To investigate potential sources of heterogeneity and test the robustness of the findings, the following analyses were conducted: subgroup analyses stratified by underlying pathology (acute appendicitis, gallstone disease, and adnexal masses); leave-one-out sensitivity analyses to assess the influence of individual studies; and meta-regression to examine the impact of publication year on outcomes, thereby assessing whether technological advancements over time influenced safety profiles. Publication bias was assessed visually using funnel plots and quantitatively using Egger’s regression test³⁹. Statistical significance was defined as a two-sided P value <0.050. A subgroup analysis by study design compared effect estimates between the RCT and the cohort studies. A sensitivity analysis was conducted by excluding the RCT from the meta-analysis. These supplementary analyses used the same random-effects model (DerSimonian–Laird method) as the primary analysis.

Results

Search results and study characteristics

The systematic literature search identified 7645 records. After the removal of duplicates and a rigorous screening of titles and abstracts, 22 studies^5,6,11–30 met the strict eligibility criteria and were included in the final meta-analysis. The cohort comprised 1 RCT²⁰ and 21 retrospective cohort studies^{5,6,11–19,21–30}, 8 of which were large-scale population-based analyses derived from national or statewide databases. The methodological quality of the studies is shown in Table S4 and the GRADE summary of findings (certainty of evidence) is shown in Table S5.

The total pooled population included 28 160 pregnant women (15 786 of whom underwent laparoscopic surgery and 12 374 of whom underwent open surgery). The underlying pathologies included acute appendicitis (13 studies)^{5,12,14–19,21,22,24,25,27}, adnexal masses (4 studies)^11,20,23,30, gallstone disease (4 studies)^13,26,28,29, and mixed pathologies (1 study)⁶. The PRISMA flow diagram details the selection process (Fig. S1).

Fetal loss

Data on fetal loss—defined as spontaneous abortion, miscarriage, or intrauterine fetal demise—were extracted from 13 studies involving 5839 patients. The pooled analysis revealed that the laparoscopic approach was associated with a statistically higher risk of fetal loss compared with open surgery (OR 2.02 (95% c.i. 1.40 to 2.92); P < 0.001). Given the preferential use of laparoscopy during the first trimester compared with open surgery, and to strictly evaluate whether the observed risk was solely an artefact of gestational age (selection bias), a subgroup analysis restricted to the first trimester was performed. Results indicated that the laparoscopic approach maintained a trend toward higher fetal loss (OR 1.35 (95% c.i. 0.84 to 2.19); Fig. S2). When stratified by pathology, this elevated risk remained consistent across subgroups. In patients with acute appendicitis, laparoscopy was associated with significantly higher odds of fetal loss (OR 2.10 (95% c.i. 1.31 to 3.37)). Similarly, the adnexal masses subgroup demonstrated an elevated risk (OR 1.99 (95% c.i. 0.87 to 4.54)). In the gallstone disease subgroup, while the point estimate favoured open surgery, the difference was not statistically significant (OR 0.96 (95% c.i. 0.13 to 7.32)). See Fig. 1.

Figure 1.

Forest plot of fetal loss comparing laparoscopic versus open abdominal surgery during pregnancy stratified by pathology

The forest plot displays the ORs and 95% confidence intervals for fetal loss derived from a random-effects model. The size of each square is proportional to the weight of the individual study in the meta-analysis. The diamonds represent the pooled ORs for each subgroup and the overall population. The vertical dashed line represents the line of no effect (OR = 1.00). An OR >1.00 indicates a higher risk of fetal loss in the laparoscopic group, favouring open surgery.

Preterm delivery

The analysis of preterm delivery, based on data from 12 studies (16 031 pregnant women), favoured the laparoscopic approach. Laparoscopy was associated with a significantly reduced risk of preterm delivery compared with laparotomy (OR 0.56 (95% c.i. 0.34 to 0.94); P = 0.020). However, this finding was graded as low certainty evidence due to substantial heterogeneity (I² = 91%). Subgroup analysis revealed distinct patterns across pathologies. The protective effect of laparoscopy was most pronounced and statistically significant in the gallstone disease subgroup (OR 0.41 (95% c.i. 0.33 to 0.51)). For acute appendicitis and adnexal masses, trends favoured laparoscopy (OR 0.69 and OR 0.65 respectively), although this was not statistically significant. See Fig. 2.

Figure 2.

Forest plot of preterm delivery comparing laparoscopic versus open abdominal surgery during pregnancy stratified by pathology

Squares represent the ORs for individual studies, with the size of the square proportional to the study’s weight. Horizontal lines indicate the 95% confidence intervals. Diamonds represent the pooled ORs for each subgroup and overall. Data were analysed using a random-effects model.

To assess the potential influence of study design, the single RCT (Chen et al.²⁰, 2014) was compared with the 21 cohort studies. For preterm delivery, the pooled OR from the 11 cohort studies that reported this outcome was 0.56 (95% c.i. 0.33 to 0.94; I² = 91%), while the RCT alone gave an OR of 1.07 (95% c.i. 0.07 to 17.9). There was no statistically significant interaction between study design and the effect estimate (P for interaction = 0.680). For fetal loss, the RCT contributed no events in either arm and therefore did not influence the pooled estimate. A sensitivity analysis that completely excluded the RCT yielded virtually identical results. The OR for preterm delivery changed from 0.56 (95% c.i. 0.34 to 0.94) to 0.56 (95% c.i. 0.33 to 0.94), and the OR for fetal loss remained unchanged at 2.02 (95% c.i. 1.40 to 2.92). These supplementary analyses confirm that the inclusion of this small RCT does not materially alter any of the primary conclusions. See Fig. S3.

Trimester-specific adverse outcomes

To elucidate the impact of gestational age on surgical safety, a stratified analysis of composite adverse fetal outcomes was performed based on the trimester of surgery (Fig. 3). In the first trimester, no significant difference in adverse outcomes was observed between laparoscopic and open surgery (OR 0.96 (95% c.i. 0.50 to 1.83); P = 0.900). However, during the second trimester—classically considered the optimal time for non-obstetric surgery—laparoscopy was associated with a significantly elevated risk of composite adverse outcomes compared with open surgery (OR 2.35 (95% c.i. 1.15 to 4.77); P = 0.020). A similar trend toward higher risk with laparoscopy persisted in the third trimester (OR 1.62 (95% c.i. 0.54 to 4.31); P = 0.390), with heterogeneity driven by a large database study⁵ that independently reported a significantly higher risk in this subgroup.

Figure 3.

Forest plot of composite adverse fetal outcomes comparing laparoscopic versus open abdominal surgery during pregnancy stratified by trimester of pregnancy

Composite adverse fetal outcomes is defined as the aggregate of spontaneous abortion, intrauterine fetal death, and preterm delivery occurring after the procedure. The forest plot displays the ORs and 95% confidence intervals for composite adverse obstetric outcomes associated with laparoscopic surgery compared with open surgery. Studies were stratified by gestational age at the time of surgery. Trimesters were defined in accordance with the ACOG guidelines: first trimester, <14 weeks; second trimester, 14–27 weeks; and third trimester, ≥28 weeks. ACOG, American College of Obstetricians and Gynecologists.

Maternal surgical outcomes and resource utilization

Regarding maternal morbidity, the laparoscopic approach demonstrated a definitive advantage. Analysis of nine studies (17 785 pregnant women) showed that women who underwent laparoscopy experienced a 55% reduction in the odds of maternal postoperative complications—such as wound infection, ileus, and thromboembolism—compared with those who underwent laparotomy (OR 0.45 (95% c.i. 0.30 to 0.68); P < 0.001; Fig. 4). Furthermore, resource utilization analysis from 10 studies (19 011 pregnant women) confirmed that laparoscopy was associated with a significantly shorter length of hospital stay, with a mean reduction of 1.62 days (95% c.i. −2.04 to −1.19 days; P < 0.001; GRADE: moderate certainty; Fig. 5).

Figure 4.

Forest plot of maternal postoperative complications comparing laparoscopic versus open abdominal surgery during pregnancy stratified by pathology

Maternal complications include surgical-site infection, wound dehiscence, postoperative ileus, venous thromboembolism, pneumonia, urinary tract infection, and readmission within 30 days related to surgery. Squares indicate the ORs for each study, with the size proportional to the study’s weight in the random-effects model. Horizontal lines represent 95% confidence intervals. Diamonds represent pooled ORs. An OR <1.00 favours the laparoscopic approach.

Figure 5.

Forest plot of length of hospital stay comparing laparoscopic versus open abdominal surgery during pregnancy stratified by pathology

Length of hospital stay is measured in days from the day of surgery to discharge. Squares indicate the MDs between the laparoscopic and open groups. Horizontal lines represent 95% confidence intervals. A negative MD indicates a shorter length of hospital stay for the laparoscopic group. Data were analysed using a random-effects model due to high heterogeneity (I²). MD, mean difference.

Sensitivity analysis and publication bias

To validate the robustness of these findings, sensitivity analyses and meta-regression were performed (presented in Figs S4–S9). Leave-one-out sensitivity analyses demonstrated that the pooled effect estimates for fetal loss and preterm delivery were not driven by any single study (Figs S4, S7). Meta-regression analysis was conducted to assess whether the safety profile of laparoscopy has improved over time due to technological advancements. The analysis revealed no significant association between publication year and the risk of fetal loss (P = 0.720; Fig. S5) or preterm delivery (P = 0.450; Fig. S8), suggesting that the relative risks observed are inherent to the surgical approaches rather than a reflection of the learning curve or historical era. Finally, assessment of publication bias via Egger’s regression tests revealed no significant asymmetry for either fetal loss (P = 0.312; Fig. S6) or preterm delivery (P = 0.415; Fig. S9), supporting the integrity of the synthesized evidence.

Discussion

This systematic review and meta-analysis, encompassing over 28 000 pregnant women, shows that although laparoscopic surgery reduces maternal complications and the risk of preterm delivery, it is associated with a higher rate of fetal loss compared with open surgery. This challenges the simplified view that laparoscopy is universally superior and suggests that surgical decision-making must be tailored to the specific gestational age and pathology.

A worrisome finding of this study is the association between laparoscopy and increased fetal loss. Historically, similar findings were interpreted as selection bias—arguing that laparoscopy is preferentially utilized in the first trimester, an interval naturally carrying a high background rate of spontaneous abortion^40,41. Open surgery was more frequently performed in the second and third trimesters¹⁷, where the natural risk of fetal loss is significantly lower⁴¹. The present study challenges this interpretation. Even when the analysis was restricted to the first trimester to neutralize the confounder of gestational age, the increased fetal loss in the laparoscopic group remained. This suggests that the elevated risk may not be an artefact but possibly attributable to intrinsic physiological stressors of the laparoscopic procedure itself. Therefore, the potential for intrinsic fetal compromise mandates heightened caution and rigorous perioperative monitoring during early-trimester laparoscopy.

For pregnancies that have progressed beyond the first trimester, the primary threat to fetal survival shifts from miscarriage to preterm delivery. The present data unequivocally support laparoscopy. The minimally invasive approach reduced the risk of preterm delivery by 44% (OR 0.56), a benefit most evident in the gallstone disease subgroup. This protective effect is biologically plausible. Open surgery often requires large incisions and manual retraction of the gravid uterus to expose the surgical field, which can irritate the uterus and trigger contractions⁴². Laparoscopy avoids this direct manipulation, thereby maintaining uterine quiescence⁴³. For patients in the late second or third trimester—where the fetus is viable but prematurity poses severe risks—laparoscopy appears to be the superior strategy for prolonging pregnancy.

A novel and important finding of this study is the trimester-specific risk profile. While the second trimester is traditionally considered the ‘safe harbour’ for non-obstetric surgery, the present analysis revealed an increase in composite adverse outcomes for laparoscopy. This finding, supported by a recent study⁵, suggests that, as the uterus enlarges, the technical difficulty of laparoscopy increases. The limited working space and the physiological effects of pneumoperitoneum may pose greater risks than previously appreciated as pregnancy advances⁸. This challenges the current SAGES guidelines that support laparoscopy ‘regardless of trimester’³. These data suggest that, for complex cases in the late second or third trimester, surgeons should not hesitate to convert to or select open surgery if technical difficulties arise, as prolonged laparoscopic manipulation may be detrimental.

From a public health perspective, the widespread adoption of laparoscopy is justified by the reductions in maternal morbidity and length of hospital stay, which translate into economic benefits and accelerated maternal recovery¹³. However, clinical guidelines should reflect a more nuanced, risk-stratified approach. While laparoscopy remains feasible in the first trimester, comprehensive patient counselling regarding the baseline risk of miscarriage is essential. On the other hand, in the second and third trimesters, the laparoscopic approach is preferred for mitigating the risk of preterm delivery, albeit necessitating heightened technical caution as the uterus enlarges. Furthermore, in clinical scenarios involving complex pathology—such as perforated appendicitis or advanced pregnancy where visualization is compromised—open surgery remains a safe and valid alternative, advocating for a tailored surgical strategy that prioritizes maternal and fetal safety over the mode of access.

This study has several limitations. Causal inference is inherently limited. Of the 22 studies, 21 of them are retrospective cohort studies and residual confounding cannot be ruled out despite adjustment for possible confounders. In particular, indication bias may play a role: open surgery could have been used more often in patients could have been more severe disease (for example perforated appendicitis, generalized peritonitis, or large adnexal masses), whereas laparoscopy was more often chosen for milder, uncomplicated conditions. This would underestimate the true risk of open surgery and may explain the higher fetal loss rate observed with laparoscopy. Without individual patient data, quantitative adjustment for disease severity markers (for example white blood cell count, C-reactive protein levels, or intraoperative findings) was not possible. Second, the high heterogeneity (I² = 91%) for preterm delivery indicates that the pooled estimate is not a single, transferable number. The heterogeneity likely arises from variations in tocolytic protocols, thresholds for operative intervention, gestational age distribution across studies, and differences in healthcare systems (for example routine versus selective fetal monitoring). A random-effects model accounts for this but it does not resolve the clinical heterogeneity. Therefore, the protective effect of laparoscopy against preterm delivery may not apply uniformly across all settings or patient subgroups. Third, outcome definitions and reporting were inconsistent across studies. Some studies defined fetal loss as spontaneous abortion before 20 weeks and other studies defined fetal loss as intrauterine death at any gestational age; similarly, preterm delivery thresholds varied (some studies used <37 weeks and other studies used <34 weeks). This introduces non-differential misclassification and may drive effect estimates toward the null and may have attenuated true differences between surgical approaches. Fourth, publication bias remains a concern despite non-significant results from Egger’s regression tests. In observational studies, negative or null results are less likely to be published, particularly for rare outcomes such as fetal loss during pregnancy. The funnel plots showed no significant asymmetry but the power of Egger’s test is low when the number of studies is small (for example, for fetal loss, only 13 studies contributed). Finally, the single small RCT contributed minimal weight (3.2% for preterm delivery) and zero events for fetal loss. While its inclusion does not alter the pooled estimates, its limited sample size and single-centre design preclude any meaningful comparative inference. The evidence base therefore remains dominated by observational data. Despite these limitations, the consistency of the findings across sensitivity analyses supports their robustness.

To address these limitations, an international prospective registry could be established with standardized collection of trimester-specific surgical details, tocolysis, fetal monitoring, and outcomes. Second, a meta-analysis of individual patient data from existing high-quality cohort studies would allow granular adjustment for confounders such as perforation or peritonitis.

In summary, this study shows that laparoscopy confers definitive maternal benefits and mitigates the risk of preterm delivery; however, its association with increased fetal loss and the paradoxical rise in adverse outcomes during the second trimester expose critical limitations in current ‘all-trimester’ safety guidelines. Surgical strategy should therefore shift from a default minimally invasive approach to a risk-stratified model, maintaining open intervention as a prudent safeguard for fetal viability in cases of advanced gestation or complex pathology.

Funding

The authors have no funding to declare.

Author contributions

Siran Wan (Conceptualization (lead), Investigation (lead), Writing—original draft (lead)), Wei Lin (Investigation (supporting), Writing—original draft (supporting)), Xue Zhang (Data curation (supporting), Validation (supporting)), Saidi Hu (Data curation (equal), Validation (supporting)), Chenchen Luo (Formal analysis (equal), Visualization (supporting)), Shunhong Zhang (Formal analysis (supporting), Visualization (equal)), Elaine dos Ramos Amado dos Ramos (Writing—review & editing (supporting)), Lucilia Fernandes de Almeida (Writing—review & editing (supporting)), and Lin-yong Zhao (Conceptualization (supporting), Methodology (lead), Project administration (lead), Supervision (lead), Writing—review & editing (lead)). All authors reviewed and approved the final version of the manuscript.

Disclosure

The authors declare no conflict of interest.

Supplementary material

Supplementary material is available at BJS online.

Data availability

The data supporting the findings of this meta-analysis are available from the corresponding author upon reasonable request.