Objective
Accurate preoperative counseling about whether an endometriotic cystectomy has a detrimental effect on the ovarian reserve has been a considerable challenge, because studies assessing the postoperative antral follicle counts and anti-Müllerian hormone levels have reported conflicting results. Our objective was to explore the impact of endometriotic cystectomy on both the anti-Müllerian hormone levels and antral follicle counts, with focus on prospective studies in which both variables were measured for each woman concurrently (overcoming unmeasured confounding), in the same setting (overcoming surgical technique differences), and at the same 3 postoperative time points, namely early (1–6 weeks), intermediate (2–6 months) and late (9–18 months), to overcome time-sensitive changes.
Data Sources
Databases of PubMed, ClinicalTrials.gov , the Cochrane Library, Web of Science, and EBSCO were searched between January 2000 and October 2020.
Study Eligibility Criteria
Only prospective cohort studies that evaluated the impact of endometriotic stripping cystectomy on anti-Müllerian hormone levels and antral follicle counts in the same women, at matching time points, and in the same setting were eligible.
Study Appraisal and Synthesis Methods
Two authors performed the screening and data extraction independently.
Results
A total of 14 prospectively designed studies were eligible for the meta-analysis and included 650 women. The included studies had a low risk of bias. The postoperative weighted mean differences in serum anti-Müllerian hormone levels dropped significantly when compared with the preoperative levels by an estimated 1.77 ng/mL (95% confidence interval, 0.77–2.77; P <.001), 1.17 ng/mL (95% confidence interval, 0.66–1.67; P <.001), and 2.13 ng/mL (95% confidence interval, 1.61–2.65; P <.001) at the early (1–6 weeks), intermediate (2–6 months), and late (9–18 months) time points, respectively. This corresponded to a mean reduction in serum anti-Müllerian hormone levels at each of the 3-time points of 44.4%, 35.1%, and 54.2%, respectively. Conversely, the postoperative weighted mean difference in the antral follicle count estimates did not change significantly at any of the 3 time points; the early antral follicle count was 0.70 (95% confidence interval, −2.71 to 3.56; P =.63), the intermediate count was −0.94 (95% confidence interval, −2.53 to 0.65; P =.25), and the late count was 2.58 (95% confidence interval, −0.43 to 5.58; P =.09). Overall, high levels of heterogeneity were encountered (I 2 ranging between 92% and 94% for the anti-Müllerian hormone levels and between 94% and 98% for the antral follicle counts at the 3 time points), which were attenuated when similar anti-Müllerian hormone assays were compared, and the meta-regression suggested that age did not contribute to heterogeneity.
Conclusion
Endometriotic cystectomies are associated with a significant reduction in the serum anti-Müllerian hormone levels but not in the antral follicle counts, with the detrimental effects on the anti-Müllerian hormone levels consistently detectable at the early-, intermediate-, and late-postoperative time points. In women with endometrioma, the anti-Müllerian hormone level may provide a more accurate assessment of the risk for iatrogenic depletion of the ovarian reserve.
Why was this study conducted?
Accurate preoperative counseling about whether an endometriotic cystectomy has a detrimental effect on the ovarian reserve has been a considerable challenge, because studies assessing the postoperative antral follicle counts (AFCs) and anti-Müllerian hormone (AMH) levels have reported conflicting results. To address important measured and unmeasured confounding factors, a systematic review of prospective studies was undertaken, in which parallel repeat measurements of both AMH and AFC were conducted for the same women, at the same time points, and in the same setting.
Key findings
Endometriotic stripping cystectomies are associated with a substantial reduction in the serum AMH levels but not in the AFCs, with detrimental effects on the AMH levels being consistently detected at the early-, intermediate-, and late-postoperative time points.
What does this add to what is known?
In women with endometrioma, AMH levels may be of greater utility than AFCs in the assessment of the risk for iatrogenic depletion of the ovarian reserve.
Introduction
Endometriosis is a common and chronic disease that affects 5% to 10% of women of reproductive age worldwide, , however, its prevalence may reach 50% among women with infertility. Endometriosis typically manifests with debilitating pain or subfertility, and, as such, affected women may present to gynecologic (endoscopic) surgeons or reproductive medicine specialists. In the clinical setting, endometriosis may present in isolation or as a combination of 3 distinct principal forms, namely superficial peritoneal lesions, endometrioma, and deep infiltrating disease. Endometrioma is considered to be the most pathognomonic feature of endometriosis; it is also the most commonly diagnosed form of the disease, because it is amenable to detection by contemporary ultrasound technology.
Surgery is a well-accepted modality used for the treatment of endometriosis. However, accurate counseling, particularly of those with endometrioma before ovarian cystectomy, is a considerable challenge, with an appraisal of their existing ovarian reserve and the potential impact of surgery being key factors in shared decision making. This challenge arises from several key issues interrelated with the patient’s management, which are sources of active debate between clinicians. These issues include the potential effect of endometrioma per se on the ovarian reserve, , the adverse effect of endometriotic cystectomy on ovarian reserve, , and the suitability of the most reliable ovarian reserve biomarkers, anti-Müllerian hormone (AMH) and antral follicle count (AFC), to accurately assess the postoperative ovarian reserve.
Studies comparing the use of AMH with that of AFC in women with ovarian endometrioma undergoing ovarian cystectomy have demonstrated conflicting results. Specifically, some studies show that serum AMH levels decrease and AFCs remain unchanged following surgery, , whereas others reported the opposite conclusion––AMH levels remained unchanged but AFCs decreased. ,
Two frequently cited systematic reviews and meta-analyses that assessed the impact of endometriotic cystectomy on ovarian reserve reached contradictory conclusions. , One group of investigators, utilizing AMH as the marker of ovarian reserve, suggested that the surgery had a deleterious effect, whereas another group using AFC as a marker found no effect. The meta-analysis assessing the AMH levels included 8 prospective studies and 237 patients and included women who underwent excisional surgery, with 1 additional study including women who underwent either a cystectomy or an oophorectomy. In contrast, the meta-analysis of the AFCs included 13 (11 prospective and 2 retrospective) studies comprising 597 women and included both excisional and nonexcisional operative techniques. The prospect of drawing conclusions about the true impact of surgery on ovarian reserves from these 2 independent reviews is limited, especially because neither performed parallel repeat AMH and AFC measurements and because there was no standardization regarding the size of the endometrioma, surgical technique, and postoperative time-interval evaluations.
Since these initial meta-analyses, the potential detrimental impact of surgery on ovarian reserves continues to be elucidated, with several histologic studies confirming that an endometriotic cystectomy is generally complicated by inadvertent removal of intact ovarian primordial follicles adjacent to the cyst’s pseudocapsule. , Furthermore, even experienced surgeons employing accurate techniques were unable to escape operative ovarian reserve damage, , which was most likely caused by unavoidable, excessive manipulation of the cortex with subsequent tearing, bleeding, and the need for coagulation. Clarification about the consequences of surgery on the ovarian reserve and which of these 2 biomarkers is more clinically appropriate and sensitive to inform counseling of patients with endometrioma is urgently required. For women considering endometrioma surgery or women with infertility still indecisive about whether to proceed with assisted reproduction or surgical treatment, accurate ascertainment of the potential risks of endometrioma surgery may have substantial implications.
This review and meta-analysis aimed to estimate the impact of an endometriotic stripping cystectomy on the ovarian reserve as determined by repeat, parallel measures of the AMH level and AFC. Specifically, only prospective studies that investigated both the AMH level and AFC before and after endometriotic cystectomy, employed an excision stripping technique, and used matching, timed measures, were analyzed. To assess the short-, medium-, and long-term effects of surgery, repeat measurements were grouped into preoperative and early (1–6 weeks), intermediate (2–6 months), and late (9–18 months) postoperative periods.
Materials and Methods
Search strategy
We performed an extensive electronic database search by employing PubMed, ClinicalTrials.gov , the Cochrane Library Web of Science and EBSCO (as interfaces to databases) to identify research articles published between January 2000 and October 2020, which examined, in parallel and in the same setting, the impact of ovarian endometrioma surgery on AMH levels and AFCs in women of reproductive age. All databases were searched when applying Web of Science and academic search complete when EBSCO was employed. The studies were restricted to those conducted in humans and published in the English language. The following 3 groups of search terms, which were further combined, were used as free text and Medical Subject Heading expressions: endometrioma OR endometriomata OR endometriotic cyst OR endometriosis AND resection OR removal OR remove OR cystectomy OR stripping OR excision AND ovarian reserve OR ovarian response OR antral follicle count OR AFC OR anti-Müllerian hormone OR Müllerian inhibiting factor OR Müllerian inhibiting substance OR AMH.
We employed the Cochrane Handbook for Systematic Reviews of Interventions as a methodologic guide for our analysis. , The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were used as reporting guidelines. All reports relevant to the research question were retrieved and their reference lists manually reviewed to identify further studies. One author (Y.B.S.) performed the search (twice) and 2 authors (N.S. and Y.B.S.) independently screened and reviewed all publications found to be relevant for this study and both selected the studies to be included in the systematic review and meta-analysis. The screening process consisted of 2 stages typical for systematic reviews, namely the title and abstract review stage followed by full-text review. We resolved any disagreements about inclusions or analyses by discussion and consensus or mediation by a third reviewer (J.S.Y.).
Study selection and quality assessment
Prospective studies that evaluated the impact of endometriotic cystectomy on ovarian reserves and in which a stripping cystectomy technique was used in a single-stage procedure in women with 2 intact ovaries were selected. The study design was planned a priori. Cohort, controlled, or randomized studies that fulfilled the inclusion criteria qualified for assessment. For studies that had 2 groups (prospective-controlled or randomized), we only included the group that matched our inclusion criteria. Only studies that assessed both the serum AMH level and AFC in the same woman at a matching time and constant setting were eligible for inclusion.
At least 1 pre- and 1 postoperative time point evaluation was needed for a study to be included. Preoperative AMH and AFC tests were performed at the time of surgery or within a maximum period of 4 weeks before surgery, whereas postoperative examinations were conducted up to 18 months postoperatively. Serum AMH levels were standardized to ng/mL (conversion factor to ng/mL=pmol/L/7.14). AFC at the early follicular measurement was determined by counting the number of antral follicles combined in both operated and nonoperated ovaries. AFC values were standardized to mean±standard deviation (SD), and for studies that reported the median, range, and size of the cohort, standard techniques for conversion for pooled quantitative analyses were used. Only studies that clearly defined the methodology used for serum AMH assay evaluation and AFC value assessment were eligible for investigation.
To reduce heterogeneity and improve data quality, the following exclusions were applied. Retrospective studies were excluded. Studies that included pregnant women were omitted. Studies that examined the impact of surgery on ovarian endometrioma following chemotherapy and/or radiotherapy were omitted to eliminate gonadotoxic effects on the ovarian reserve. Reports that included women who had a history of ovarian surgery, such as an ovarian cystectomy or unilateral oophorectomy, were excluded. Studies that included women with polycystic ovary syndrome or irregular menstrual cycles were omitted, because ovarian reserve test results may differ in these women. Studies that included patients with a history of ovarian failure or premenopausal follicle stimulating hormone levels were excluded. Studies that included women who were taking oral contraceptives or other hormonal therapies at least 3 months before recruitment or postoperatively were omitted to eliminate the possible impact on ovarian reserve test results. Studies reporting other surgical modalities such as ablation technologies, microsurgery or cyst deroofing techniques, or multiple-stage procedures were excluded, because their impact on ovarian reserves may vary widely compared with the commonly practiced stripping procedure. Data presented exclusively as abstracts in scientific meetings were excluded, because of the limited information provided.
We used the Newcastle-Ottawa Scale to assess the quality of the prospective studies eligible for quantitative analysis based on the recommendation of the Cochrane Collaboration. , ,
Outcome measures
The primary outcome was the impact of an endometriotic cystectomy on parallel repeat measurements of AMH and AFC over a short-, medium-, and long-term period. Repeat measurements were grouped into preoperative and early- (1–6 weeks), intermediate- (2–6 months), and late-postoperative (9–18 months) periods. In studies in which more than 1 repeat measurement was performed in the same postoperative period, for example at 12 months and at 18 months (both classed as late), the mean of these values was calculated. The AMH assay and AFC technique used were recorded because both the assay and ultrasound methodology may differ among studies. The size and number of the endometrioma, the revised American Society for Reproductive Medicine (rASRM) classification, surgical techniques, and hemostasis methods were recorded.
Data extraction
We generated descriptive tables for the population and study characteristics for all eligible studies to prepare for the quantitative analysis. For each eligible study, we recorded the first author, publication year, country, study design, sample size, and mean age of participants. The indication for cystectomy, size of endometrioma, number of endometriomata (whether mono- or multicystic), rASRM stage of disease, type of surgery (laparoscopy or laparotomy), the technique of surgery (endometrioma stripping or other), and method of hemostasis (bipolar coagulation or suture) were recorded. Ovarian reserve test methodology (AMH and AFC testing methods), the timing of their assessment, and their results were recorded. When data were missing in the studies included for analysis, we attempted to contact the authors on at least 2 separate occasions.
Data analysis and statistical methods
Continuous measures were extracted as means and SDs and meta-analyzed to produce a weighted mean difference (WMD). The analysis was performed with the open-source software OpenMeta [Analyst] ( www.cebm.brown.edu/openmeta ).
We assessed statistical heterogeneity by Cochran Q, and inconsistency was estimated by using the I 2 statistic. A chi-square value that was larger than its degree of freedom or an I 2 with a value greater than 50% provided evidence for substantial heterogeneity among studies. Anticipating substantial heterogeneity, we used random-effects models.
Results
Search results and excluded studies
Our search found 1222 studies, and among them there were 678 duplicates; 199 were omitted after reading the title and 288 were excluded after reading the abstract ( Figure 1 ). Of the final 57 full-text papers assessed for eligibility, 31 were excluded because they were abstracts, short proceedings in conferences, or planned studies. Of the remaining 26 full-text articles, 12 were excluded from the analysis. Six studies did not include all pre- and postoperative AMH serum levels and AFC values for both ovaries. In 4 of the studies, the methodology for the AMH and/or AFC tests were not self-evident or adequately and explicitly clarified. , In 1 study, a discrepancy was noted in the AMH values between the manuscript itself and its online supplements, and in the AFC values between the tables and the text. One study was retrospective in design, 1 included women taking hormonal therapy, and another did not specifically report the number of women who underwent endometriotic cystectomy.
The authors of 12 publications were contacted to complete the missing data. One author replied and provided further information.
Included studies
A total of 14 studies were eligible for quantitative synthesis. , , , All eligible studies focused on the impact of an endometrioma and endometriotic cystectomy on serum AMH levels and AFC values, which were measured in parallel for the same women within the same setting and with the same timing for repeat measures. All 14 studies were prospective in design and included an aggregate of 650 women. The weighted mean age±weighted SD of women in the meta-analysis was 28.6±5.0 years. The time intervals of postoperative AMH and AFC evaluations were divided into 3 different time strata, selected based on the data available in the eligible studies. These strata included early-, intermediate-, and late-postoperative periods corresponding to 1 to 6 weeks, 3 to 6 months, and 9 to 18 months, respectively, following surgery.
Of note, 5 of the eligible studies were prospective cohort studies comprising women with an endometrioma who underwent cystectomies. , , , , A further 4 studies involved prospective-controlled reports; in 3 of these, the study groups comprised women with an endometrioma and the control groups included women with benign, nonendometriotic cysts , or a normal ovary. In the fourth study, both groups had endometriotic cystectomy but by different surgical modalities. The remaining 5 studies were randomized clinical trials that compared endometrioma stripping with other alternative surgical modalities. In 1 of the studies, laser vaporization and ablation were employed, cutting and coagulation were used in another, the cyst deroofing method was used in 1 study, another employed microsurgery technique, and 1 employed a 3-step procedure, including gonadotropin-releasing hormone agonist therapy. All groups that employed alternative surgical modalities were excluded from the quantitative assessment.
The descriptive characteristics of the eligible studies, including design, number of women, mean age, indications for surgery, endometriotic cyst diameter, surgical technique, and method of hemostasis, are summarized in Tables 1 and 2 . Selected information regarding the size of the treated endometrioma was available in all included studies, however, the presentation of these data varied ( Table 2 ).
| Author, y | Publication site | Country | Design | Number of patients | Age±SD (y) Study group | Ovarian reserve tests | AMH assay | ||
|---|---|---|---|---|---|---|---|---|---|
| Study group | Unilateral | Bilateral | |||||||
| Tsolakidis et al, 2010 | Fertil Steril | Greece | Prospective RCT | 10 | ND | 32.8±5.3 | AMH, AFC, FSH, LH, E 2 , Inhibin B | DSL | |
| Biacchiardi et al, 2011 | Reprod Biomed Online | Italy | Prospective cohort | 43 | 33 | 10 | 34.2±5.4 | AMH, AFC, FSH, LH, E 2 , Inhibin B | IOT |
| Ercan et al, 2011 | Eur J Obstet Gynecol Reprod Biol | Turkey | Prospective cohort | 36 | 36 | — | 29.4±4.6 | AMH, AFC | DSL |
| Celik et al, 2012 | Fertil Steril | Turkey | Prospective cohort | 65 | 46 | 19 | 28.4±5.7 | AMH, AFC, FSH, LH, E 2 | DSL |
| Uncu et al, 2013 | Hum Reprod | Turkey | Prospective controlled | 30 | 15 | 15 | 29.0±5.4 | AMH, AFC | DSL |
| Urman et al, 2013 | Reprod Biomed Online | Turkey | Prospective cohort | 25 | 25 | — | 32.7±6.1 | AMH, AFC | GEN II |
| Zaitoun et al, 2013 | J Ovarian Res | Egypt | Prospective RCT | 61 | 61 | — | 24.2 ±3.1 | AMH, AFC, FSH | DSL |
| Aboul Gheit et al, 2014 | Middle East Fertil Soc J | Egypt | Prospective controlled | 80 | 64 | 16 | 28.1±4.2 | AMH, AFC | ND |
| Alborzi et al, 2014 | Fertil Steril | Iran | Prospective cohort | 193 | 121 | 72 | 28.4±5.3 | AMH, AFC FSH, E 2 | DSL |
| Jang et al, 2014 | Obstet Gynecol Sci | Korea | Prospective controlled | 12 | 6 | 6 | 27.0±4.30 | AMH, AFC FSH, E 2 | IOT |
| Salihoğlu et al, 2016 | J Minim Invasive Gynecol | Turkey | Prospective controlled | 34 | ND | 18–40 | AMH, AFC FSH, E 2 | ND | |
| Candiani et al, 2018 | Hum Reprod | Italy | Prospective RCT | 30 | 27 | 3 | 30.3±5.2 | AMH, AFC | GEN II |
| Sweed et al, 2019 | J Minim Invasive Gynecol | Egypt | Prospective RCT | 61 | 34 | 27 | 27.1±4.6 | AMH, AFC | Glory Science |
| Rawat et al, 2019 | Int J Clin Obstet Gynaecol | India | Prospective RCT | 10 | ND | ND | 23.5±4.47 | AMH, AFC FSH, LH, E 2 , Inhibin B | IOT |
| Author, y | Indications for surgery | Endometrioma diameter (mm) (mean±SD) | rASRM score | Technique of cyst removal | Hemostasis |
|---|---|---|---|---|---|
| Tsolakidis et al, 2010 |
| At least 30 37.9±15.2 | Stage III—50% Stage IV—50% | Stripping a | Bipolar |
| Biacchiardi et al, 2011 |
| First 36.8±10.8 Second 35.2±8.7 | Stage II—16.3% Stage III—60.5% Stage IV—23.3% | Stripping | Bipolar |
| Ercan et al, 2011 | ND | 52±14 | Stage III—69% Stage IV—31% | Stripping | Bipolar |
| Celik et al, 2012 | ND | At least 30 ≥50 (61.5 %) <50 (38.5 %) 58.7±21.0 | Stage III or IV | Stripping | Bipolar |
| Uncu et al, 2013 |
| 44.4 ±10.2 | ND | Stripping | Bipolar |
| Urman et al, 2013 |
| 51.5±14.5 | ND | Stripping | Bipolar |
| Zaitoun et al, 2013 | ND | ND | ND | Stripping | Bipolar |
| Aboul Gheit et al, 2014 |
| 78±16 | ND | Stripping | Bipolar |
| Alborzi et al, 2014 |
| > 30 (91%) < 30 (9%) | Stage III—36% Stage IV—62% | Stripping | Bipolar±Suture |
| Jang et al, 2014 | ND | ND | ND | Stripping | Bipolar |
| Salihoğlu et al, 2016 | ND | At least 40 | Stage III or IV | Stripping | Bipolar |
| Candiani et al, 2018 |
| ≥30 and ≤80 49.0±15.0 | Stage III or IV | Stripping | ND |
| Sweed et al, 2019 |
| At least 30 52.0±6.0 | Stage III or IV | Stripping | Bipolar |
| Rawat et al, 2019 | ND | ≥30 and ≤80 mean size 84.7 mm 3 | ND | Stripping | ND |
a Stripping is the technique of endometrioma removal in which 2 atraumatic grasping forceps are used to pull the cyst wall and the normal ovarian parenchyma in opposite directions, thus developing the cleavage plane.
Serum AMH levels and AFC data from eligible studies, before and after surgery, and the length of follow-up in the studies are shown in Table 3 . The timing of postoperative serum AMH level measurements were segregated, on the basis of the available data, into 3 time strata, namely 1 to 6 weeks (early), 2 to 6 months (intermediate), and 9 to 18 months (late) after surgery. Table 4 summarizes the postoperative rate of change of AMH levels and AFCs from baseline at each postoperative checkpoint.
| Author, y | AMH assay | Timing of AMH testing and serum levels (ng/mL) | Timing of AFC testing and values | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Baseline | After surgery | Baseline | After surgery | ||||||
| 1–6 wk | 2–6 mo | 9–18 mo | 1–6 wk | 2–6 mo | 9–18 mo | ||||
| Tsolakidis et al, 2010 | DSL | 3.90±1.26 | — | 2.90±0.63 | — | 2.00±3.16 | — | 2.40±2.53 | — |
| Biacchiardi et al, 2011 | IOT | 3.00±0.40 | — | 1.40±0.20 | 1.30±0.30 | 11.70±5.42 | — | 15.00±5.58 | 13.20±4.53 |
| Ercan et al, 2011 | DSL | 2.03±0.41 | — | 1.95±0.62 | — | 9.70±2.85 | 8.30±3.27 | 10.10±2.76 | — |
| Celik et al, 2012 | DSL | 1.78±1.71 | 1.32±1.29 | 0.72±0.79 | — | 4.90±2.20 | 5.10±2.40 | 6.40±2.20 | — |
| Uncu et al, 2013 | DSL | 2.81±2.15 | 2.07±1.47 | 1.82±1.29 | — | 9.73±4.77 | 11.00±5.37 | 10.40±4.16 | — |
| Urman et al, 2013 | GEN II | 3.42±1.88 | 2.61±1.73 | 2.60±1.73 | — | 8.76±2.34 | 7.80±2.48 | 7.44±2.35 | — |
| Zaitoun et al, 2013 | DSL | 4.50±0.80 | — | 2.40±0.50 | 2.60±0.46 | 6.60±2.30 | — | 3.00±2.50 | 3.80±2.32 |
| Aboul Gheit et al, 2014 | ND | 8.90±4.50 | 3.00±2.10 | — | — | 2.90±1.20 | 6.40±2.08 | — | — |
| Alborzi et al, 2014 | DSL | 3.86±3.58 | 1.66±1.92 | 2.06±2.50 | 1.77±1.76 | 7.81±3.22 | — | 10.75±3.68 | — |
| Jang et al, 2014 | IOT | 5.15±3.32 | — | 2.45±1.77 | — | 10.28±3.77 | — | 11.05±4.61 | — |
| Salihoğlu et al, 2016 | ND | 3.10±1.90 | — | 2.50±1.60 | — | 5.00±2.00 | — | 7.00±2.25 | — |
| Candiani et al, 2018 | GEN II | 2.60±1.40 | — | 1.80±0.80 | — | 11.17±3.60 | — | 15.00±4.80 | — |
| Sweed et al, 2019 | Glory Science | 4.23±0.87 | 1.66±1.02 | — | 1.39±0.76 | 8.90±2.60 | 3.20±1.30 | — | 3.17±1.36 |
| Rawat et al, 2019 | IOT | 4.70±0.94 | 4.27±1.02 | — | — | 8.30±1.16 | 6.60±2.06 | — | — |
| Author, y | Postoperative AMH percentage change (%) | Postoperative AFC percentage change (%) | ||||
|---|---|---|---|---|---|---|
| 1–6 wk | 2–6 mo | 9–18 mo | 1–6 wk | 2–6 mo | 9–18 mo | |
| Tsolakidis et al, 2010 | — | −25.6 | — | — | 20.0 | — |
| Biacchiardi et al, 2011 | — | −53.3 | −56.7 | — | 28.2 | 12.8 |
| Ercan et al, 2011 | — | −3.9 | — | −14.4 | 4.1 | — |
| Celik et al, 2012 | −25.8 | −59.6 | — | 4.1 | 30.6 | — |
| Uncu et al, 2013 | −26.3 | −35.2 | — | 13.1 | 6.9 | — |
| Urman et al, 2013 | −23.7 | −24.0 | — | −11.0 | −15.1 | — |
| Zaitoun et al, 2013 | — | −46.7 | −42.2 | — | −54.5 | −42.4 |
| Aboul Gheit et al, 2014 | −66.3 | — | — | 120.7 | — | — |
| Alborzi et al, 2014 | −57.0 | −46.6 | −54.1 | — | 37.6 | — |
| Jang et al, 2014 | — | −52.4 | — | — | 7.5 | — |
| Salihoğlu et al, 2016 | — | −19.4 | — | — | 40.0 | — |
| Candiani et al, 2018 | — | −30.8 | — | — | 34.3 | — |
| Sweed et al, 2019 | −60.9 | — | −67.3 | −64.0 | — | −64.4 |
| Rawat et al, 2019 | −9.1 | — | — | −20.5 | — | — |
Risk of bias assessment
Quality assessment of the 14 studies that qualified for meta-analysis, employing the Newcastle-Ottawa Scale, is provided in Table 5 . Overall, the quality assessment of these studies showed a low risk of bias. Of the 9 possible stars given to assess the 3 main categories, namely selection, comparability, and outcomes, eligible studies received 7 to 9 stars.
