Pain relief in office gynaecology: a systematic review and meta-analysis

European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

Contents lists available at ScienceDirect

European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb

Review

Pain relief in office gynaecology: a systematic review and meta-analysis Gaity Ahmad a, Shatha Attarbashi b, Helena O’Flynn c, Andrew J.S. Watson b,* a

Department of Obstetrics and Gynaecology, Pennine Acute Trust, Greater Manchester, UK Department of Obstetrics and Gynaecology, Tameside General Hospital, Lancashire, UK c University Hospital of South Manchester Foundation Trust, Manchester, UK b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 5 July 2010 Received in revised form 21 October 2010 Accepted 26 November 2010

Hysteroscopy, hysterosalpingography (HSG), sonohysterography and endometrial ablation are increasingly performed in an outpatient setting. The primary reason for failure to complete these procedures is pain. The objective of this review was to compare the effectiveness and safety of different types of pharmacological intervention for pain relief in office gynaecological procedures. A systematic search of medical databases including PubMed, EMBASE, Cochrane Central register of controlled trials, PsychInfo and CINHAL was conducted in 2009. Randomised controlled trials (RCTs) investigating the use of local anaesthetics, opioid analgesics, non-opioid analgesics and intravenous sedation for pain relief during and after hysteroscopy, HSG, sonohysterography and endometrial ablation were reviewed. Secondary outcomes included adverse effects and failure to complete procedures. Where RCTs were not identified, the best available evidence was sought. Each study was assessed against inclusion criterion. Results for each study were expressed as a standardised mean difference (SMD) with 95% confidence intervals and combined for meta-analysis with Revman 5 software. Meta-analysis revealed beneficial effect of the use of local anaesthetics during and within 30 min after hysteroscopy; SMD 0.45 (95% CI 0.73, 0.17) and SMD 0.51 (95% CI 0.81, 0.21) respectively. No beneficial effect was noted during HSG. One RCT found evidence of benefit for pain relief during hysterosalpingo-contrastsonography; SMD 1.04 [95% CI 1.44, 0.63]. There was no significant difference in failure to complete hysteroscopy due to cervical stenosis between the intervention and control groups (OR 1.31 (95% CI 0.66, 2.59)), but the incidence of failure to complete the procedure due to pain was significantly less in the intervention group (OR 0.29 (0.12, 0.69)). There is evidence of benefit for the use of local anaesthetics for outpatient hysteroscopy and hysterosalpingo-contrastsonography. Local anaesthetics may be considered when performing hysteroscopy in postmenopausal women to reduce the failure rate. ß 2011 Published by Elsevier Ireland Ltd.

Key words: Office gynaecology Local anaesthesia Anaesthetics Opioid

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Search methods for identification of studies . 2.2. Data collection . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Excluded studies . . . . . . . . . . . . . . . . . . . . . . . 3.2. Pain relief for hysteroscopy . . . . . . . . . . . . . . 3.3. Pain relief for hysterosalpingography . . . . . . 3.4. Pain relief for sonohysterography . . . . . . . . . 3.5. Pain relief for endometrial ablation . . . . . . . . 3.6. Secondary outcomes . . . . . . . . . . . . . . . . . . . . 3.6.1. Adverse effects. . . . . . . . . . . . . . . . . 3.6.2. Failure to complete the procedure .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

* Corresponding author. E-mail address: [email protected] (A.J.S. Watson). 0301-2115/$ – see front matter ß 2011 Published by Elsevier Ireland Ltd. doi:10.1016/j.ejogrb.2010.11.018

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . .

4 4 4 4 4 4 6 6 6 6 8 10 10 10

4

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

4. 5.

Discussion . . . . . . . Conclusions . . . . . . Acknowledgements References . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

1. Introduction A number of gynaecological procedures, including hysteroscopy, endometrial ablation (thermal balloon technique), hysterosalpingography (HSG), hysterosalpingo-contrast sonography (sonohysterography) and insertion of intrauterine contraceptive device (IUCD) are increasingly being performed in an office setting. This is beneficial to both healthcare providers and patients, with avoidance of anaesthetic complications, shorter recovery time, patient preference and decreased costs [1–3]. There is no consensus in the literature as to the optimal method of pain relief for office gynaecological procedures. We are unaware of published guidelines on the routine use of analgesia during office hysteroscopy, HSG, sonohysterography or endometrial ablation. A recent national survey of gynaecologists in the United Kingdom reported wide variation in clinical practice in the use of analgesia for HSG. It is likely that this variation is replicated in other office gynaecology procedures [4]. In the absence of clear guidance and conflicting results from randomised controlled trials (RCTs) in terms of use of analgesics, practice remains varied and optimal analgesia for office gynaecological procedures will not be achieved. There are several causes of pain during outpatient gynaecological procedures, the first being cervical manipulation. Pain from the cervix and vagina is conducted by visceral afferent fibres to the S2 to S4 spinal ganglia via the pudendal and pelvic splanchnic nerves, along with parasympathetic fibres [5]. Pain from intraperitoneal structures, such as the uterine body, is conducted by visceral afferent fibres with sympathetic fibres via the hypogastric nerves to the T12 to L2 spinal ganglia [5]. As the cervix and uterus are insensitive to heat or fine touch stimuli, dilatation of cervix and uterine distension account for the pain [6]. There may also be further delayed pain caused by the release of prostaglandins. The aim of the review was to establish the most effective method of pain relief for the various office gynaecological procedures. 2. Methods A systematic review was undertaken with a prospective protocol designed using Cochrane methodology for systematic reviews. Only RCTs comparing different pharmacological interventions for pain relief during and after outpatient gynaecological procedures were included. Where RCTs were not identified, the best available evidence was sought. Trials comparing intravenous analgesics and general anaesthetic were excluded. Primary outcome measures include pain score during and after the procedure. This was further sub-grouped into pain within or more than 30 min after the procedure. If a study reported a mean pain score on two separate occasions within the same group, the earliest mean pain score was used. Secondary outcomes include the rate of completion of the procedure and adverse effects.

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

10 12 12 12

choice, thermal ablation, microwave, intracervical, topical, intrauterine analgesics, opioid, narcotic, lignocaine, morphine, benzodiazepines and pain relief. Electronic databases were searched using Ovid software: MEDLINE (1950 to July 2009), EMBASE (1980 to July 2009), CINAHL (1982 to July 2009), PsychINFO (1806 to July 2009). The search was not restricted by language. A copy of the search strategy is available. 2.2. Data collection Data were extracted by three independent authors (GA, HO’F and SA) using a data extraction form which can be obtained from the authors on request. Disagreements were resolved by AW. Studies were selected in a two-stage process. First, all abstracts or titles in the electronic searches were critiqued and full manuscripts of potentially eligible citations were obtained. Second, studies meeting our predefined criteria were selected, as outlined in Fig. 1. Trials were assessed for the following: source, eligibility, methodological details, analgesia interventions (routes of delivery, doses, and timing), descriptive data of participants (age, comorbidities), outcomes and funding sources. Data were managed within a Microsoft Access database, in the form of a data collection form, before input into RevMan data software [7]. 2.3. Data analysis The Cochrane Collaboration’s tool for assessing risk of bias was used to assess the risk of bias for each study [8]. Results for each study were expressed as weighted mean difference with 95% confidence intervals (CI) unless differing validated scales were reported, in which case a standard mean difference (SMD) was used and combined for meta-analysis with RevMan software. When interpreting the results of the comparison between an intervention and a placebo, a negative SMD indicates that an intervention is superior to a placebo in terms of ability to reduce levels of pain and a positive SMD indicates that a placebo is superior to an intervention. The Chi-squared distribution test (x2) was performed and the I-squared statistic (I2) was calculated to determine whether there is significant heterogeneity. Adverse effects were reported as odds ratios, with a fixed-effect model. Where significant heterogeneity was identified, the causes were explored. With significant heterogeneity, a sensitivity analysis was performed using the random-effects model. The subgroups analyzed were: (a) Opioid versus placebo or no treatment; (b) Non-opioid analgesics including local anaesthetics versus placebo or no treatment; (c) Opioid analgesics versus non-opioid analgesics; (d) Other comparisons within classes. Sensitivity analysis using a fixed-effects model in comparison to a random-effects model was carried out in the meta-analysis.

2.1. Search methods for identification of studies 3. Results The search strategy was designed by the Cochrane Menstrual Disorders and Subfertility Group. The Menstrual Disorders and Subfertility Group Specialised Register of Controlled Trials was searched for any trials with the following keywords: gynaecologic surgical procedures, hysteroscopy, ablation, cryoablation, therma-

Twenty-five studies which met the inclusion criteria were included. Fifteen of the included studies investigated the use of analgesics during hysteroscopy, [9–23] nine studies during HSG [24–31] and one study during sonohysterography [32]. There was

[()TD$FIG]

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

5

Fig. 1. Trials identification and selection.

[()TD$FIG] Control Std. Mean Difference Analgesic type Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI 1.1.1 Opioid versus placebo or no treatment Lin 2005 Subtotal (95% CI)

3.3

1.1

80 80

3.2

1.3

84 100.0% 84 100.0%

Std. Mean Difference IV, Random, 95% CI

0.08 [-0.22, 0.39] 0.08 [-0.22, 0.39]

Heterogeneity: Not applicable Test for overall effect: Z = 0.53 (P = 0.60) 1.1.2 Local Anaesthetics versus placebo or no comparison Cicinelli 1998 Soriano 2000 Esteve 2002 Cicinelli 1997 Lau 1999 Vercellini 1994 Lau 2000 Costello 1998 Wong AYK 2000 Subtotal (95% CI)

1.55 1.38 2.2 1.9 2.6 2.2 9.22 3.56 2.4 2.4 2 4.5 3 3.1 3.1 2.3 1.57 0.77

36 6.66 3.94 36 62 3.7 2.5 56 34 4.3 3 28 40 11.32 3.75 40 49 3.3 3.3 50 87 4.9 2.2 90 45 3.6 3.3 44 49 3.4 2.6 50 250 1.58 0.8 250 652 644

9.1% 11.5% 9.5% 10.4% 11.1% 12.5% 10.8% 11.1% 13.9% 100.0%

-1.71 [-2.26, -1.17] -0.68 [-1.05, -0.30] -0.65 [-1.16, -0.13] -0.57 [-1.02, -0.12] -0.31 [-0.71, 0.09] -0.19 [-0.48, 0.11] -0.16 [-0.57, 0.26] -0.12 [-0.52, 0.27] -0.01 [-0.19, 0.16] -0.45 [-0.73, -0.17]

Heterogeneity: Tau² = 0.15; Chi² = 44.78, df = 8 (P < 0.00001); I² = 82% Test for overall effect: Z = 3.10 (P = 0.002) 1.1.3 Oral NSAIDs versus placebo or no comparison Tam 2001 Subtotal (95% CI)

2.2

2.1

92 92

2.1

2

89 100.0% 89 100.0%

0.05 [-0.24, 0.34] 0.05 [-0.24, 0.34]

Heterogeneity: Not applicable Test for overall effect: Z = 0.33 (P = 0.74)

-2 -1 0 1 2 Favours experimentalFavours control Fig. 2. Any analgesic versus placebo or no treatment – mean pain score during the procedure – hysteroscopy. Copyright Cochrane Collaboration, reproduced with permission [48].

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

6

no study that met the inclusion criteria for the use of analgesia during endometrial ablation; therefore the best available evidence was sought [6,33–36]. Details of the included studies are shown in Tables S1–3 (Copyright Cochrane Collaboration, reproduced with permission) [48,49]. Risk of bias of individual studies is outlined in Table S4 (Copyright Cochrane Collaboration, reproduced with permission) [48,49]. Sensitivity analysis for all results using a fixed effects model in comparison to a random effects model demonstrated no change in the significance of results. 3.1. Excluded studies A total of twenty studies were excluded. Six studies were excluded because either general anaesthetic or an intravenous (IV) analgesic was used as an intervention. Twelve studies were excluded due to the method of presentation of results; mean pain scores were either not reported, or grouped into dichotomous data, pain scores were provided in percentages, mean pain scores were only presented graphically, or no standard deviations were given. One study was excluded as it was not randomised and one study was excluded as the intervention investigated was non pharmacological. 3.2. Pain relief for hysteroscopy No evidence of benefit was demonstrated for the use of oral opioids (SMD 0.08 (95% CI 0.22 to 0.39)) during hysteroscopy. No evidence of benefit was demonstrated for oral non-steroidal antiinflammatory drugs (NSAIDs) during or more than 30 min after hysteroscopy (SMD 0.05 (95% CI 0.24 to 0.34)) and (SMD 0.07 (95% CI 0.36 to 0.23)) shown in Figs. 2 and 4. Evidence of an overall beneficial effect was demonstrated for [()TD$FIG] use of local anaesthetics during and within 30 min after the

hysteroscopy (SMD 0.45 ((95% CI 0.73, 0.17) I2 = 82%) and (SMD 0.51 (95% CI 0.81, 0.21 I2 = 54%) respectively as shown in Figs. 2 and 3. No significant beneficial effect with the use of local anaesthetics was identified after 30 min, however (SMD 0.08 (95% CI 0.33, to 0.17) I2 = 0%) as shown in Fig. 4. Evidence of a significant reduction in the mean pain score more than 30 min after hysteroscopy was demonstrated (SMD 0.34 [95% CI 0.06, 0.61] I2 = 0%) with the use of IV sedation in comparison to paracervical block (Fig. 5). No significant evidence of benefit was demonstrated during hysteroscopy, however (SMD 0.26 [95% CI 0.01, 0.54] I2 = 0%). 3.3. Pain relief for hysterosalpingography No significant evidence of a beneficial effect during, within 30 min or more than 30 min after HSG with the use of oral NSAIDs was demonstrated; (SMD 0.21 (95% CI 0.89, 0.48) I2 = 62%), (SMD 1.38 (95% CI 4.79, 2.02) I2 = 94%) and (SMD 0.07 (95% CI 0.51, 0.37) I2 = 0%) respectively as shown in Figs. 6, 7 and 8. No significant evidence of a beneficial effect during or within 30 min after HSG with the use of local anaesthetics was demonstrated (SMD 0.04 (95% CI 0.24, 0.15) I2 = 0%) and (SMD 1.01 (95% CI 0.26, 2.29) I2 = 96%)) respectively (Figs. 6 and 8). Evidence of a significant reduction in the mean pain score more than 30 min after HSG with the use of local anaesthetics was identified (SMD 2.40 (95% CI 3.04, 1.76)) as shown in Fig. 8. 3.4. Pain relief for sonohysterography One RCT was identified which compared an intrauterine lidocaine infusion during saline solution infusion for sonohysterography with placebo [32]. Analysis revealed that intrauterine

Control Std. Mean Difference Analgesic type Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI 1.2.1 Opioid analgesic versus placebo or no treatment Subtotal (95% CI) 0 0 Not estimable

Std. Mean Difference IV, Random, 95% CI

Heterogeneity: Not applicable Test for overall effect: Not applicable 1.2.2 Local Anaesthetics versus placebo or no comparison Cicinelli 1997 Cicinelli 1998 Esteve 2002 Giorda 2000 Subtotal (95% CI)

2.67 2.66 0.55 0.93 1.7 1.7 5.3 1.1

40 36 34 121 231

3.87 2.47 1.8 6.3

3.13 2.7 1.7 2.18

40 36 28 119 223

23.4% 21.1% 20.6% 34.9% 100.0%

-0.41 [-0.85, 0.03] -0.94 [-1.43, -0.45] -0.06 [-0.56, 0.44] -0.58 [-0.84, -0.32] -0.51 [-0.81, -0.21]

Heterogeneity: Tau² = 0.05; Chi² = 6.56, df = 3 (P = 0.09); I² = 54% Test for overall effect: Z = 3.29 (P = 0.0010) 1.2.3 Oral NSAIDs versus placebo or no comparison Subtotal (95% CI) 0 0

Not estimable

Heterogeneity: Not applicable Test for overall effect: Not applicable 1.2.4 Other comparisons within classes Subtotal (95% CI) 0

0

Not estimable

Heterogeneity: Not applicable Test for overall effect: Not applicable Total (95% CI)

231

223 100.0%

Heterogeneity: Tau² = 0.05; Chi² = 6.56, df = 3 (P = 0.09); I² = 54% Test for overall effect: Z = 3.29 (P = 0.0010)

-0.51 [-0.81, -0.21] -1 -0.5 0 0.5 1 Favours experimentalFavours control

Fig. 3. Any analgesic versus placebo or no treatment (hysteroscopy) – mean pain score within 30 min of the procedure. Copyright Cochrane Collaboration, reproduced with permission [48].

[()TD$FIG]

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

Experimental Control Std. Mean Difference IV, Random, 95% CI SD Total Mean SD Total Weight Mean

Study or Subgroup

1.3.1 Opioid analgesic versus placebo or no treatment 0 0 Subtotal (95% CI)

7

Std. Mean Difference IV, Random, 95% CI

Not estimable

Heterogeneity: Not applicable Test for overall effect: Not applicable 1.3.3 Non-opioid analgesics (Local Anaesthetics) versus placebo or no comparison -0.14 [-0.64, 0.36] 0.9 1.2 34 1.1 1.6 28 14.3% Esteve 2002 -0.22 [-0.61, 0.18] 1.4 1.6 49 1.8 2 50 22.9% Lau 1999 0.12 [-0.29, 0.54] 1.4 1.8 45 1.2 1.4 44 20.7% Lau 2000 -0.08 [-0.33, 0.17] 57.9% 122 128 Subtotal (95% CI) Heterogeneity: Tau² = 0.00; Chi² = 1.45, df = 2 (P = 0.49); I² = 0% Test for overall effect: Z = 0.61 (P = 0.54) 1.3.4 Non-opioid analgesics (Oral NSAIDs) versus placebo or no comparison -0.07 [-0.36, 0.23] 1.2 1.7 89 42.1% Tam 2001 1.1 1.3 92 -0.07 [-0.36, 0.23] 89 42.1% Subtotal (95% CI) 92 Heterogeneity: Not applicable Test for overall effect: Z = 0.44 (P = 0.66) 1.3.5 Other comparisons within classes 0 Subtotal (95% CI)

Not estimable

0

Heterogeneity: Not applicable Test for overall effect: Not applicable 211

220

Total (95% CI)

100.0%

-0.07 [-0.26, 0.12]

Heterogeneity: Tau² = 0.00; Chi² = 1.45, df = 3 (P = 0.69); I² = 0% Test for overall effect: Z = 0.75 (P = 0.45) Test for subgroup differences: Chi² = 0.00, df = 1 (P = 0.95), I² = 0%

-1 -0.5 0 0.5 1 Favours experimental Favours control

Fig. 4. Any analgesic versus placebo or no treatment more than 30 min after the procedure – hysteroscopy. Copyright Cochrane Collaboration, reproduced with permission [48].

[()TD$FIG] Experimental Control Std. Mean Difference Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI 2.3.1 Local intracervical anaesthesia compared to combined intracervical and paracervical anaesthesia. Al-Sunaidi 2007 Subtotal (95% CI)

1

0.2

42 42

1.7

0.2

42 42

19.7% 19.7%

Std. Mean Difference IV, Random, 95% CI

-3.47 [-4.15, -2.78] -3.47 [-4.15, -2.78]

Heterogeneity: Not applicable Test for overall effect: Z = 9.90 (P < 0.00001) 2.3.2 Local paracervical anaesthesia versus conscious sedation Guida 2003 Sharma 2009 Subtotal (95% CI)

1.5 1.1 2.53 0.81

82 20 102

1.2 0.6 2.23 0.94

84 20 104

20.7% 19.9% 40.6%

0.34 [0.03, 0.64] 0.34 [-0.29, 0.96] 0.34 [0.06, 0.61]

Heterogeneity: Tau² = 0.00; Chi² = 0.00, df = 1 (P = 0.99); I² = 0% Test for overall effect: Z = 2.40 (P = 0.02) 2.3.3 Anti-spasmodic/ NSAID versus local paracervical anaesthesia Sharma 2009 Subtotal (95% CI)

1.78 0.89

20 20

2.53 0.81

20 20

19.8% 19.8%

-0.86 [-1.51, -0.21] -0.86 [-1.51, -0.21]

20 20

19.9% 19.9%

-0.48 [-1.11, 0.15] -0.48 [-1.11, 0.15]

186 100.0%

-0.81 [-2.07, 0.44]

Heterogeneity: Not applicable Test for overall effect: Z = 2.60 (P = 0.009) 2.3.4 Anti spasmodic/ NSAID versus IV sedation Sharma 2009 Subtotal (95% CI)

1.78 0.89

20 20

2.23 0.94

Heterogeneity: Not applicable Test for overall effect: Z = 1.50 (P = 0.13) Total (95% CI)

184

Heterogeneity: Tau² = 1.96; Chi² = 105.60, df = 4 (P < 0.00001); I² = 96% Test for overall effect: Z = 1.27 (P = 0.20)

-4 -2 0 2 4 Favours experimental Favours control

Fig. 5. Any analgesic versus any other analgesic more than 30 min after the procedure (hysteroscopy). Copyright Cochrane Collaboration, reproduced with permission [48].

[()TD$FIG]

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

8

Treatment Control Std. Mean Difference Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI 1.1.1 Non-Opoiod Analgesic versus Placebo or No Treatment Elson, 2000 Owens 2, 1985 Owens, 1985 Subtotal (95% CI)

47.9 31.7 1.47 0.24 1.07 0.28

39 15 15 69

16.6% 3.9% 3.1% 23.6%

0.02 [-0.40, 0.44] 0.26 [-0.60, 1.12] -1.12 [-2.09, -0.15] -0.21 [-0.89, 0.48]

21.0% 24.2% 16.4% 14.8% 76.4%

-0.05 [-0.43, 0.32] 0.00 [-0.35, 0.35] -0.07 [-0.49, 0.36] -0.07 [-0.52, 0.37] -0.04 [-0.24, 0.15]

265 100.0%

-0.05 [-0.23, 0.12]

49 8 7 64

47.2 33.8 1.4 0.29 1.4 0.29

Std. Mean Difference IV, Random, 95% CI

Heterogeneity: Tau² = 0.23; Chi² = 5.26, df = 2 (P = 0.07); I² = 62% Test for overall effect: Z = 0.59 (P = 0.56) 1.1.2 Topical Analgesic versus Placebo or No Treatment Costello, 2002 Frishman, 2004 Kafali, 2003 Liberty 2007 Subtotal (95% CI)

5.3 0.4 1.4 2.7

7.3 5.3 4.1 2.8

55 63 41 41 200

7.6 5.3 4.2 3

5.8 0.4 1.5 2.6

55 64 45 37 201

Heterogeneity: Tau² = 0.00; Chi² = 0.09, df = 3 (P = 0.99); I² = 0% Test for overall effect: Z = 0.44 (P = 0.66) 269

Total (95% CI)

Heterogeneity: Tau² = 0.00; Chi² = 5.41, df = 6 (P = 0.49); I² = 0% Test for overall effect: Z = 0.63 (P = 0.53)

-1 -0.5 0 0.5 1 Favours Treatment Favours Control

Fig. 6. Any analgesic versus placebo or no treatment – (HSG) – mean pain score during the procedure. Copyright Cochrane Collaboration, reproduced with permission [49].

lidocaine infusion was significantly beneficial during sonohysterography (SMD of 1.04 (95% CI 1.44, 0.63)) and within 30 min after the procedure (SMD of 1.10 (95% CI 1.51, 0.69)). 3.5. Pain relief for endometrial ablation There were no RCTs identified comparing the effectiveness of different types of pain relief used in outpatient endometrial [()TD$FIG]

ablative techniques. Therefore for this group the best available evidence was reviewed. This included five observational studies which evaluated the feasibility of outpatient endometrial ablation rather than comparison of analgesics [6,33–36]. Various ablative devices used were: HydroThermAblator [6], Thermablate EA system [36], Thermachoice III [33–35]. The analgesic regimen used in most studies was a combination of oral/rectal NSAIDs and intracervical/paracervical local anaesthetic;

Experimental Control Std. Mean Difference SD Total Mean IV, Random, 95% CI Mean SD Total Weight Study or Subgroup 3.2.1 Opioid analgesic versus placebo or no treatment Not estimable 0 0 Subtotal (95% CI)

Std. Mean Difference IV, Random, 95% CI

Heterogeneity: Not applicable Test for overall effect: Not applicable 3.2.2 Non-opioid analgesic (Local Anaesthetics) versus placebo or no comparison Costello 2002 Frishman, 2004 Kafali 2003 Subtotal (95% CI)

3.4 1.8 4

4.5 0.3 1.4

55 63 41 159

1.2 1.2 3.9

1.8 0.2 1.6

55 64 45 164

21.3% 21.1% 21.2% 63.7%

0.64 [0.25, 1.02] 2.34 [1.89, 2.80] 0.07 [-0.36, 0.49] 1.01 [-0.26, 2.29]

Heterogeneity: Tau² = 1.22; Chi² = 55.35, df = 2 (P < 0.00001); I² = 96% Test for overall effect: Z = 1.56 (P = 0.12) 3.2.3 Non-opioid analgesic versus placebo or no comparison Owens 2, 1985 Owens, 1985 Subtotal (95% CI)

1.47 0.73

0.24 0.22

15 15 30

1.4 0.16 1.4 0.16

8 7 15

19.5% 16.8% 36.3%

0.31 [-0.55, 1.18] -3.16 [-4.53, -1.79] -1.38 [-4.79, 2.02]

Heterogeneity: Tau² = 5.69; Chi² = 17.73, df = 1 (P < 0.0001); I² = 94% Test for overall effect: Z = 0.80 (P = 0.43) 3.2.4 Other comparisons within classes 0 Subtotal (95% CI)

0

Not estimable

Heterogeneity: Not applicable Test for overall effect: Not applicable Total (95% CI)

189

179

100.0%

Heterogeneity: Tau² = 1.63; Chi² = 89.95, df = 4 (P < 0.00001); I² = 96% Test for overall effect: Z = 0.29 (P = 0.77) Test for subgroup differences: Chi² = 16.87, df = 1 (P < 0.0001), I² = 94.1%

0.17 [-1.00, 1.34] -10 -5 0 5 10 Favours experimental Favours control

Fig. 7. Any analgesic versus placebo or no treatment (HSG) – mean pain score within 30 min of the procedure. Copyright Cochrane Collaboration, reproduced with permission [49].

[()TD$FIG]

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

Study or Subgroup

Experimental Control Std. Mean Difference IV, Random, 95% CI SD Total Weight Mean SD Total Mean

3.3.1 Opioid analgesic versus placebo or no treatment Subtotal (95% CI) 0 0

9

Std. Mean Difference IV, Random, 95% CI

Not estimable

Heterogeneity: Not applicable Test for overall effect: Not applicable 3.3.2 Non-opioid analgesic (Local Anaesthetics) versus placebo or no comparison Kafali 2003 -1.54 [-2.03, -1.06] 41 4.7 1.8 45 26.8% 2.3 1.2 Subtotal (95% CI) 26.8% 41 45 -1.54 [-2.03, -1.06] Heterogeneity: Not applicable Test for overall effect: Z = 6.23 (P < 0.00001) 3.3.3 Non-opioid analgesic versus Elson, 2000 28.4 33.7 Owens 2, 1985 0.27 0.59 Owens, 1985 0.13 0.35 Subtotal (95% CI)

placebo or no comparison 39 15 15 69

24.5 28.8 0.27 0.8 0.27 0.8

49 8 7 64

27.3% 23.2% 22.7% 73.2%

0.12 [-0.30, 0.55] 0.00 [-0.86, 0.86] -0.26 [-1.16, 0.65] 0.05 [-0.30, 0.40]

Heterogeneity: Tau² = 0.00; Chi² = 0.57, df = 2 (P = 0.75); I² = 0% Test for overall effect: Z = 0.26 (P = 0.79) 3.3.4 Other comparisons within classes 0 Subtotal (95% CI)

0

Not estimable

Heterogeneity: Not applicable

109

110

Total (95% CI)

100.0%

-0.44 [-1.37, 0.50]

Heterogeneity: Tau² = 0.78; Chi² = 27.77, df = 3 (P < 0.00001); I² = 89% Test for overall effect: Z = 0.92 (P = 0.36) Test for subgroup differences: Chi² = 27.20, df = 1 (P < 0.00001), I² = 96.3%

-2 -1 0 1 2 Favours experimental Favours control

Fig. 8. Any analgesic versus placebo or no treatment – (HSG) – mean pain score more than 30 min after the procedure. Copyright Cochrane Collaboration, reproduced with permission [49].

[()TD$FIG] Control Experimental Study or Subgroup Events Total Events Total Weight 6.5.1 Opioid analgesic versus palcebo or no comparison 0 0 Subtotal (95% CI)

Odds Ratio M-H, Fixed, 95% CI

Odds Ratio M-H, Fixed, 95% CI

Not estimable

0

0 Total events Heterogeneity: Not applicable Test for overall effect: Not applicable

6.5.2 Non-opioid versus placebo or no comparison (Local Anaesthetics) Giorda 2000 Lau 1999 Lau 2000 Subtotal (95% CI)

7 0 0

121 50 45 216

21 0 1

119 93.1% 50 45 6.9% 214 100.0%

0.29 [0.12, 0.70] Not estimable 0.33 [0.01, 8.22] 0.29 [0.12, 0.69]

22 7 Total events Heterogeneity: Chi² = 0.01, df = 1 (P = 0.94); I² = 0% Test for overall effect: Z = 2.81 (P = 0.005) 6.5.3 Non-opioid versus placebo or no comparison (NSAIDs) Tam 2001 Subtotal (95% CI)

0 0

0

0 Total events Heterogeneity: Not applicable Test for overall effect: Not applicable

0

Total (95% CI)

216

Not estimable Not estimable

0

Not estimable

0

6.5.4 Other comparisons within classes 0 Subtotal (95% CI) 0 Total events Heterogeneity: Not applicable Test for overall effect: Not applicable

0 0

0

214 100.0%

22 7 Total events Heterogeneity: Chi² = 0.01, df = 1 (P = 0.94); I² = 0% Test for overall effect: Z = 2.81 (P = 0.005) Test for subgroup differences: Not applicable

0.29 [0.12, 0.69]

0.01 0.1 1 10 100 Favours experimental Favours control

Fig. 9. Secondary outcome – failure to complete the procedure in hysteroscopy-due to pain. Copyright Cochrane Collaboration, reproduced with permission [48].

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

10

with the details shown in Table S3. The lowest pain scores were noted in one study [35] in which, in addition to the above interventions, patients received intramuscular (IM) ketorolac 60 mg, oral hydrocodone/acetaminophen 5/500 mg, IM promethazine 50 mg IM and oral 1 mg alprazolam. One study used oral ibuprofen only [33]. The authors reported that most women (87%) successfully completed a full 8-min treatment. However; five women (13%) requested the procedure to be stopped due to pain and 36% required rescue analgesia. 3.6. Secondary outcomes 3.6.1. Adverse effects Adverse effects were sub-grouped into three categories; vasovagal reactions, non-pelvic pain and allergic reactions. Eight RCTs investigating hysteroscopy reported adverse effects [9–16]. For all three sub-groups, there was no significant difference demonstrated between the intervention and control groups. Within the vasovagal reactions subgroup, meta-analysis demonstrated an OR of 0.52 [95% CI 0.15, 1.76], for non-pelvic pain an OR of 1.87 [95% CI 0.61, 5.73] and for allergic reactions an OR of 2.93 [95% CI 0.12, 72.99]. Within the HSG group, no studies reported any useable data on adverse effects. The only sonohysterography study reported transient palpitations in two participants in the intervention group and one case of vomiting in the placebo group [32]. There

[()TD$FIG] Experimental Control Study or Subgroup Events Total Events Total Weight 6.4.1 Opioid analgesic versus placebo or no comparison 0 0 Subtotal (95% CI) 0 Total events Heterogeneity: Not applicable Test for overall effect: Not applicable

was no significant difference between the intervention and placebo groups (OR 2.04 (95% CI 0.18, 23.19)). 3.6.2. Failure to complete the procedure Reasons for failure to complete the procedure were subgrouped into cervical stenosis or pain. Within the hysteroscopy group, six RCTs reported failure to complete the procedure due to cervical stenosis [11,12,14,18,19,21]. There was no significant difference between the intervention and control groups (Fig. 9). Two RCTs reported failure to complete procedure due to pain [11,13]. Meta-analysis demonstrated significantly fewer incidents of failure to complete the procedure in the intervention group than in the control group (Fig. 10). 4. Discussion With an increasing emphasis on ambulatory gynaecological procedures, a move towards performing hysteroscopy and endometrial ablation in an outpatient setting has been advocated due to substantial health and economic benefits. Although these procedures are safe, studies of their acceptability have displayed various completion rates, the principal reason for failure being recognised as pain. De Iaco et al. [37] observed that this procedure is painful even when performed by an experienced surgeon using a non-traumatic technique. With the vast potential for its use in the primary care setting, a consensus on the safest and most efficacious

Odds Ratio M-H, Fixed, 95% CI

Odds Ratio M-H, Fixed, 95% CI

Not estimable

0

6.4.2 Non-opioid versus placebo or no comparison (Local Anaesthetics) Costello 1998 Esteve 2002 Giorda 2000 Lau 1999 Lau 2000 Soriano 2000 Vercellini 1994 Subtotal (95% CI)

1 2 12 1 0 1 2

50 36 121 50 45 63 89 454

0 1 11 0 0 1 2

3.2% 50 7.0% 29 119 66.7% 3.2% 50 45 7.0% 56 92 12.8% 441 100.0%

3.06 [0.12, 76.95] 1.65 [0.14, 19.13] 1.08 [0.46, 2.56] 3.06 [0.12, 76.95] Not estimable 0.89 [0.05, 14.52] 1.03 [0.14, 7.51] 1.23 [0.62, 2.43]

15 19 Total events Heterogeneity: Chi² = 0.84, df = 5 (P = 0.97); I² = 0% Test for overall effect: Z = 0.59 (P = 0.55) 6.4.3 Non-opioid analgesic versus placebo or no comparison (Oral NSAIDs) Not estimable 0 0 Subtotal (95% CI) 0 Total events Heterogeneity: Not applicable Test for overall effect: Not applicable

0

6.4.4 Other comparisons within classes 0 Subtotal (95% CI) 0 Total events Heterogeneity: Not applicable Test for overall effect: Not applicable Total (95% CI)

454

0

Not estimable

0

441 100.0%

15 19 Total events Heterogeneity: Chi² = 0.84, df = 5 (P = 0.97); I² = 0% Test for overall effect: Z = 0.59 (P = 0.55) Test for subgroup differences: Not applicable

1.23 [0.62, 2.43]

0.2 0.5 1 2 5 Favours experimental Favours control

Fig. 10. Secondary outcome – failure to complete the procedure in hysteroscopy due to cervical stenosis. Copyright Cochrane Collaboration, reproduced with permission [48].

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

method of pain relief for outpatient hysteroscopy would be advantageous. The purpose of the meta-analysis was therefore to compare the effectiveness and safety of different types of pharmacological intervention for office gynaecological procedure. Effective pain relief from local anaesthetic depends on various factors including; route of administration, concentration, classification of drug and sufficient time interval between the administration of the analgesic and start of the procedure. Various methods of administration of local anaesthesia were investigated and found to be effective, including paracervical block, intracervical, intrauterine, and topical anaesthesia. The three RCTs [10,12,21] using paracervical block as an intervention for pain relief during hysteroscopy reported different time intervals between the administration of anaesthesia and the start of the procedure. In the only RCT [10] showing significant beneficial effect during the procedure, the authors allowed 10 min between administration of the drug and the start of the procedure, while using a higher concentration of drug (Mepivicaine). This was in contrast to the other two studies, which failed to show any benefit using lower concentrations and shorter time intervals. The studies which supported the use of paracervical anaesthesia were restricted to postmenopausal women [10,11]. This group tends to have a higher incidence of cervical atrophy and stenosis, and may experience more pain during the procedure, thereby demonstrating a greater contrast between the placebo and intervention. Deep nerve endings in the myometrium, cervical stroma and visceral peritoneum may be insufficiently blocked at 5 min after administration. This may explain the beneficial effect found in studies allowing a longer time interval between administration of anaesthetic and start of the procedure [29]. Soriano et al., however, although using a topical lignocaine spray, demonstrated effective pain relief within 5 min from instillation of drug to start of the procedure [14]. A number of studies included in the meta-analysis suggested that paracervical block may well decrease the pain caused by cervical manipulation, but is unable to affect the pain arising due to uterine distension [12,21]. Anatomically, sensory innervations of the pelvic organs are from the superior hypogastric plexus or pre sacral nerve, pelvic nerves, ovarian plexus. Sensory fibres from the upper part of the vagina, uterus, proximal portion of the tubes, bladder, urethra and rectum running through the paracervical tissue and within the uterosacral folds meet in the hypogastric and pelvic nerves. Thus, on the basis of the above anatomical observations, paracervical block should block not only cervical but uterine pain [38,39]. Two RCTs failed to demonstrate a beneficial effect of the use of intrauterine local anaesthetic [9,13]. This may be attributed to bypassing the sensory fibres located in the paracervical tissues and uterosacral ligaments and thus being unable to inhibit cervical pain. Lignocaine sprays, gels and creams provide anaesthesia to superficial pain receptors. The advantages of topical methods include painless application and a decreased rate of infection [14]. Two studies included in the meta-analysis investigated their use [14,15]. Soriano et al. [14] displayed significant beneficial results during hysteroscopy with the use of lidocaine spray in comparison to Wong et al., who used lignocaine gel. The procedure, however, was performed immediately after application of the gel [15], and this may not have allowed sufficient time for the effect of the anaesthesia. The use of spray may allow more consistent application under a pressurized environment in comparison to a gel applied with gauze. The only RCT identified investigating the use of an NSAID given 1–2 h prior to procedure did not demonstrate any beneficial effect in terms of pain relief during or after hysteroscopy [16]. Inconsistent time interval between administration of NSAID and

11

start of the procedure introduces bias and does not allow for accurate interpretation of the results. Delayed pain may be attributed to the release of prostaglandins from cervical manipulation and uterine distension; therefore, the use of a prostaglandinsynthetase inhibitor would seem a logical intervention for pain relief after rather than during the procedure. An NSAID administered 2 h prior to hysteroscopy will have its peak analgesic effect before the completion of the procedure, thus having a suboptimal effect on delayed pain. Mean peak plasma concentrations are attained after 20–60 min after one tablet of 50 mg Diclofenac [40]. Although buprenorphine is thirty-five times more potent an opioid than morphine, no benefit was noted when compared with placebo in the only RCT identified [20]. Non-significant effects on pain relief may be due to the opioid being given 40 min before the procedure, when the peak pain relief due to this drug appears to occur between 1 and 2 h after sublingual administration of buprenorphine [41]. A high incidence of nausea and drowsiness limits its use in the outpatient setting. Sharma et al. demonstrated significant reduction in pain during and after the procedure with the use of drotaverine hydrochloride/ mefenamic acid in comparison to both paracervical block and IV sedation [23]. Mefenamic acid is an NSAID which acts as a potent inhibitor of prostaglandin synthesis and has a peak plasma concentration 2–4 h following administration [42]. Drotaverine hydrochloride is an anti-spasmodic agent and selective phosphodiesterase inhibitor with a peak plasma concentration at 1 h [43], which may account for the combined analgesic effect of both during and after hysteroscopy. A wide variation in the bioavailability of drotaverine hydrochloride following oral administration has been demonstrated, however, and individual response may therefore be variable [44]. Meta-analysis did not show any benefit of using any analgesia in preference to a placebo for HSG. These studies, however, examined the role of relatively weak analgesics; any beneficial effect could have been masked by the benefits of a placebo effect [45]. The non-opiates which were analyzed in this study included paracetamol and NSAIDs; fenoprofen and naproxen, administered at various times before the procedure. The effect of paracetamol is mainly mediated peripherally, although it also has some central effects. As paracetamol has no anti-inflammatory properties, it is unlikely to have a significant effect on pain caused by prostaglandin release as a result of uterine distension or cervical instrumentation – two recognized causes of pain during HSG. Therefore NSAIDs, which inhibit prostaglandin release, would seem a logical intervention. Meta-analysis of all the studies involving oral analgesics found they were no more or less effective than a placebo in alleviating pain during HSG. Meta-analysis revealed no preferential effect of topical lignocaine to modulate pain during or after HSG. Overall, meta-analysis revealed beneficial effect of the use of local anaesthetics for hysteroscopy but not for HSG. This effect may be due to the fact that more RCTs were identified investigating the role of local anaesthetics in hysteroscopy. Moreover, there is evidence to suggest that higher intrauterine pressure is required in HSG, resulting in the procedure being more painful than hysteroscopy. The median intrauterine perfusion pressure that will produce spill from the fallopian tubes into the peritoneal cavity is 100 mm Hg, with no spill occurring at pressures less than 70 mm Hg, in comparison to hysteroscopy, where pressures of up to 40 mm Hg are considered adequate to distend the uterus [46,47]. Within the hysteroscopy group, adverse effects were subgrouped into three categories: vasovagal reactions, non-pelvic pain and allergic reactions. Eight RCTs reported adverse effects but meta-analysis failed to demonstrate any significant difference between the analgesic and placebo groups. Cicinelli et al.

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13

12

demonstrated that local anaesthesia reduced the incidence of vasovagal reactions [9,10]. This finding was significant, especially for postmenopausal women, who are more susceptible to vasovagal reactions due to uterine involution and cervical stenosis. In addition, the age-related reduced responsiveness of the cardiovascular system and possible onset of cardiac arrhythmia increase the risk posed by vagal stimulation in these women. In contrast, Lau et al. failed to demonstrate a reduced incidence of vasovagal reaction with paracervical anaesthesia [12]. In fact, this study revealed a significantly higher incidence of symptomatic bradycardia or hypotension in the lignocaine group than in the placebo group. The authors postulated that the side effect was probably the result of inadvertent intravasation of lignocaine rather than a genuine vasovagal reaction. Lau et al. failed to confirm this through the use of transcervical anaesthesia, which avoids intravasation [13]. Transcervical anaesthesia may bypass the cervical plexus at the internal cervical os and thus may fail to demonstrate any beneficial effect of the local anaesthetics in avoiding vasovagal reactions. Reasons for failure to complete the procedure were subgrouped into cervical stenosis or pain. Cervical stenosis is predominately a problem in post-menopausal women [11]. Meta-analysis demonstrated a significantly lower incidence of failure to complete procedure in the intervention group than in the control group. Giorda et al. investigated failure to complete the procedure in postmenopausal women exclusively; all other studies enrolled predominately pre- and peri-menopausal women, in which the use of local anaesthesia and cervical dilatation in routine hysteroscopy is significantly less often required than in postmenopausal women, who may have a degree of cervical stenosis [11]. The use of local anaesthetic thus safely and effectively reduces the pain associated with hysteroscopy. 5. Conclusions There is evidence of benefit for the use of local anaesthetics for outpatient hysteroscopy and sonohysterography but not for HSG. Local anaesthetics may be considered when performing hysteroscopy in postmenopausal women to reduce the failure rate due to pain. Local anaesthetics may also relieve pain during endometrial ablation. Further high quality, well powered trials should be undertaken in order to provide the data necessary to estimate the efficacy of oral analgesics and optimal route of administration and dose of local anaesthetics. Conflict of interest None. Contributions GA: Main review author, designed the protocol; screen the search results, organise the retrieval of the RCTs, screened them against the inclusion criteria, extract the data from the RCTs, manage the data, interpret the results, and supervise HO’F and SA throughout the process. HO’F: Co-review author, helped write the protocol; organised the retrieval of the RCTs, screened them against the inclusion criteria, extracted the data from RCTs, wrote to authors when required, managed the data, and help interpret the results. SA: Co-review author, helped write the protocol; organised the retrieval of the RCTs, screened them against the inclusion criteria, extracted the data from RCTs, wrote to authors when required, managed the data, and help interpret the results. AW: Helped design the review, supervised all the steps undertaken for the review.

Funding None Acknowledgements We would like to thank Cochrane Menstrual Disorders and Fertility Subgroup for their support. Parts of this manuscript have been derived from two Cochrane reviews and were included in the dissertation for MSc in Advanced Gynaecological Endoscopy at Univeristy of Surrey. We would like to thank John Wiley & Sons Ltd for their permission to use published data and also Dr Karen Ballard at the University of Surrey. Copyright Cochrane Collaboration, reproduced with permission [48,49].

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ejogrb.2010.11.018. References [1] Marana R, Marana E, Catalano GF. Current practical application of office endoscopy. Curr Opin Obstet Gynecol 2001;13:383–7. [2] Fothergill RE. Endometrial ablation in the office setting. Obstet Gynecol Clin North Am 2008;35(2):317–30. [3] Kremer C, Duffy S, Moroney M. Patient satisfaction with outpatient hysteroscopy versus day case hysteroscopy: randomised controlled trial. BMJ 2000;29(320):279–82. [4] Duffy JM, Ahmad G, Watson AJ. Pain relief during hysterosalpingography: a national survey. Human Fertil (Cambridge England) 2008;11(2):119–21. [5] Moore. Clinical orientated anatomy, fifth ed., London: Lipincott Williams & Wilkins; 2006. [6] Farrugia M, Hussain SY. Hysteroscopic endometrial ablation using the Hydro ThermAblator in an outpatient hysteroscopy clinic: feasibility and acceptability. J Minim Invasive Gynecol 2006;13:178–82. [7] Review Manager (RevMan) [Computer Program]. Version 5.0. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration; 2008. [8] Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions, Version 5.0.1. The Cochrane Collaboration; 2008. available from http://www.cochrane-handbook.org [updated September 2008]. [9] Cicinelli E, Didonna T, Ambrosi G, Schonauer LM, Fiore G, Matteo MG. Topical anaesthesia for diagnostic hysteroscopy and endometrial biopsy in post menopausal women: a randomised placebo-controlled double-blind study. Br J Obstet Gynaecol 1997;104:1326–7. [10] Cicinelli E, Didonna T, Schonauer LM, Stragapede S, Falco N, Pansini N. Paracervical anaesthesia for hysteroscopy and endometrial biopsy in postmenopausal women. J Reprod Med 1998;43(12):1014–8. [11] Giorda G, Scarabelli C, Franceschi S, Campagnutta E. Feasibility and pain control in outpatient hysteroscopy in postmenopausal women: a randomized trial. Acta Obstet Gynaecol Scand 2000;79:593–7. [12] Lau W, Lo W, Tam W, Yuen P. Paracervical anaesthesia in outpatient hysteroscopy: a randomised double blind placebo-controlled trial. Br J Obstet Gynaecol 1999;106:356–9. [13] Lau WC, Tam WH, Lo WK, Yuen PM. A randomised double-blind placebocontrolled trial of transcervical intrauterine local anaesthesia in outpatient hysteroscopy. Br J Obstet Gynaecol 2000;107:610–3. [14] Soriano D, Ajaj S, Chuong T, Deval B. Lidocaine spray and outpatient hysteroscopy: randomized placebo-controlled study. Obstet Gynecol 2000;96:661–4. [15] Wong AYK, Wong KS, Tang LCH. Stepwise pain score analysis of the effect of local lignocaine gel on outpatient hysteroscopy: a randomized, double-blind, placebo-controlled trial. Fertil Steril 2000;73:1234–7. [16] Tam WH, Yuen PM. Use of diclofenac as an analgesic in outpatient hysteroscopy: a randomized, double-blind, placebo-controlled study. Fertil Steril 2001;75(5):1070–2. [17] Sunaidi M, Tulandi T. A randomized trial comparing local intracervical and combined local and paracervical anaesthesia in outpatient hysteroscopy. J Minim Invasive Gynecol 2007;14:153–5. [18] Costello M, Horrowitz S, Williamson M. A prospective randomized doubleblind placebo-controlled study of local anaesthetic injected through the hysteroscope for outpatient hysteroscopy and endometrial biopsy. Gynaecol Endosc 1998;7:121–6. [19] Esteve M, Schindler S, Borges Machado S, Argollo Borges S, Ramos Santos C, Coutinho E. The efficacy of intracervical lidocaine in outpatient hysteroscopy. Gynaecol Endosc 2002;11:33–6. [20] Lin YH, Hwang JL, Huang LW, Chen HJ. Use of sublingual burpenorphine for pain relief in office hysteroscopy. J Minim Invasive Surg 2005;12:347–50.

G. Ahmad et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 155 (2011) 3–13 [21] Vercellini P, Oldani S, Colombo A, Bramante T, Mauro F, Crosignani PG. Paracervical anaesthesia for outpatient hysteroscopy. Fertil Steril 1994;62(5): 1083–5. [22] Guida M, Pellicano M, Zullo F, et al. Outpatient operative hysteroscopy with bipolar electrode: a prospective multicentre randomized study between local anaesthesia and conscious sedation. Hum Reprod 2003;18(4):840–3. [23] Sharma JB, Aruna J, Kumar P, Roy KK, Malhotra N, Kumar S. Comparison of efficacy of oral drotaverine plus mefenamic acid with paracervical block and with intravenous sedation for pain relief during hysteroscopy and endometrial biopsy. Indian J Med Sci 2009;63(June (6)):244–52. [24] Costello CF, Horrowitz S, Steigrad S, Saif N, Bennett M, Ekangaki A. Transcervical intra-uterine topical local anesthetic at hysterosalpingography: a prospective, randomised double-blind placebo-controlled trial. Fertil Steril 2002;78(5):1116–22. [25] Elson EM, Ridley NTF. Paracetamol as a prophylactic analgesic for hysterosalpingography: a double blind randomized control trial. Clin Radiol 2000; 55:675–8. [26] Frishman GN, Spencer PK, Weitzen S, Plosker S, Shafi F. The use of intrauterine lidocaine to minimize pain during hysterosalpingography: a randomised trial. Gynecol Oncol 2004;103(6):1261–6. [27] Jacobs SL, Luciano AA, Raphael D, Litt M. The use of topical endometrial lidocaine in hysterosalpingogram and endometrial biopsies. In: 47th annual meeting of American Fertility Society. 1991. p. S78. [28] Kafali H, Cangiz M, Demir N. Intrauterine lidocaine gel application for pain relief during and after hysterosalpingography. Int J Gynecol Obstet 2003; 83:65–7. [29] Liberty G, Gal M, Halvey-Shalem. et al. Lidocaine-Prilocaine (EMLA) cream as analgesia for hysterosalpingography: a prospective, randomized, controlled, double blind study. Hum Reprod 2007;22(5):1335–9. [30] Owens OM, Schiff I, Kaul AF, Crammer DC, Burt RAP. Reduction of pain following hysterosalpingogram by prior analgesic administration. Fertil Steril 1985;43(1):146–8. [31] Peters AAW, Witte EH, Damen ACH, et al. Pain relief during and following outpatient curettage and hysterosalpingography: a double blind study to compare the efficacy and safety of Tramadol versus Naproxen. Eur J Obstet Gynecol Reprod Biol 1996;51:51–6. [32] Guney M, Oral B, Bayhan G, Mungan T. Intrauterine lidocaine infusion for pain relief during saline solution infusion sonohysterography: a randomised controlled trial. J Minim Invasive Gynecol 2007;14:304–10. [33] Marsh F, Thewlis J, Duffy S. Randomized controlled trial comparing Thermachoice III in the outpatient versus daycase setting. Fertil Steril 2007;87:624–50.

13

[34] Varma R, Soneja H, Samuel N, Sangha E, Clark TJ, Gupta JK. Hospital recovery following Thermachoice ablation is not dependent on setting (outpatient or daycase) or rescue analgesia: Unexpected result. Eur J Obstet GynecolReprod Biol 2008;140:76–81. [35] Chapa HO. Utility of in-office endometrial ablation a prospective cohort study of endometrial ablation under local anesthesia. J Reprod Med 2008; 53:827–31. [36] Prasad P, Powell MC. Prospective observational study of Thermablate Endometrial Ablation System as an outpatient procedure. J Minim Invasive Gynecol 2008;15:476–9. [37] De Iaco P, Marabini A, Stefanetti M, Del Vecchio C, Bovicelli L. Acceptability and pain of outpatient hysteroscopy. J Am Assoc Gynecol Laparosc 2000;7(1):71–5. [38] Bonica JJ. The management of pain, 2nd ed., Philadelphia: Lea & Febiger; 1990. [39] Rapkin AL. Neuroanatomy, neurophysiology and neuropharmacology of pelvic pain. Clin Obstet Gynecol 1990;33:119–29. [40] Medsafe – New ZealandMedicines and Medical Devices Safety Authority. Data Sheet: Voltaren1 Rapid 25. Information for Health Professionals. 2007, http:// www.medsafe.govt.nz/profs/datasheet/v/voltarenrapidtab.htm. [41] Dobkin AB, Su JP. Newer anesthetics and their uses. Clin Pharmacol Ther 1966;7(5):648–82. [42] Foye WO, Lemke TL, Williams DA. Foye’s principles of medicinal chemistry, 6th ed., Baltimore: Lipincott Williams & Wilkins; 2008. [43] Blasko G. Pharmocology: a mechanism of action and clinical significance of a convenient antispasmodic agent: Drotaverine. J Am Med Assoc India 1998; 1:63–9. [44] Bolaji OO, Onyeji CO, Ogundaini AO, Olugbade TA, Ogunbona FA. Pharmacokinetics and bioavailability of drotaverine in humans. Eur J Drug Metab Pharmacokinet 1996;21(3):217–21. [45] Turner JA, Deyo RA, Loeser JD, Von Korff M, Fordyce WE. The importance of placebo effects in pain treatment and research. JAMA 1994;271:1609–14. [46] Baker VL, Adamson GD. Minimum intrauterine pressure required for uterine distention. J Am Assoc Gynecol Laparosc 1998;5(1):51–3. [47] Baker VL, Adamson GD. Threshold intrauterine perfusion pressures for intraperitoneal spill during hydrotubation and correlation with tubal adhesive disease. Fertil Steril 1995;64(6):1066–9. [48] Ahmad G, O’Flynn H, Attarbashi S, Duffy JMN, Watson A. Pain relief for outpatient hysteroscopy. Cochrane Database of Systematic Reviews 2010; (11). Art. No.: CD007710. doi:10.1002/14651858.CD007710.pub2. [49] Ahmad G, Duffy JMN, O’Flynn H, Watson A. Pain relief in hysterosalpingography. Cochrane Database of Systematic Reviews 2007; (2). Art. No.: CD006106. doi:10.1002/14651858.CD006106.pub2.