Scolaris Content Display Scolaris Content Display

卵巢手术用于缓解多囊卵巢综合征女性的症状

Collapse all Expand all

研究背景

多囊卵巢综合征(polycystic ovary syndrome, PCOS)是一种常见的内分泌疾病,约有十分之一的女性受其影响。PCOS由以下三种特征中的两种定义:少排卵或无排卵、临床或生化雄激素过多症或两者兼有、或多囊卵巢。

患有PCOS的女性可能会出现一系列健康问题,包括月经稀发且不规律、多毛和痤疮,以及生育能力低于正常水平。长期的健康问题包括心脏病、糖尿病和子宫癌前病变的风险增加。

研究目的

系统综述旨在评价卵巢手术作为干预措施缓解PCOS患者多毛症、痤疮和月经不调症状的有效性和危害。

检索策略

我们检索了Cochrane妇科和生殖专业注册库(Cochrane Gynaecology and Fertility Group Specialised Register)、CENTRAL、MEDLINE、Embase和PsycINFO(从建库至2016年10月17日)。我们手动检索了引文列表、正在进行的试验和会议记录。

纳入排除标准

我们纳入了关于多囊卵巢女性患者的随机对照试验(randomized controlled trials, RCTs),试验将卵巢钻孔术与未进行治疗、药物治疗或针对PCOS症状的其他形式的手术治疗进行比较 。

资料收集与分析

我们采用了Cochrane推荐的标准方法学程序。主要结局指标是月经规律和PCOS的雄激素症状(多毛症、痤疮)的改善;次要结局指标包括危害、体重指数(body mass index, BMI)变化、腰围、雄激素水平、代谢指标和生活质量。我们使用GRADE方法评价证据的质量。

主要结果

我们纳入了22项随机对照试验(分析了2278名女性)涉及到患有PCOS和痤疮、多毛症或月经周期不规律症状的受试者,所有试验都包括将腹腔镜卵巢钻孔术( laparoscopic ovarian drilling, LOD)作为干预措施。

两项研究报告了其资金来源(Farquhar2002年‐部分由奥克兰医学研究基金会支持;Sarouri2015年‐作者感谢桂兰医科大学研究副校长对本项目的资金支持)。

证据的质量从极低到中等不等。主要局限性是因研究数量少而导致的不精确性、不一致性以及因未对受试者施盲而导致的偏倚风险。评价发表偏倚风险的研究极少。

月经规律

两项研究比较了LOD与二甲双胍(n=226),但无法就月经规律得出结论,因为其研究结果不一致且不适合合并。当LOD与包括二甲双胍+克罗米芬(OR=1.02,95% CI [0.64, 1.64],2项研究,332名女性,I2=13%,低质量证据),来曲唑(OR=1.08,95% CI [0.64, 1.84],1项研究,260名女性,低质量证据),或二甲双胍+来曲唑(OR=0.95,95% CI [0.49, 1.81],1项研究,146名女性,低质量证据)在内的药物治疗相比时,报告月经规律改善的女性比率差异很小或几乎没有差异。然而,一项研究报告称LOD优于促性腺激素(OR=19.2,95% CI [3.17, 116.45],1项研究,35名女性,极低质量证据)。

与单侧LOD相比,进行双侧单极LOD(OR=1.51,95% CI [0.62, 3.71],2项研究,104名女性,I2=0%,中等质量证据),经阴道超声引导LOD(OR=1.23,95% CI [0.64, 2.37],1项研究,147名女性,低质量证据),根据卵巢体积调整热剂量的LOD(OR=0.42,95% CI [0.16, 1.14],1项研究,115名女性,低质量证据)或双极LOD(OR=1.00,95% CI [0.05, 18.57],1项研究,18名女性,低质量证据)的女性报告月经规律改善的比率差异很小或几乎没有差异。

与穿刺两次及以下相比,每个卵巢穿刺四到五次可能会提高女性报告月经规律的比率(OR=16.04,95% CI [4.19, 61.34],2项研究,73名女性,I2=0%,低质量证据)。

雄激素症状

LOD与二甲双胍(OR=1.00,95% CI [0.42, 2.37],1项研究,126名女性,中等质量证据)或促性腺激素相比,雄性激素症状改善差异很小或几乎没有差异;痤疮(OR=3.20,95% CI [0.33, 30.94],1项研究,25名女性,低质量证据),多毛症(OR=2.31,95% CI [0.22, 23.89],1项研究,25名女性,低质量证据)。

LOD与经阴道超声引导下的LOD相比,雄激素症状改善差异很小或几乎没有差异,如多毛症(OR=1.09,95% CI [0.30, 3.91],1项研究,39名女性,低质量证据)或痤疮(OR=0.84,95% CI [0.20, 3.50],1项研究,31名女性,低质量证据)。

伤害

与二甲双胍联合克罗米芬相比,LOD产生较少的胃肠道副作用(OR=0.05,95% CI [0.01, 0.36],2项研究,332名女性,I2=0%,中等质量证据)。一项研究表明,LOD和促性腺激素之间的卵巢过度刺激综合征发生率差异很小或几乎没有差异(OR=0.08,95% CI [0.00, 1.61],1项研究,33名女性,低质量证据)。

与LOD相比,经阴道注水式腹腔镜检查的粘连更少(OR=0.10,95% CI [0.05, 0.18],1项研究,246名女性,中等质量证据)。将可变能量LOD与标准LOD进行比较时,粘连方面的差异很小或几乎没有差异(OR=0.96,95% CI [0.32, 2.88],1项研究,64名女性,低质量证据)。另一项研究(44名女性)报告称,在传统或单侧LOD术后再次手术的女性中,均未发现粘连。

作者结论

与纳入研究中使用的大多数药物治疗相比,没有明确的证据表明LOD可以改善月经规律或PCOS的雄激素症状。与二甲双胍和克罗米芬相比,LOD产生较少的胃肠道副作用。

同样没有明确的证据表明不同类型的LOD有不同的疗效,但每个卵巢穿刺4到5次的LOD可能比穿刺2次或更少的LOD更有效。几乎没有证据将LOD与不同类型的手术进行比较,只有一项研究得出经阴道注水式腹腔镜检查比LOD有更低的粘连风险。

一项小型研究的证据表明,与促性腺激素相比,LOD对月经调节有益。然而,促性腺激素很少用于这一适应症。

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

卵巢手术用于缓解多囊卵巢综合征女性的症状

系统综述问题

Cochrane研究人员回顾了有关卵巢手术对多囊卵巢综合征(polycystic ovary syndrome, PCOS)症状影响的证据。我们发现22项将其与手术和非手术治疗以及手术技术的变化进行比较的研究。测量的主要结局是月经规律改善,多毛和痤疮(雄性激素症状)的减少。我们还关注了治疗带来的危害、体重变化、睾酮水平变化、代谢指标和生活质量的变化。

系统综述背景

患有PCOS的女性可能会出现一系列健康问题,包括月经稀发且不规律、多毛和痤疮,以及生育能力低于正常水平。长期的健康问题包括心脏病、糖尿病和子宫癌前病变的风险增加。

目前的大多数研究都着眼于卵巢手术对提高PCOS女性生育能力的影响。我们的系统综述旨在研究腹腔镜(钻孔)卵巢手术(laparoscopic ovarian surgery, LOD)对于改善PCOS其他症状的影响。

检索日期

证据检索截至2016年10月。

研究特征

我们纳入了22项随机对照试验(randomised controlled trials, RCTs),涉及2278名女性。随机对照试验是一种医学实验,受试者在研究中将随机接受一种或其他不同的治疗。受试者是来自世界各地不同环境的PCOS患者。

二十二项RCT中有十项比较了LOD与药物治疗。这些治疗包括二甲双胍、克罗米芬、促性腺激素、来曲唑和罗格列酮。二十二项研究中有十项比较了传统LOD与手术技术的变化。二十二项RCT中有两项关注了在LOD期间使用不同的能量水平或卵巢钻孔数量。

研究资金来源

两项研究报告了其资金来源(Farquhar2002年‐部分由奥克兰医学研究基金会支持;Sarouri2015年‐作者感谢桂兰医科大学研究副校长对本项目的资金支持)。

主要研究结果

LOD可能比促性腺激素更能调节月经周期。然而,大多数医生会考虑其他一线治疗方案。在多囊卵巢综合征患者中,每个卵巢穿刺四或五个钻孔比穿刺两个或更少的钻孔在月经调节方面更有效。

没有足够的证据来说明LOD与其他医学治疗或手术技术的变化在改善月经规律或雄激素症状方面是否存在差异。

与二甲双胍和克罗米芬相比,LOD产生较少的胃肠道副作用,但涉及手术并且不是月经紊乱或多毛症的标准治疗方法。与LOD相比,经阴道注水式腹腔镜检查的瘢痕组织较少。

总体来说,LOD可以被认为具有较低的危害风险,并且可作为PCOS症状管理的一种选择。

证据质量

证据的质量从极低到中等不等。主要局限性是因研究数量少而导致的不精确性、不一致性以及因未对受试者施盲(向其隐瞒治疗类型)而导致的偏倚风险。评价发表偏倚风险的研究极少。

Authors' conclusions

available in

Implications for practice

There was no clear evidence that LOD improves menstrual regularity or the androgenic symptoms of PCOS, compared to most of the medical treatments used in the included studies. LOD was associated with fewer gastrointestinal side effects compared to metformin and clomiphene.

There was also no clear evidence of different effectiveness between types of LOD, except that LOD with four to five punctures per ovary may be more effective than with two or fewer punctures. There was little evidence comparing LOD with different types of surgery, although one study concluded that transvaginal hydrolaparoscopy had a lower risk of adhesions than LOD.

There was evidence from one small study of benefit from LOD compared to gonadotrophins in terms of menstrual regulation. However, gonadotrophins are seldom used for this indication.

Implications for research

Studies are required to compare LOD to placebo or to no treatment in order to clarify the impact of LOD alone on the management of symptoms of PCOS. Our conclusions on LOD versus gonadotrophins are based on the results of one small RCT and correlation with further data would be beneficial. Most studies enrolled women for fertility treatment, and primary RCT evidence of surgical treatments of symptoms of PCOS would enhance the body of evidence. Implications of both medical and surgical treatments were generally analyzed in the short and medium term. Long‐term, prospective follow‐up of women with treatments for PCOS would add to our knowledge of the relative effects and harms involved.

Summary of findings

Open in table viewer
Summary of findings for the main comparison. LOD compared to medical interventions for symptom relief in women with polycystic ovary syndrome

LOD compared to medical interventions for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: Laparoscopic ovarian drilling (LOD)
Comparison: medical interventions

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD

With LOD

Difference

Menstrual regularity at 6 months

LOD vs metformin
N of participants: 236 (2 RCTs)

LOD vs metformin + clomiphene
N of participants: 332 (2 RCTs)

LOD vs gonadotropins
N of participants: 35 (1 RCT)

LOD vs letrozole
N of participants: 260 (1 RCT)

LOD vs metformin + letrozole
N of participants: 146 (1 RCT)

Findings inconsistent and data unsuitable for pooling

Not calculable

⊕⊝⊝⊝
VERY LOW 1, 2, 3

OR 1.02
(0.64 to 1.64)

70.6%

71.0%
(60.5 to 79.7)

0.4% more
(10 fewer to 9.2 more)

⊕⊕⊝⊝
LOW 1, 3

OR 19.20
(3.17 to 116.45)

11.1%

70.6%
(28.4 to 93.6)

59.5% more
(17.3 more to 82.5 more)

⊕⊕⊝⊝
LOW 1, 3

OR 1.08
(0.64 to 1.84)

68.8%

70.4%
(58.5 to 80.2)

1.6% more
(10.3 fewer to 11.4 more)

⊕⊕⊝⊝
LOW 1, 3

OR 0.95
(0.49 to 1.81)

52.1%

50.8%
(34.7 to 66.3)

1.3% fewer
(17.3 fewer to 14.2 more)

⊕⊕⊝⊝
LOW 1, 3

Improvement in androgenic symptoms at 6 months (hirsutism/acne) ‐ LOD vs metformin
N of participants: 126
(1 RCT)

OR 1.00
(0.42 to 2.37)

79.4%

79.4%
(61.8 to 90.1)

0.0% fewer
(17.6 fewer to 10.7 more)

⊕⊕⊝⊝
LOW 1, 3

Improvement in androgenic symptoms at 6 months (hirsutism/acne) ‐ LOD vs gonadotrophins
N of participants: 50 (1 RCT)

Acne: OR 3.20 (0.33 to 30.94)

Hirsutism: OR 2.31, (0.22 to 23.89)

See comments

⊕⊝⊝⊝VERY LOW 1, 4

Acne: 4/29 without LOD, 1/21 with LOD

Hirsutism: 3/29 without LOD, 1/21 with LOD

Harms: GI Upset at 6 months ‐ LOD vs metformin + clomiphene
N of participants: 332 (2 RCTs)

Harms: OHSS rates at 6 months ‐ LOD vs gonadotrophins
N of participants: 33 (1 RCT)

OR 0.05
(0.01 to 0.36)

10.4%

0.6%
(0.1 to 4.0)

9.9% fewer
(10.3 fewer to 6.4 fewer)

⊕⊕⊕⊝
MODERATE 1

OR 0.08
(0.00 to 1.61)

25.0%

2.6%
(0.0 to 34.9)

22.4% fewer
(25 fewer to 9.9 more)

⊕⊕⊝⊝
LOW 1, 3

*The risk in the intervention group (and its 95% confidence interval) is based on the mean risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio; OHSS: ovarian hyperstimulation syndrome

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded, and in some cases methods of randomization unclear.
2Downgraded two levels for very serious and unexplained heterogeneity: I2 = 85%, direction of effect inconsistent
3Downgraded one level for serious imprecision.
4Downgraded two levels for very serious imprecision: Broad confidence interval, very few events.

Open in table viewer
Summary of findings 2. LOD compared to other surgical interventions for symptom relief in women with polycystic ovary syndrome

LOD compared to other surgical interventions for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: LOD
Comparison: other surgical interventions

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD

With LOD

Difference

Menstrual regularity ‐ LOD vs unilateral LOD
N of participants: 104
(2 RCTs)

OR 1.51
(0.62 to 3.71)

71.2%

78.8%
(60.5 to 90.1)

7.7% more
(10.7 fewer to 19 more)

⊕⊕⊕⊝
MODERATE 1

1 study follow‐up at 3 months. 1 study follow‐up at 12 months

Menstrual regularity at 6 months ‐ LOD vs ultrasound‐guided transvaginal ovarian drilling
N of participants: 147
(1 RCT)

OR 1.23
(0.64 to 2.37)

54.7%

59.7%
(43.6 to 74.1)

5.1% more
(11.1 fewer to 19.4 more)

⊕⊕⊝⊝
LOW 1 2

Menstrual regularity at 12 months
Laser LOD vs harmonic scalpel
N of participants: 34
(1 RCT)

OR 2.13
(0.17 to 26.03)

88.2%

94.1%
(56.0 to 99.5)

5.9% more
(32.2 fewer to 11.3 more)

⊕⊕⊝⊝
LOW 1, 3

Note control group is NdYAG Laser

Improvement in androgenic symptoms at 6 months (Acne) ‐ LOD vs USS guided
N of participants:31
(1 RCT)

OR 0.84 (0.20 to 3.5)

47.1%

42.7%
(15.1 to 75.7)

4.3% fewer
(32 fewer to 28.6 more)

⊕⊕⊝⊝
LOW 1, 2

Improvement in androgenic symptoms at 6 months (Hirsutism) ‐ LOD vs USS‐guided
N of participants: 39
(1 RCT)

OR 1.09
(0.30 to 3.91)

40.0%

42.1% (16.7 to 72.3)

2.1% more (23.3 fewer to 32.3 more)

⊕⊕⊝⊝
LOW 1, 2

Harms: Adhesions at 6 months ‐ LOD vs THL
N of participants: 246
(1 RCT)

OR 0.10
(0.05 to 0.18)

59.3%

12.7%
(6.8 to 20.8)

46.6% fewer
(52.5 fewer to 38.5 fewer)

⊕⊝⊝⊝
VERY LOW 1, 4

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio; THL: transvaginal hydrolaparoscopy

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded or unclear allocation concealment or unclear randomization method.
2Downgraded one level for serious imprecision.
3Single study, narrow confidence interval.
4Downgraded two levels for very serious imprecision: Broad confidence interval, very few events.

Open in table viewer
Summary of findings 3. LOD 4‐5 compared to 2 or fewer punctures for symptom relief in women with polycystic ovary syndrome

LOD 4 ‐ 5 punctures compared to 2 or fewer punctures for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: LOD 4 ‐ 5 punctures
Comparison: 2 or fewer punctures

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD 4‐5

With LOD 4‐5

Difference

Menstrual regularity at 6 months ‐ LOD 4 ‐ 5 coagulation points compared to 2 or fewer
N of participants: 73
(2 RCTs)

OR 16.04
(4.19 to 61.34)

13.9%

72.1%
(40.3 to 90.8)

58.2% more
(26.4 more to 76.9 more)

⊕⊕⊝⊝
LOW 1, 2

Menstrual regularity at 6 months ‐ LOD (4 ‐ 5 laser coagulation points) vs 1 laser coagulation point per ovary
N of participants: 40
(1 RCT)

OR 19.00
(2.12 to 170.38)

5.0%

50.0%
(10.0 to 90.0)

45.0% more
(5 more to 85 more)

⊕⊕⊝⊝
LOW 1, 2

Menstrual regularity at 6 months ‐ LOD 4 punctures vs 2 punctures per ovary
N of participants: 33
(1 RCT)

OR 14.00
(2.60 to 75.41)

25.0%

82.4%
(46.4 to 96.2)

57.4% more
(21.4 more to 71.2 more)

⊕⊕⊝⊝
LOW 1, 2

Improvement in androgenic symptoms

No data available

Harms

LOD 4 ‐ 5 versus fewer punctures

No data available

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio;

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded or methods of randomization unclear.
2Downgraded one level for serious imprecision.

Open in table viewer
Summary of findings 4. LOD compared to LOD variable energy for symptom relief in women with polycystic ovary syndrome

LOD compared to LOD variable energy for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: LOD
Comparison: LOD variable energy

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD

With LOD

Difference

Menstrual regularity at 6 months ‐ LOD vs adjusted thermal dose
N of participants: 115
(1 RCT)

OR 0.42
(0.16 to 1.14)

87.9%

75.4%
(53.8 to 89.3)

12.6% fewer
(34.1 fewer to 1.3 more)

⊕⊝⊝⊝
VERY LOW 1, 2

Menstrual regularity at 3 months ‐ LOD unipolar vs LOD bipolar
N of participants: 20
(1 RCT)

OR 1.00
(0.05 to 18.57)

90.0%

90.0%
(31.0 to 99.4)

0.0% fewer
(59 fewer to 9.4 more)

⊕⊝⊝⊝
VERY LOW 1, 2

Groups had different metabolic characteristics at baseline

Improvement in androgenic symptoms

No data available

Harms: Adhesions at 6 months
№ of participants: 64
(1 study)

OR 0.96
(0.32 to 2.88)

28.6%

27.7%
(11.3 to 53.5)

0.8% fewer
(17.2 fewer to 25 more)

⊕⊝⊝⊝

VERY LOW1, 2

Women that remained enrolled for second‐look laparoscopy

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio;

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded or unclear allocation concealment.
2Downgraded two levels for very serious imprecision: Broad confidence interval, very few events.

Background

available in

Description of the condition

Polycystic ovary syndrome (PCOS) is one of the most common endocrine disorders in women of childbearing age. Typically, women will have menstrual disturbances, increased hair growth (hirsutism), acne, infrequent or absent menstrual periods and subfertility. PCOS affects approximately 10% of women worldwide (Adams 1986; Homburg 2008).

A proportion of women with PCOS will present clinically with hirsutism, acne and androgen‐dependent alopecia. Hirsutism in women is defined as an increased growth of terminal hair in a male pattern. The prevalence of clinical signs of hyperandrogenism in women of reproductive age is around 5% to 25% (Azziz 2000). It is characterised biochemically by raised serum concentrations of androgens, particularly testosterone and androstenedione. These features are associated with hypersecretion of luteinising hormone (LH) but with normal or low serum concentrations of follicle stimulating hormone (FSH). In 80% of women with PCOS, the testosterone concentration will exceed the upper limit of 2.4 nmol/l (Taylor 2003).

Women with PCOS also often present with menstrual disorders such as heavy menstrual bleeding and oligomenorrhoea. It is estimated that 30% to 70% of women with PCOS are obese (Vrbikova 2009), and many will develop type 2 diabetes, metabolic disorders and cardiovascular disease (Giallauria 2008; Kiddy 1992).

The Rotterdam meeting in 2003 endorsed a consensus regarding the diagnosis of the syndrome (ESHRE/ASRM PCOS Consensus 2004a; ESHRE/ASRM PCOS Consensus 2004b). This defines PCOS, after the exclusion of related disorders, as a clinical condition exhibiting at least two of the following three features: oligo‐ or anovulation (infrequent or absent menstrual periods); clinical or biochemical, or both, signs of hyperandrogenism (typically, hirsutism or acne); or polycystic ovaries seen on ultrasound.

A further meeting in Greece in 2007 proposed evidence‐based management for the treatment of PCOS‐related subfertility (Thessaloniki ESHRE/ASRM PCOS Consensus 2008). However, very little consensus guidance addresses the management of the non‐fertility‐related symptoms of this condition.

Conservative management with advice on lifestyle changes such as acupuncture (Lim 2016), dieting and weight loss has variable effectiveness. Medical treatment with statins (Raval 2011), hormonal therapy or insulin sensitising agents (Costello 2007) can be associated with significant side effects and is often contraindicated in women wishing to conceive. Laparoscopic ovarian drilling (LOD) is a surgical alternative to medical treatment of anovulation in women with PCOS and subfertility. However, whilst studies have examined the effectiveness of LOD in relation to the improvement of conception, reports on the effect of LOD on non‐fertility‐related PCOS symptoms are contradictory.

Description of the intervention

Ovarian wedge resection was first described in 1935 by Stein and Leventhal (Homburg 2008). This surgery was primarily aimed at reducing ovarian mass by bilateral ovarian wedge resections. With the advent of laparoscopic surgery, a variant of the traditional wedge resection was developed using this technique. During LOD, uni‐ or bipolar electrocautery at various energy levels for a variable duration is used to puncture the ovary. These are typically 2 to 4 mm deep, penetrating into the cortex. The number of puncture holes, the type of energy source and the duration of treatment vary between practitioners. Recent advances in transvaginal hydrolaparoscopy have also allowed ovarian drilling to be performed by the transvaginal route (Gordts 2009), using a hydrolaparoscope or fertiloscope. Ovarian surgery is traditionally performed to induce ovulation in anovulatory women with PCOS. The documented advantage of LOD compared with hormone treatment is the reduction in complication rates associated with ovarian hyperstimulation syndrome (OHSS) and a decreased risk of multiple pregnancy (Farquhar 2007).

How the intervention might work

The mechanism of action of ovarian surgery, wedge resection or more commonly LOD is largely unexplained. Compared to medical treatment, where the treatment effect is dependent on continuous administration of the medication, many LOD‐induced effects appear to be long‐term. It is, however, not known whether LOD exerts its action through a direct effect on the ovary or through a systemic endocrine mechanism. LOD has been shown to reduce long‐term androgen serum levels through a number of mechanisms (Amer 2002). The destruction of androgen‐producing ovarian stroma and the subsequent reduction in substrate for steroid aromatisation may contribute to an overall reduction in androgen production, with improvements in acne and hirsutism. High testosterone levels increase terminal hair growth, and therefore 5α‐reductase inhibitors are implemented in the treatment of hirsutism. A more generalised increased 5α‐reductase activity may be important for increased cortisol metabolism in PCOS (Glintborg 2010). In LOD, it has been hypothesised that the drilling of punctures within the androgen‐producing stroma may depress serum androgen concentrations and offer symptomatic relief. A further possible mechanism is an impact of LOD on insulin resistance. Whilst there is currently very little evidence which documents the effect of LOD on metabolic changes and insulin levels in women with PCOS, small studies have shown that LOD (by electrocautery) can reduce insulin resistance in women with PCOS through decreased IRS‐1 Ser312 phosphorylation (Seow 2007). Other more recent studies have also suggested that LOD can decrease anti‐müllerian hormone (AMH) concentrations and ovarian stroma blood flow in women with PCOS compared with controls (Elmashad 2011).

Why it is important to do this review

The aim of this review is to evaluate the impact of ovarian surgery on symptomatic control of PCOS. Whilst much research has focused on the effect of ovarian surgery in improving fertility in clomiphene‐resistant women with PCOS, the impact of LOD on symptoms of PCOS has not been reviewed.

Objectives

available in

To assess the effectiveness and harms of ovarian surgery as a treatment for symptomatic relief of hirsutism, acne and menstrual disturbances in women with PCOS.

Methods

available in

Criteria for considering studies for this review

Types of studies

We include only randomized controlled trials (RCTs) of surgical interventions for the treatment of symptoms associated with PCOS. Only pre‐crossover data from cross‐over studies was to be included in this study. Other data from cross‐over studies would not be included, even if there was a wash‐out period, as the effect of LOD is long term (for example some women will ovulate regularly after one year).

Types of participants

Women with PCOS (as defined by the Rotterdam Criteria) who are symptomatic with acne, hirsutism or irregular menstrual cycles. Oligomennorhoea is here defined as a menstrual cycle lasting more than 45 days.

Types of interventions

Ovarian surgical interventions for the symptomatic treatment of women with PCOS. Types of comparisons include the following:

1. Surgical (laparotomy or laparoscopy) versus non‐surgical intervention (including placebo or non‐treatment).

2. Comparison between various surgical methods or techniques, e.g.:

  • Comparing the number of drill holes administered.

  • Comparison of various energy modalities for the ovarian drilling procedure (e.g. uni‐ versus bipolar electrocautery, or electrocautery versus laser versus harmonic scalpel).

  • Laparotomy versus laparoscopy.

Types of outcome measures

Primary outcomes

1. Rates of women reporting an improvement in menstrual regularity as defined by the number of cycles per year (or other time frame, e.g. six months) and recorded as regular or irregular.

2. Improvement in androgenic symptoms of PCOS or rates of androgenic symptoms: improvements in hirsutism or acne, as defined by scoring systems such as the Ferriman Gallwey Score (FGS), Global Acne Grading Score (GAGS) or the Leeds revised Acne Grading system (LRAGS).

  • FGS divides the body into nine areas identified for assessment: upper lip, chin, chest, upper back, lower back, upper abdomen, lower abdomen, upper arms and thighs. Each area is assigned a score between 0 and 4 based on hair growth, giving a potential score of 0 to 36. In white populations a score over seven indicates hyperandrogenism. There is ethnic variation of 'normal'.

  • GAGS divides the face, chest and back into six areas identified for assessment: forehead, each cheek, nose, chin, chest and back. Each area is assigned a factor of 1, 2 or 3 based on area. Acne lesions are given a value based on severity: no lesions = 0, comedones = 1, papules = 2, pustules = 3 and nodules = 4. Scores of 1 to 18 are considered mild, 19 to 30 are considered moderate, 31 to 38 are severe, and more than 39 very severe.

  • LRAGS assesses acne on the face, chest and back. A scale of 1 to 10 is used. 

Secondary outcomes

1. Harms of surgical and non‐surgical interventions

2. Change in body weight or body mass index

3. Change in waist circumference

4. Testosterone levels, free testosterone or free androgen index

5. Metabolic measures: fasting glucose or insulin levels or haemoglobin A1C (HbA1C)

6. Changes in quality of life

Search methods for identification of studies

We sought all published and unpublished RCTs on surgery for PCOS, using the following search strategy, without language restriction and in consultation with the Gynaecology and Fertility Group (CGF) Information Specialist.

Electronic searches

We searched the following electronic databases, trials registers and websites:

Cochrane Gynaecology and Fertility Group (CGF) Specialised Register (from inception to 17 October 2016, Procite platform) (Appendix 1).
Central Register of Studies Online (CRSO) (searched 17 October 2016, web platform) (Appendix 2).
Ovid MEDLINE® In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE® (from 1946 to 17 October 2016, Ovid platform) (Appendix 3).
Ovid Embase (from 1980 to 17 October 2016, Ovid platform) (Appendix 4).
Ovid PsycINFO (from 1806 to 17 October 2016, Ovid platform) (Appendix 5).

We searched other electronic sources of trials 17 October 2016, including:

Database of Abstracts of Reviews of Effects (DARE) in the Cochrane Library (www.cochrane.org/index.htm);
ClinicalTrials.gov (clinicaltrials.gov/ct2/home);
World Health Organization International Clinical Trials Registry Platform search portal (apps.who.int/trialsearch/);
OpenSigle for grey literature from Europe (opensigle.inist.fr/);
China Academic Journal Electronic full text Database in China National Knowledge Infrastructure;
Index to Chinese Periodical Literature.

LILACS and other Spanish and Portuguese language databases (LatinAmerican and Caribbean Health Science Information database). This is in the Virtual Health library Regional Portal (VHL) (bvsalud.org/portal/?lang=en).

There were no language or date restrictions in these searches. The search strategies are in the appendices.

Searching other resources

We handsearched the reference lists of articles retrieved by the search and made personal contact with experts in the field. We handsearched ESHRE conference abstracts that were not covered in the CGF Specialised Register, in collaboration with the Information Specialist.

Data collection and analysis

Selection of studies

The four review authors undertook the selection of studies. We used the search strategy described above to obtain titles and, where possible, abstracts of studies that were potentially relevant to the review. We were overly inclusive rather than risk losing relevant studies. All review authors independently assessed whether the studies were eligible for inclusion, with disagreements planned for consultation with a third author, although this was not required. Where papers had insufficient information to enable an accurate assessment of eligibility for inclusion, we sought further information from authors.

The selection process is documented with a PRISMA flow chart in Figure 1.

Data extraction and management

Two review authors independently extracted data and assessed risks of bias using forms designed according to Cochrane guidelines. We resolved disagreements about study eligibility by discussion, without requiring referral to a third review author. Where studies had multiple publications, we used the main trial report as the primary reference and derived additional details from secondary papers. We corresponded with study investigators to clarify further data on methods and results.

Assessment of risk of bias in included studies

Two review authors independently assessed the included studies for risks of bias, using the Cochrane 'Risk of bias' assessment tool (Higgins 2011) to evaluate: selection (random sequence generation and allocation concealment); performance (blinding of participants and personnel); detection (blinding of outcome assessors); attrition (incomplete outcome data); reporting (selective reporting); and other potential bias. We assigned judgements as recommended in the Cochrane Handbook for Systematic Reviews of Interventions Section 8.5 (Higgins 2011), resolving disagreements by discussion.

We present our judgements in a 'Risk of bias' table, which we incorporate into the interpretation of review findings by means of sensitivity analyses (see below).

Measures of treatment effect

We performed statistical analysis in accordance with the guidelines and methods developed by Cochrane (Higgins 2011). We expressed dichotomous data results as a Mantel‐Haenszel odds ratio (OR) and continuous data results as a mean difference (MD). We present a 95% confidence interval (CI) for all outcomes.

Unit of analysis issues

The primary analysis was by each woman randomized. Data reported that did not allow valid analysis (e.g. "per cycle" rather than "per woman" where women contributed more than one cycle) were to be briefly summarized in an additional table and would not be meta‐analysed.

Dealing with missing data

In the event that data were missing or not available, we planned to contact the authors to provide further assistance. Where we could obtain no further information, the review authors state this in the review. The review authors determined the individual weight of missing data. We conducted an intention‐to‐treat analysis as far as possible, but otherwise we analyzed only the available data.

Assessment of heterogeneity

We assessed heterogeneity using the I2 statistic, according to the guidelines set out in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). An I2 value greater than 50% indicates substantial heterogeneity. Where we detected substantial heterogeneity we have explored possible explanations in subgroup analyses (e.g. differing populations) and sensitivity analyses (e.g. differing risks of bias). We have taken heterogeneity in to account in our interpretation of the results.

Assessment of reporting biases

In view of the difficulty in detecting and correcting for publication bias and other reporting bias, we minimised their potential impact by ensuring a comprehensive search for eligible studies and by being aware of duplicated data. In the case where 10 or more studies contributed to our analysis, we planned to use a funnel plot to explore the possibility of a small‐study effect.

Data synthesis

We used a fixed‐effect model inverse variance meta‐analysis for combining data where trials examined the same intervention, and where we judged the trial populations and methods to be sufficiently similar.

An increase in the odds of a particular outcome which was beneficial (for example, return of regular menses) or detrimental (for example, surgical complications) is displayed graphically in the meta‐analyses. The aim was to define analyses that were comprehensive and mutually exclusive so that all eligible study results could be fitted into only one of the strata.

Subgroup analysis and investigation of heterogeneity

We planned, where data allowed, to carry out subgroup analyses to explore possible sources of heterogeneity (for example, differences between participants, interventions and study quality).

Sensitivity analysis

We intended to conduct sensitivity analyses for the primary outcomes to explore the effect of risks of bias assessed by adequate methodology versus poor methodology, where adequate methodology was defined a: adequate randomization method, adequate allocation concealment, analysis by intention‐to‐treat, and losses to follow‐up of less than 20%. However, sensitivity analysis in this respect was not necessary and therefore not performed.

These analyses also included consideration of whether the review conclusions would have differed if:

1. A random‐effects model had been adopted.

2. The summary effect measure had been relative risk rather than odds ratio.

Overall quality of the body of evidence: 'Summary of findings' table

We have prepared a 'Summary of findings' table using GRADEpro and Cochrane methods. This table evaluates the overall quality of the body of evidence for the main review outcomes (improvement in menstrual regularity, improvement in androgenic symptoms of PCOS, harms of medical and surgical intervention) for the main review comparison (LOD versus medical intervention). summary of findings Table for the main comparison.

We prepared additional 'Summary of findings' tables for the main review outcomes for other important comparisons (LOD versus other surgical interventions, LOD four to five versus two or fewer punctures, LOD versus variable energy source). summary of findings Table 2; summary of findings Table 3; summary of findings Table 4.

We assessed the quality of the evidence using GRADE criteria: risk of bias, consistency of effect, imprecision, indirectness and publication bias. Two review authors working independently judged evidence quality (high, moderate, low or very low), with disagreements resolved by discussion. Judgements were justified, documented, and incorporated into the reporting of results for each outcome.

Results

Description of studies

Only pre‐cross‐over data from cross‐over studies were to be included in this review. We do not include other data from cross‐over studies, even if there is a wash‐out period, as the effect of LOD is long‐term (for example, some women will ovulate regularly after one year).

Results of the search

The search retrieved 1115 articles. Thirty‐one studies were potentially eligible and retrieved in full text, with 22 meeting our inclusion criteria and included in the final analysis. See study tables: Characteristics of included studies; Characteristics of excluded studies.

See Figure 1.


Study flow diagram

Study flow diagram

Included studies

Study Design and setting

22 randomized controlled trials included in this review were set in the high‐, middle‐ and low‐income countries, in private, public and university hospitals. All trials had a parallel design, and all but one were single‐centre studies.

Participants

The studies included 2278 women. All the participants had polycystic ovary syndrome. The characteristics of the participants can be found in the table for each included study. Participants were included regardless of whether their treatment goal was symptom management of PCOS or, more commonly, fertility outcomes.

Interventions

Ten of the 22 studies were designed to assess the impact of LOD on the outcome measures specified earlier, compared to a medical treatment. The comparisons were as follows:

Traditional LOD compared to;

Ten of 22 studies were designed to assess the impact of LOD on the outcome measures specified compared to a different surgical technique/method of ovarian drilling. The comparisons were as follows:

Traditional LOD compared to;

Two of 22 studies were head‐to‐head comparisons between different dosage/number of punctures of LOD (Selim 2011; Zhu 2010)

No studies compared LOD with no treatment/placebo, or laparotomy. No studies compared LOD to wedge resection.

Outcomes

The following studies reported on the specified outcome measures in this review:

1. Menstrual regularity: Abdelhafeez 2013; Ashrafinia 2009; Badawy 2009; Elgafor 2013; Hamed 2010; Hashim 2010; Hashim 2011; Kaya 2005; Palomba 2010; Roy 2009; Selim 2011; Takeuchi 2002; Zakherah 2011; Zhu 2010

2. Androgenic symptoms: Ashrafinia 2009; Badawy 2009; Farquhar 2002

3. Harms of surgical intervention: short‐ and long‐term surgical complications:

4. Body weight or body mass index (kg/m2): Elgafor 2013; Hamed 2010; Farquhar 2002; Zakherah 2011

5. Change in waist circumference: None of the studies reported on waist circumference.

6. Testosterone levels (nmol/l), free testosterone (pg/ml) or free androgen index: Ashrafinia 2009; Badawy 2009; Elgafor 2013; Hamed 2010; Roy 2009; Roy 2010; Sarouri 2015; Selim 2011; Sharma 2006; Takeuchi 2002; Taskin 1996; Youssef 2007; Zakherah 2011; Zhu 2010; Zullo 2000

7. Metabolic measures: fasting glucose (nmol/L) or insulin levels (pmol/L) or haemoglobin A1C (HbA1C): Elgafor 2013; Hamed 2010; Roy 2010

8. Quality of life: Farquhar 2002 provided a questionnaire on the acceptability and convenience of both procedures.

Excluded studies

We excluded nine studies from the review. Reasons for exclusion include studies not being RCTs (three studies), one study was retracted, one study did not measure testosterone as planned, and the remainder did not meet our inclusion criteria.

Risk of bias in included studies

Risk of bias assessment focused on seven main domains: random sequence generation, allocation concealment, blinding, incomplete data outcome, selective reporting and other potential bias. The findings are summarized in Figure 2 and Figure 3, and details appear in individual 'Risk of bias' included studies tables.


Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.


Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Sequence generation

Eighteen trials were at low risk of selection bias related to sequence generation. Four studies had unclear risk due to a lack of detail about their methods of sequence generation.

Allocation concealment

Twelve trials were at low risk of bias for allocation concealment, with the other 10 trials at unclear risk of bias.

Blinding

Blinding of participants

Due to the nature of the studies, blinding was often not possible. Fourteen trials had no means of blinding due to the surgical study design and so were at high risk of performance and detection bias; the remaining eight trials were deemed to be at high risk as they stated a lack of participant blinding.

Blinding of assessment

Most of the studies (20/22) were at unclear risk of assessment blinding. Two studies were at low risk, as they specified that the assessors were blinded.

Incomplete outcome data

Attrition rates in all studies were low and hence we rated all studies at low risk for this domain. One study (Giampaolino 2016) had attrition rates of 45/246, with 19 lost from one arm and 26 from the other.

Selective reporting

All the studies reported on the outcome specified.

Other potential sources of bias

One study (Sharma 2006) showed a statistically significant difference in baseline levels of glucose between the control and intervention groups.

Effects of interventions

See: Summary of findings for the main comparison LOD compared to medical interventions for symptom relief in women with polycystic ovary syndrome; Summary of findings 2 LOD compared to other surgical interventions for symptom relief in women with polycystic ovary syndrome; Summary of findings 3 LOD 4‐5 compared to 2 or fewer punctures for symptom relief in women with polycystic ovary syndrome; Summary of findings 4 LOD compared to LOD variable energy for symptom relief in women with polycystic ovary syndrome

We include 22 studies in this review.

As the studies were sufficiently similar, we combined the data using a fixed‐effect model in the following comparisons:

  1. LOD versus medical interventions (stratified by type of medication)

  2. LOD versus other surgical interventions (stratified by type of surgery)

  3. LOD with four to five punctures versus LOD with two or fewer punctures

  4. LOD versus variable energy

1. LOD versus medical treatments

Primary outcomes
1.1 Improvement of menstrual regularity

Seven studies reported this outcome. We did not pool studies comparing different medications, as this resulted in high statistical heterogeneity (I2 = 66%). We interpret the underlying cause of the heterogeneity to be attributable to the differences between the interventions.

1.1.1. LOD versus metformin: Two studies reported this comparison. Findings were inconsistent and the data were unsuitable for pooling due to high statistical heterogeneity (I2=85%). There was no obvious difference between the studies that might explain the heterogeneity. One study (Hamed 2010) suggested a benefit in the LOD arm (OR 2.32, 95% CI 1.02 to 5.28, 110 women) while the other study (Ashrafinia 2009) had the opposite direction of effect but found no conclusive evidence of a difference between the groups (OR 0.55, 95% CI 0.27 to 1.13, 126 women). See Analysis 1.1; Figure 4.


Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.1 Menstrual regularity.

Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.1 Menstrual regularity.

1.1.2. LOD versus metformin + clomiphene: There was no clear evidence of a difference between the groups at six months (OR 1.02, 95% CI 0.64 to 1.64, 2 RCTs, n = 332, I2 = 13%, low‐quality evidence) (Hashim 2011; Palomba 2010).

1.1.3 LOD versus gonadotrophin: One study showed evidence of benefit of LOD over gonadotrophin at six months (OR 19.20, 95% CI 3.17 to 116.45, 1 RCT, n = 35, low‐quality evidence) (Kaya 2005).

1.1.4. LOD versus letrozole: There was no clear evidence of a difference between the groups at six months (OR 1.08, 95% CI 0.64 to 1.84, 1 RCT, n = 260, moderate‐quality evidence) (Hashim 2010).

1.1.5. LOD versus metformin + letrozole: There was no clear evidence of a difference between the groups at six months (OR 0.95, 95% CI 0.49 to 1.81, 1 RCT, n = 146, moderate‐quality evidence) (Elgafor 2013).

See Analysis 1.2; Figure 5


Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.2 Menstrual regularity.

Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.2 Menstrual regularity.

1.2 Improvement of androgenic symptoms (hirsutism or acne, or both)

1.2.1. LOD versus metformin: There was no clear evidence of a difference between the groups at six months (OR 1.00, 95% CI 0.42 to 2.37, 1 RCT, n = 126, moderate‐quality evidence) (Ashrafinia 2009).

1.2.2. LOD versus gonadotrophins: There was no clear evidence of a difference between the groups in improvement of acne (OR 3.20, 95% CI 0.33 to 30.94, 1 RCT, n = 25, low‐quality evidence) or hirsutism at six months (OR 2.31, 95% CI 0.22 to 23.89, 1 RCT, n = 25, very low‐quality evidence) (Farquhar 2002).

Secondary outcomes
1.3 Harms

1.3.1. LOD versus metformin + clomiphene: There were more gastrointestinal side effects in the metformin + clomiphene group at six months (OR 0.05, 95% CI 0.01 to 0.36, 2 RCTs, n = 332 I2 = 0%, moderate‐quality evidence) (Hashim 2011; Palomba 2010).

1.3.2. LOD versus gonadotrophin: There was no clear evidence of a difference between the groups in OHSS rate at six months (OR 0.08, 95% CI 0.00 to 1.61, 1 study, n = 33, low‐quality evidence) (Kaya 2005).

1.3.3. LOD versus rosiglitazone: There was no occurrence of OHSS in either group at six months (Roy 2010).

1.4 Body weight or body mass index (kg/m2)

1.4.1. LOD versus metformin: There was a higher body mass index (BMI) at six months follow‐up in the LOD group compared to the metformin group (MD 3.60, 95% CI ‐6.28 to 13.48, 1 RCT, n = 110) (Hamed 2010).

1.4.2. LOD versus metformin + letrozole: There was no clear evidence of a difference in BMI at six months follow‐up between groups (MD 0.47, 95% CI ‐0.90 to 1.84, 1 RCT, n = 146) (Elgafor 2013).

1.5 Waist circumference

None of the studies reported on waist circumference.

1.6 Testosterone levels (nmol/L), free testosterone (pg/ml) or free androgen index

1.6.1. LOD versus metformin: There was no clear evidence of a difference in free testosterone levels post‐surgery at six months follow‐up, (MD ‐0.34, 95% CI ‐0.46 to ‐0.22; n = 126 ; 1 RCT) (Ashrafinia 2009) There was no clear evidence of a difference in mean testosterone levels post‐surgery at six months follow‐up, (MD 0.20, 95% CI ‐0.61 to 1.01, 1 RCTs, n = 110) Hamed 2010).

1.6.2. LOD versus gonadotropin‐releasing hormone antagonist (GnRHa) + oral contraceptive pill (OCP): There was a higher testosterone level one week post‐surgery or at the first menstruation after treatment in the LOD group compared to the GnRHa + OCP group (MD 0.3, 95% CI 0.01 to 0.59, 1 RCT, n = 17) (Taskin 1996).

1.6.3. LOD versus metformin + letrozole: There was a higher testosterone level after LOD compared to metformin + letrozole use (MD 0.60, 95% CI 0.23 to 0.97, 1 RCT, n = 146) (Elgafor 2013).

1.6.4. LOD versus rosiglitazone: There was no clear evidence of a difference between the groups in testosterone levels between rosiglitazone and LOD (MD 0.08, 95% CI ‐0.11 to 0.27, 1 RCT, n = 43) (Roy 2010).

1.7 Metabolic measures: fasting glucose or insulin levels or ratios or haemoglobin A1C (HbA1C)

1.7.1. LOD versus rosiglitazone: There was a lower glucose/insulin ratio in the LOD group compared to the rosiglitazone group (MD ‐1.98, 95% CI ‐2.61 to ‐1.35, 1 RCT, n = 43) (Roy 2010).

1.7.2. LOD versus metformin: There was a lower glucose/insulin ratio in the LOD group compared to the metformin group (MD ‐2.60, 95% CI ‐4.55 to ‐0.65, 1 RCT, n = 110) (Hamed 2010).

1.7.3. LOD versus metformin + letrozole: There was a lower glucose/insulin ratio in the LOD group compared to the metformin + letrozole group (MD ‐1.35, 95% CI ‐2.21 to ‐0.49, 1 RCT, n = 146) (Elgafor 2013).

1.8 Quality of life

1.8.1. LOD versus gonadotrophins: Farquhar 2002 provided a questionnaire to the 19 women who underwent both LOD and gonadotrophins regarding the acceptability and convenience of both procedures. Seventeen returned the questionnaire; 15 women preferred LOD and two preferred gonadotrophins. LOD was described as "less traumatic".

2. Bilateral LOD versus other surgical techniques

Primary outcomes
2.1 Improvement of menstrual regularity

2.1.1. LOD versus unilateral LOD: There was no clear evidence of a difference between the groups (OR 1.51, 95% CI 0.62 to 3.71, 2 RCTs, n = 104, I2 = 0%, moderate‐quality evidence) (Abdelhafeez 2013; Roy 2009). Abdelhafeez 2013 reported at follow‐up of three months (OR 1.41, 95% CI 0.45 to 4.45, 1 RCT, n = 60), while Roy 2009 reported at follow‐up of 12 months (OR 1.69, 95% CI 0.40 to 7.07, 1 RCT, n = 44). We have pooled these data, as they are otherwise similar studies.

2.1.2. LOD versus transvaginal ultrasound‐guided LOD: There was no clear evidence of a difference between the groups at six months (OR 1.23, 95% CI 0.64 to 2.37, 1 RCT, n = 147, low‐quality evidence) (Badawy 2009).

2.1.3. LOD with Hd‐YAG laser versus LOD with harmonic scalpel: There was no clear evidence of a difference between the groups at 12 months (OR 2.13, 95% CI 0.17 to 26.03, 1 RCT, n = 34, low‐quality evidence) (Takeuchi 2002).

See Analysis 2.1; Figure 6.


Forest plot of comparison: 2 LOD vs other surgical interventions, outcome: 2.1 Menstrual regularity.

Forest plot of comparison: 2 LOD vs other surgical interventions, outcome: 2.1 Menstrual regularity.

2.2 Improvement of androgenic symptoms (hirsutism or acne, or both)

LOD versus transvaginal ultrasound‐guided LOD: There was no clear evidence of a difference between the groups in hirsutism (OR 1.09, 95% CI 0.30 to 3.91, 1 RCT, n = 39, low‐quality evidence) or acne at six months (OR 0.84, 95% CI 0.20 to 3.50, 1 RCT, n = 31, low‐quality evidence) (Badawy 2009).

Secondary outcomes
2.3 Harms

2.3.1. LOD versus unilateral LOD: There were no post‐surgical peri‐ovarian adhesions found in a subgroup of women who returned for repeat laparoscopy or caesarean section up to 12 months post‐intervention (Roy 2009).

2.3.2. LOD versus minilaparoscopic LOD under local anaesthetic and conscious sedation: There were no complications from local anaesthetic (Zullo 2000).

2.3.3 LOD versus transvaginal hydrolaparoscopy: There were significantly fewer women with post‐procedure adhesions with THL compared to LOD at six months (OR 0.10, 95% CI 0.05 to 0.18, 1 RCT, n = 246, very low‐quality evidence) (Giampaolino 2016).

2.4 Body weight or body mass index (kg/m2)

None of the studies reported on this outcome.

2.5 Waist circumference

None of the studies reported on this outcome.

2.6 Testosterone levels (nmol/L), free testosterone (pg/ml) or free androgen index

2.4.1. LOD versus transvaginal ultrasound‐guided LOD: The mean testosterone levels were found to be higher in the LOD group compared to the ultrasound‐guided transvaginal LOD group (MD 0.30, 95% CI 0.13 to 0.47, 1 RCT, n = 163) (Badawy 2009).

2.4.2. LOD versus minilaparoscopic LOD under local anaesthetic and conscious sedation: There was no clear evidence of a difference between the groups (MD 0.00, 95% CI ‐0.43 to 0.43, 1 RCT, n = 62) (Zullo 2000).

2.4.3. LOD versus unilateral LOD: There was no clear evidence of a difference in mean testosterone between the groups (MD 0.03, 95% CI ‐0.04 to 0.09, 2 RCTs, n = 112, I2 = 0%) (Roy 2009; Youssef 2007). There was no clear evidence in free testosterone between the groups (MD ‐0.30, 95% CI ‐0.84 to 0.24, 1 RCTs, n = 90) (Sarouri 2015).

2.4.4. LOD with Hd‐YAG Laser versus LOD with harmonic scalpel: There was no clear evidence of difference between the groups (MD 0.0, 95% CI ‐0.28 to 0.28, 1 RCT, n = 34) (Takeuchi 2002).

2.7 Metabolic measures: fasting glucose or insulin levels or ratio or haemoglobin A1C (Hb1AC)

None of the studies reported on this outcome.

2.8 Quality of life

None of the studies reported on this outcome.

3. LOD with different numbers of punctures per ovary

Primary outcomes
3.1 Improvement of menstrual regularity

LOD four to five punctures versus two or fewer punctures per ovary: There is evidence of benefit of applying four or five punctures per ovary compared to two or fewer punctures at six months (OR 16.04, 95% CI 4.19 to 61.34, 2 RCTs, n = 73, I2 = 0%, low‐quality evidence) (Zhu 2010; Selim 2011).

See Analysis 3.1; Figure 7.


Forest plot of comparison: 3 LOD 4 ‐ 5 vs 2 or fewer punctures, outcome: 3.1 Menstrual regularity.

Forest plot of comparison: 3 LOD 4 ‐ 5 vs 2 or fewer punctures, outcome: 3.1 Menstrual regularity.

3.2 Improvement of androgenic symptoms (hirsutism or acne, or both)

None of the studies reported on this outcome.

Secondary outcomes
3.3 Harms

LOD four to five punctures versus two or fewer punctures: There were no surgical adverse events in any group (Zhu 2010).

3.4 Body weight or body mass index (kg/m2)

None of the studies reported on this outcome.

3.5 Waist circumference

None of the studies reported on this outcome.

3.6 Testosterone levels (nmol/L) or free androgen index

Testosterone

3.6.1. LOD four to five punctures versus two or fewer punctures: There is evidence of a difference in applying more punctures per ovary compared to two or fewer punctures (MD ‐0.90, 95% CI ‐1.12 to ‐0.68, 2 RCTs, n = 73, I2 = 0%) (Selim 2011; Zhu 2010) (Analysis 3.2).

Free androgen index (FAI)

3.6.2. LOD four to five punctures versus two or fewer punctures: There was no clear evidence of a difference between the groups (MD ‐1.50, 95% CI ‐3.21 to 0.21, 1 RCT, n = 33) (Selim 2011) (Analysis 3.2).

3.7 Metabolic measures: fasting glucose or insulin levels or haemoglobin A1C (HbA1C)

None of the studies reported on this outcome.

3.8 Quality of life

None of the studies reported on this outcome.

4. LOD with various energy modalities

Primary outcomes
4.1 Improvement of menstrual regularity

4.1.1. LOD versus LOD using adjusted thermal dose in accordance with the ovarian volume: There was no clear evidence of a difference between the groups at six months (OR 0.42, 95% CI 0.16 to 1.14, 1 RCT, n = 115, very low‐quality evidence) (Zakherah 2011).

4.1.2. Unipolar LOD versus bipolar LOD: There was no clear evidence of a difference between the groups at three months (OR 1.00, 95% CI 0.05 to 18.57, 1 RCT, n = 20, very low‐quality evidence) (Sharma 2006).

See Analysis 4.1; Figure 8.


Forest plot of comparison: 4 LOD vs LOD variable energy, outcome: 4.1 Menstrual regularity.

Forest plot of comparison: 4 LOD vs LOD variable energy, outcome: 4.1 Menstrual regularity.

4.2 Improvement of androgenic symptoms (hirsutism or acne, or both)

None of the studies reported on this outcome.

Secondary outcomes
4.3 Harms

LOD versus LOD using adjusted thermal dose in accordance with ovarian volume: There was no clear evidence of a difference between the groups in the adhesions found at second‐look laparoscopy at six months (OR 0.96, 95% CI 0.32 to 2.88, 1 RCT, n = 64, very low‐quality evidence) (Zakherah 2011).

4.4 Body weight or body mass index (kg/m2)

LOD versus LOD using adjusted thermal dose in accordance with ovarian volume: There was no clear evidence of a difference between the groups (MD 0.20, 95% CI ‐0.71 to 1.11, 1 RCT, n = 115) (Zakherah 2011).

4.5 Waist circumference

None of the studies reported on this outcome.

4.6 Testosterone levels (nmol/L) or free androgen index

4.6.1. LOD versus LOD using adjusted thermal dose in accordance with ovarian volume: Testosterone levels were lower in the thermal dose group (MD 0.70, 95% CI 0.57 to 0.83, 1 RCT, n = 115) (Zakherah 2011) (Analysis 4.4).

4.6.2. Unipolar LOD versus bipolar LOD: There was no clear evidence of a difference between the groups in the testosterone levels (MD 0.30, 95% CI ‐2.01 to 2.61, 1 RCT, n = 20) (Sharma 2006) (Analysis 4.4).

4.7 Metabolic measures: fasting glucose or insulin levels or haemoglobin A1C (HbA1C)

4.5.1. Unipolar LOD versus bipolar LOD: There was a significantly lower glucose insulin ratio at three months after surgery in the group who had unipolar LOD compared to those with bipolar LOD (MD 4.10, 95% CI 3.08 to 5.12, 1 RCT, n = 20) (Sharma 2006). However, in this study there was a significant disparity between the baseline characteristics of the glucose insulin ratio between the two groups.

4.8 Quality of life

None of the studies reported on this outcome.

5. LOD versus laparotomy

There were no studies that investigated this.

Other analyses

Upon completing study selection and data extraction we considered no subgroups to be relevant to our analysis.

We conducted sensitivity analyses as planned. The sensitivity analyses resulted in no significant modification of the results or the strength of associations after applying the above considerations to our data. No imputation of data was required. We are unable to address publication bias in this review.

Discussion

available in

Summary of main results

For women with menstrual irregularity, acne or hirsutism due to PCOS there was evidence of benefit of LOD over gonadotrophin, and for high puncture number versus low puncture number in menstrual regularity. These outcomes were based on one and two RCTs respectively. As gonadotrophins are unlikely to be prescribed solely for these indications, it is important to note that these outcomes are secondary.

There was no clear evidence of a difference in the rates of improvement of menstrual regularity or androgenic symptoms when LOD was compared to other medical or other surgical interventions.

Studies reported low adverse effects with LOD, and metformin and clomiphene is associated with greater gastrointestinal side effects compared to LOD. Given the evidence supporting metformin as beneficial in the improvement of menstrual regularity (Tang 2012), LOD may be considered a non‐inferior but invasive alternative in those unable to tolerate metformin.

There were significantly fewer post‐procedure adhesions with transvaginal hydrolaparoscopy when compared with LOD, but this was based on a single RCT.

Overall completeness and applicability of evidence

The studies included addressed the review question, and the participants and outcomes were all relevant. Certain comparisons of surgical technique will not be relevant in centres where the equipment is not available. There were no studies comparing LOD versus no treatment and hence our conclusions can only be drawn from head‐to‐head comparisons of LOD with medical or surgical treatments. Current evidence is largely limited by data obtained from single or small numbers of randomised controlled studies. When confronted with a woman with symptoms of PCOS, this evidence needs to be applied in the wider context of their symptoms and goals, as neither LOD nor the medical interventions are considered first‐line treatments. There may be additional benefit, however, if these interventions are otherwise indicated; they should form part of patient counselling and may guide certain decisions.

Quality of the evidence

The quality of the evidence ranged from very low to moderate quality. The main limitations were imprecision associated with the low number of studies, inconsistency and risk of bias associated with the inability to blind participants (see summary of findings Table for the main comparison; summary of findings Table 2; summary of findings Table 3; summary of findings Table 4). The risk of performance bias in individual studies was high, due to a lack of blinding of outcome assessors, which can be expected with surgical comparisons. There were too few studies to assess risk of publication bias. Other bias was rare, with the exception of Sharma 2006 which we rated at high risk of other bias due to a significant disparity in the baseline characteristics of glucose/insulin ratio.

Potential biases in the review process

We executed an inclusive search, minimising as far as possible incomplete identification of studies and the risk of reporting bias. We made every effort to identify all potentially eligible studies, and sought additional data from study authors as necessary. However, it is possible that there are unpublished studies that we did not retrieve.

Agreements and disagreements with other studies or reviews

We found no other studies or reviews on this topic.

Study flow diagram
Figures and Tables -
Figure 1

Study flow diagram

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Figures and Tables -
Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Figures and Tables -
Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.1 Menstrual regularity.
Figures and Tables -
Figure 4

Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.1 Menstrual regularity.

Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.2 Menstrual regularity.
Figures and Tables -
Figure 5

Forest plot of comparison: 1 LOD vs medical interventions, outcome: 1.2 Menstrual regularity.

Forest plot of comparison: 2 LOD vs other surgical interventions, outcome: 2.1 Menstrual regularity.
Figures and Tables -
Figure 6

Forest plot of comparison: 2 LOD vs other surgical interventions, outcome: 2.1 Menstrual regularity.

Forest plot of comparison: 3 LOD 4 ‐ 5 vs 2 or fewer punctures, outcome: 3.1 Menstrual regularity.
Figures and Tables -
Figure 7

Forest plot of comparison: 3 LOD 4 ‐ 5 vs 2 or fewer punctures, outcome: 3.1 Menstrual regularity.

Forest plot of comparison: 4 LOD vs LOD variable energy, outcome: 4.1 Menstrual regularity.
Figures and Tables -
Figure 8

Forest plot of comparison: 4 LOD vs LOD variable energy, outcome: 4.1 Menstrual regularity.

Comparison 1 LOD vs medical interventions, Outcome 1 Menstrual regularity.
Figures and Tables -
Analysis 1.1

Comparison 1 LOD vs medical interventions, Outcome 1 Menstrual regularity.

Comparison 1 LOD vs medical interventions, Outcome 2 Menstrual regularity.
Figures and Tables -
Analysis 1.2

Comparison 1 LOD vs medical interventions, Outcome 2 Menstrual regularity.

Comparison 1 LOD vs medical interventions, Outcome 3 Improvement in androgenic symptoms (hirsutism/acne).
Figures and Tables -
Analysis 1.3

Comparison 1 LOD vs medical interventions, Outcome 3 Improvement in androgenic symptoms (hirsutism/acne).

Comparison 1 LOD vs medical interventions, Outcome 4 Harms.
Figures and Tables -
Analysis 1.4

Comparison 1 LOD vs medical interventions, Outcome 4 Harms.

Comparison 1 LOD vs medical interventions, Outcome 5 BMI.
Figures and Tables -
Analysis 1.5

Comparison 1 LOD vs medical interventions, Outcome 5 BMI.

Comparison 1 LOD vs medical interventions, Outcome 6 Testosterone and free androgen index.
Figures and Tables -
Analysis 1.6

Comparison 1 LOD vs medical interventions, Outcome 6 Testosterone and free androgen index.

Comparison 1 LOD vs medical interventions, Outcome 7 Fasting Glucose:Insulin.
Figures and Tables -
Analysis 1.7

Comparison 1 LOD vs medical interventions, Outcome 7 Fasting Glucose:Insulin.

Comparison 2 LOD vs other surgical interventions, Outcome 1 Menstrual regularity.
Figures and Tables -
Analysis 2.1

Comparison 2 LOD vs other surgical interventions, Outcome 1 Menstrual regularity.

Comparison 2 LOD vs other surgical interventions, Outcome 2 Improvement in androgenic symptoms (hirsutism/acne).
Figures and Tables -
Analysis 2.2

Comparison 2 LOD vs other surgical interventions, Outcome 2 Improvement in androgenic symptoms (hirsutism/acne).

Comparison 2 LOD vs other surgical interventions, Outcome 3 Harms: Adhesions.
Figures and Tables -
Analysis 2.3

Comparison 2 LOD vs other surgical interventions, Outcome 3 Harms: Adhesions.

Comparison 2 LOD vs other surgical interventions, Outcome 4 Testosterone and free androgen index.
Figures and Tables -
Analysis 2.4

Comparison 2 LOD vs other surgical interventions, Outcome 4 Testosterone and free androgen index.

Comparison 3 LOD 4 ‐ 5 vs 2 or fewer punctures, Outcome 1 Menstrual regularity.
Figures and Tables -
Analysis 3.1

Comparison 3 LOD 4 ‐ 5 vs 2 or fewer punctures, Outcome 1 Menstrual regularity.

Comparison 3 LOD 4 ‐ 5 vs 2 or fewer punctures, Outcome 2 Testosterone and free androgen index.
Figures and Tables -
Analysis 3.2

Comparison 3 LOD 4 ‐ 5 vs 2 or fewer punctures, Outcome 2 Testosterone and free androgen index.

Comparison 4 LOD vs LOD variable energy, Outcome 1 Menstrual regularity.
Figures and Tables -
Analysis 4.1

Comparison 4 LOD vs LOD variable energy, Outcome 1 Menstrual regularity.

Comparison 4 LOD vs LOD variable energy, Outcome 2 Harms.
Figures and Tables -
Analysis 4.2

Comparison 4 LOD vs LOD variable energy, Outcome 2 Harms.

Comparison 4 LOD vs LOD variable energy, Outcome 3 BMI.
Figures and Tables -
Analysis 4.3

Comparison 4 LOD vs LOD variable energy, Outcome 3 BMI.

Comparison 4 LOD vs LOD variable energy, Outcome 4 Testosterone and free androgen index.
Figures and Tables -
Analysis 4.4

Comparison 4 LOD vs LOD variable energy, Outcome 4 Testosterone and free androgen index.

Comparison 4 LOD vs LOD variable energy, Outcome 5 Metabolic measures.
Figures and Tables -
Analysis 4.5

Comparison 4 LOD vs LOD variable energy, Outcome 5 Metabolic measures.

Summary of findings for the main comparison. LOD compared to medical interventions for symptom relief in women with polycystic ovary syndrome

LOD compared to medical interventions for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: Laparoscopic ovarian drilling (LOD)
Comparison: medical interventions

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD

With LOD

Difference

Menstrual regularity at 6 months

LOD vs metformin
N of participants: 236 (2 RCTs)

LOD vs metformin + clomiphene
N of participants: 332 (2 RCTs)

LOD vs gonadotropins
N of participants: 35 (1 RCT)

LOD vs letrozole
N of participants: 260 (1 RCT)

LOD vs metformin + letrozole
N of participants: 146 (1 RCT)

Findings inconsistent and data unsuitable for pooling

Not calculable

⊕⊝⊝⊝
VERY LOW 1, 2, 3

OR 1.02
(0.64 to 1.64)

70.6%

71.0%
(60.5 to 79.7)

0.4% more
(10 fewer to 9.2 more)

⊕⊕⊝⊝
LOW 1, 3

OR 19.20
(3.17 to 116.45)

11.1%

70.6%
(28.4 to 93.6)

59.5% more
(17.3 more to 82.5 more)

⊕⊕⊝⊝
LOW 1, 3

OR 1.08
(0.64 to 1.84)

68.8%

70.4%
(58.5 to 80.2)

1.6% more
(10.3 fewer to 11.4 more)

⊕⊕⊝⊝
LOW 1, 3

OR 0.95
(0.49 to 1.81)

52.1%

50.8%
(34.7 to 66.3)

1.3% fewer
(17.3 fewer to 14.2 more)

⊕⊕⊝⊝
LOW 1, 3

Improvement in androgenic symptoms at 6 months (hirsutism/acne) ‐ LOD vs metformin
N of participants: 126
(1 RCT)

OR 1.00
(0.42 to 2.37)

79.4%

79.4%
(61.8 to 90.1)

0.0% fewer
(17.6 fewer to 10.7 more)

⊕⊕⊝⊝
LOW 1, 3

Improvement in androgenic symptoms at 6 months (hirsutism/acne) ‐ LOD vs gonadotrophins
N of participants: 50 (1 RCT)

Acne: OR 3.20 (0.33 to 30.94)

Hirsutism: OR 2.31, (0.22 to 23.89)

See comments

⊕⊝⊝⊝VERY LOW 1, 4

Acne: 4/29 without LOD, 1/21 with LOD

Hirsutism: 3/29 without LOD, 1/21 with LOD

Harms: GI Upset at 6 months ‐ LOD vs metformin + clomiphene
N of participants: 332 (2 RCTs)

Harms: OHSS rates at 6 months ‐ LOD vs gonadotrophins
N of participants: 33 (1 RCT)

OR 0.05
(0.01 to 0.36)

10.4%

0.6%
(0.1 to 4.0)

9.9% fewer
(10.3 fewer to 6.4 fewer)

⊕⊕⊕⊝
MODERATE 1

OR 0.08
(0.00 to 1.61)

25.0%

2.6%
(0.0 to 34.9)

22.4% fewer
(25 fewer to 9.9 more)

⊕⊕⊝⊝
LOW 1, 3

*The risk in the intervention group (and its 95% confidence interval) is based on the mean risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio; OHSS: ovarian hyperstimulation syndrome

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded, and in some cases methods of randomization unclear.
2Downgraded two levels for very serious and unexplained heterogeneity: I2 = 85%, direction of effect inconsistent
3Downgraded one level for serious imprecision.
4Downgraded two levels for very serious imprecision: Broad confidence interval, very few events.

Figures and Tables -
Summary of findings for the main comparison. LOD compared to medical interventions for symptom relief in women with polycystic ovary syndrome
Summary of findings 2. LOD compared to other surgical interventions for symptom relief in women with polycystic ovary syndrome

LOD compared to other surgical interventions for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: LOD
Comparison: other surgical interventions

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD

With LOD

Difference

Menstrual regularity ‐ LOD vs unilateral LOD
N of participants: 104
(2 RCTs)

OR 1.51
(0.62 to 3.71)

71.2%

78.8%
(60.5 to 90.1)

7.7% more
(10.7 fewer to 19 more)

⊕⊕⊕⊝
MODERATE 1

1 study follow‐up at 3 months. 1 study follow‐up at 12 months

Menstrual regularity at 6 months ‐ LOD vs ultrasound‐guided transvaginal ovarian drilling
N of participants: 147
(1 RCT)

OR 1.23
(0.64 to 2.37)

54.7%

59.7%
(43.6 to 74.1)

5.1% more
(11.1 fewer to 19.4 more)

⊕⊕⊝⊝
LOW 1 2

Menstrual regularity at 12 months
Laser LOD vs harmonic scalpel
N of participants: 34
(1 RCT)

OR 2.13
(0.17 to 26.03)

88.2%

94.1%
(56.0 to 99.5)

5.9% more
(32.2 fewer to 11.3 more)

⊕⊕⊝⊝
LOW 1, 3

Note control group is NdYAG Laser

Improvement in androgenic symptoms at 6 months (Acne) ‐ LOD vs USS guided
N of participants:31
(1 RCT)

OR 0.84 (0.20 to 3.5)

47.1%

42.7%
(15.1 to 75.7)

4.3% fewer
(32 fewer to 28.6 more)

⊕⊕⊝⊝
LOW 1, 2

Improvement in androgenic symptoms at 6 months (Hirsutism) ‐ LOD vs USS‐guided
N of participants: 39
(1 RCT)

OR 1.09
(0.30 to 3.91)

40.0%

42.1% (16.7 to 72.3)

2.1% more (23.3 fewer to 32.3 more)

⊕⊕⊝⊝
LOW 1, 2

Harms: Adhesions at 6 months ‐ LOD vs THL
N of participants: 246
(1 RCT)

OR 0.10
(0.05 to 0.18)

59.3%

12.7%
(6.8 to 20.8)

46.6% fewer
(52.5 fewer to 38.5 fewer)

⊕⊝⊝⊝
VERY LOW 1, 4

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio; THL: transvaginal hydrolaparoscopy

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded or unclear allocation concealment or unclear randomization method.
2Downgraded one level for serious imprecision.
3Single study, narrow confidence interval.
4Downgraded two levels for very serious imprecision: Broad confidence interval, very few events.

Figures and Tables -
Summary of findings 2. LOD compared to other surgical interventions for symptom relief in women with polycystic ovary syndrome
Summary of findings 3. LOD 4‐5 compared to 2 or fewer punctures for symptom relief in women with polycystic ovary syndrome

LOD 4 ‐ 5 punctures compared to 2 or fewer punctures for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: LOD 4 ‐ 5 punctures
Comparison: 2 or fewer punctures

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD 4‐5

With LOD 4‐5

Difference

Menstrual regularity at 6 months ‐ LOD 4 ‐ 5 coagulation points compared to 2 or fewer
N of participants: 73
(2 RCTs)

OR 16.04
(4.19 to 61.34)

13.9%

72.1%
(40.3 to 90.8)

58.2% more
(26.4 more to 76.9 more)

⊕⊕⊝⊝
LOW 1, 2

Menstrual regularity at 6 months ‐ LOD (4 ‐ 5 laser coagulation points) vs 1 laser coagulation point per ovary
N of participants: 40
(1 RCT)

OR 19.00
(2.12 to 170.38)

5.0%

50.0%
(10.0 to 90.0)

45.0% more
(5 more to 85 more)

⊕⊕⊝⊝
LOW 1, 2

Menstrual regularity at 6 months ‐ LOD 4 punctures vs 2 punctures per ovary
N of participants: 33
(1 RCT)

OR 14.00
(2.60 to 75.41)

25.0%

82.4%
(46.4 to 96.2)

57.4% more
(21.4 more to 71.2 more)

⊕⊕⊝⊝
LOW 1, 2

Improvement in androgenic symptoms

No data available

Harms

LOD 4 ‐ 5 versus fewer punctures

No data available

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio;

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded or methods of randomization unclear.
2Downgraded one level for serious imprecision.

Figures and Tables -
Summary of findings 3. LOD 4‐5 compared to 2 or fewer punctures for symptom relief in women with polycystic ovary syndrome
Summary of findings 4. LOD compared to LOD variable energy for symptom relief in women with polycystic ovary syndrome

LOD compared to LOD variable energy for symptom relief in women with polycystic ovary syndrome

Patient or population: Women with symptoms of PCOS
Setting: Clinic or hospital
Intervention: LOD
Comparison: LOD variable energy

Outcomes

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Quality of the evidence
(GRADE)

What happens

Without LOD

With LOD

Difference

Menstrual regularity at 6 months ‐ LOD vs adjusted thermal dose
N of participants: 115
(1 RCT)

OR 0.42
(0.16 to 1.14)

87.9%

75.4%
(53.8 to 89.3)

12.6% fewer
(34.1 fewer to 1.3 more)

⊕⊝⊝⊝
VERY LOW 1, 2

Menstrual regularity at 3 months ‐ LOD unipolar vs LOD bipolar
N of participants: 20
(1 RCT)

OR 1.00
(0.05 to 18.57)

90.0%

90.0%
(31.0 to 99.4)

0.0% fewer
(59 fewer to 9.4 more)

⊕⊝⊝⊝
VERY LOW 1, 2

Groups had different metabolic characteristics at baseline

Improvement in androgenic symptoms

No data available

Harms: Adhesions at 6 months
№ of participants: 64
(1 study)

OR 0.96
(0.32 to 2.88)

28.6%

27.7%
(11.3 to 53.5)

0.8% fewer
(17.2 fewer to 25 more)

⊕⊝⊝⊝

VERY LOW1, 2

Women that remained enrolled for second‐look laparoscopy

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio;

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1Downgraded one level for serious risk of bias: Included studies not double‐blinded or unclear allocation concealment.
2Downgraded two levels for very serious imprecision: Broad confidence interval, very few events.

Figures and Tables -
Summary of findings 4. LOD compared to LOD variable energy for symptom relief in women with polycystic ovary syndrome
Comparison 1. LOD vs medical interventions

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Menstrual regularity Show forest plot

2

Odds Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.1 LOD vs Metformin alone

2

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

2 Menstrual regularity Show forest plot

5

Odds Ratio (M‐H, Fixed, 95% CI)

Subtotals only

2.1 LOD vs Metformin + Clomiphene

2

332

Odds Ratio (M‐H, Fixed, 95% CI)

1.02 [0.64, 1.64]

2.2 LOD vs Gonadotropins

1

35

Odds Ratio (M‐H, Fixed, 95% CI)

19.2 [3.17, 116.45]

2.3 LOD vs Letrozole

1

260

Odds Ratio (M‐H, Fixed, 95% CI)

1.08 [0.64, 1.84]

2.4 LOD vs Metformin + Letrozole

1

146

Odds Ratio (M‐H, Fixed, 95% CI)

0.95 [0.49, 1.81]

3 Improvement in androgenic symptoms (hirsutism/acne) Show forest plot

2

Odds Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.1 LOD vs Metformin

1

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

3.2 LOD vs Gonadotrophins

1

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

4 Harms Show forest plot

3

Odds Ratio (M‐H, Fixed, 95% CI)

Subtotals only

4.1 LOD vs Metformin + Clomiphen

2

332

Odds Ratio (M‐H, Fixed, 95% CI)

0.05 [0.01, 0.36]

4.2 LOD vs Gonadotrophins

1

33

Odds Ratio (M‐H, Fixed, 95% CI)

0.08 [0.00, 1.61]

5 BMI Show forest plot

2

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

5.1 LOD vs Metformin

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

5.2 LOD vs Metformin + Letrozole

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

6 Testosterone and free androgen index Show forest plot

5

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

6.1 LOD vs Metformin

2

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

6.2 LOD vs GnRHa + OCP

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

6.3 LOD vs Metformin + Letrozole

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

6.4 LOD versus Rosiglitazone

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

7 Fasting Glucose:Insulin Show forest plot

3

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

7.1 LOD vs Rosiglitazone

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

7.2 LOD vs Metformin

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

7.3 LOD vs Metformin + Letrozole

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

Figures and Tables -
Comparison 1. LOD vs medical interventions
Comparison 2. LOD vs other surgical interventions

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Menstrual regularity Show forest plot

4

Odds Ratio (M‐H, Fixed, 95% CI)

Subtotals only

1.1 LOD vs Unilateral LOD

2

104

Odds Ratio (M‐H, Fixed, 95% CI)

1.51 [0.62, 3.71]

1.2 LOD vs Ultrasound guided transvaginal ovarian drilling

1

147

Odds Ratio (M‐H, Fixed, 95% CI)

1.23 [0.64, 2.37]

1.3 Laser LOD vs Harmonic Scalpel

1

34

Odds Ratio (M‐H, Fixed, 95% CI)

2.13 [0.17, 26.03]

2 Improvement in androgenic symptoms (hirsutism/acne) Show forest plot

1

Odds Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.1 Hirsutism

1

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

2.2 Acne

1

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

3 Harms: Adhesions Show forest plot

1

Odds Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4 Testosterone and free androgen index Show forest plot

6

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

4.1 LOD vs Ultrasound guided transvaginal ovarian drilling

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

4.2 LOD versus mini laparoscopy with sedation

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

4.3 LOD vs unilateral LOD

3

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

4.4 Laser LOD vs Harmonic Scalpel

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

Figures and Tables -
Comparison 2. LOD vs other surgical interventions
Comparison 3. LOD 4 ‐ 5 vs 2 or fewer punctures

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Menstrual regularity Show forest plot

2

73

Odds Ratio (M‐H, Fixed, 95% CI)

16.04 [4.19, 61.34]

1.1 LOD (4‐5 laser coagulation points) vs 1 laser coagulation point per ovary

1

40

Odds Ratio (M‐H, Fixed, 95% CI)

19.0 [2.12, 170.38]

1.2 LOD 4 punctures vs 2 punctures per ovary LOD

1

33

Odds Ratio (M‐H, Fixed, 95% CI)

14.0 [2.60, 75.41]

2 Testosterone and free androgen index Show forest plot

2

Mean Difference (IV, Fixed, 95% CI)

Subtotals only

2.1 LOD 4‐5 punctures vs 2 or fewer punctures (Testosterone)

2

73

Mean Difference (IV, Fixed, 95% CI)

‐0.90 [‐1.12, ‐0.68]

2.2 LOD 4 punctures vs 2 punctures (FAI)

1

33

Mean Difference (IV, Fixed, 95% CI)

‐1.5 [‐3.21, 0.21]

Figures and Tables -
Comparison 3. LOD 4 ‐ 5 vs 2 or fewer punctures
Comparison 4. LOD vs LOD variable energy

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Menstrual regularity Show forest plot

2

Odds Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.1 LOD vs Adjusted thermal dose

1

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

1.2 LOD unipolar vs LOD bipolar

1

Odds Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

2 Harms Show forest plot

1

Odds Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3 BMI Show forest plot

1

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

4 Testosterone and free androgen index Show forest plot

2

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

4.1 LOD vs Adjusted thermal dose

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

4.2 LOD unipolar vs LOD bipolar

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

5 Metabolic measures Show forest plot

1

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

5.1 LOD (unipolar) vs bipolar

1

Mean Difference (IV, Fixed, 95% CI)

0.0 [0.0, 0.0]

Figures and Tables -
Comparison 4. LOD vs LOD variable energy