ORIGINAL ARTICLE: GYNECOLOGY AND MENOPAUSE
Pain in diagnostic hysteroscopy: a multivariate analysis after a randomized, controlled trial Ivan Mazzon, M.D.,a Alessandro Favilli, M.D.,b Mario Grasso, M.D.,a Stefano Horvath, M.D.,a Vittorio Bini, B.Sc.,b Gian Carlo Di Renzo, M.D., Ph.D.,b and Sandro Gerli, M.D.b a ‘‘Arbor Vitae’’ Center, Clinica Nuova Villa Claudia, Rome; and b Department of Obstetrics and Gynecology, University of Perugia, Perugia, Italy
Objective: To study which variables are able to influence women's experience of pain during diagnostic hysteroscopy. Design: Multivariate analysis (phase II) after a randomized, controlled trial (phase I). Setting: Endoscopic gynecologic center. Patient(s): In phase I, 392 patients were analyzed. Group A: 197 women with carbon dioxide (CO2); group B: 195 women with normal saline. In phase II, 392 patients were assigned to two different groups according to their pain experience as measured by a visual analogue scale (VAS): group VAS >3 (170 patients); group VAS %3 (222 patients). Intervention(s): Free-anesthesia diagnostic hysteroscopy performed using CO2 or normal saline as distension media. Main Outcome Measure(s): Procedure time, VAS score, image quality, and side effects during and after diagnostic hysteroscopy. Result(s): In phase I the median pain score in group A was 2, whereas in group B it was 3. In phase II the duration of the procedure, nulliparity, and the use of normal saline were significantly correlated with VAS >3. A higher presence of cervical synechiae was observed in the group VAS >3. The multivariate analysis revealed an inverse correlation between parity and a VAS >3, whereas the use of normal saline, the presence of synechiae in the cervical canal, and the duration of the hysteroscopy were all directly correlated to a VAS score >3. Conclusion(s): Pain in hysteroscopy is significantly related to the presence of cervical synechiae, to the duration of the procedure, and to the use of normal saline; conversely, parity seems to have a protective role. Use your smartphone Clinical Trial Registration Number: NCT01873391. (Fertil SterilÒ 2014;-:-–-. Ó2014 by to scan this QR code American Society for Reproductive Medicine.) and connect to the Key Words: Diagnostic hysteroscopy, pain, saline solution, carbon dioxide, distension media Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/mazzoni-pain-diagnostic-hysteroscopy/
D
iagnostic hysteroscopy is the gold standard in the study of uterine cavity morphology and in the diagnosis of endocavitary pathologies (1). Despite being considered an outpatient procedure (2) that is associated with a high success rate and a low complication rate (3), hysteroscopy is largely viewed as a painful procedure, poorly tolerated by patients (4) and often carried out under general or local anesthesia. In an attempt to make the examination less painful, a new generation
of hysteroscopes has been developed, characterized by being of smaller caliber and potentially less invasive (5–8). Other studies have resorted to pharmacologic approaches (9–11), which at times can be relatively invasive (10, 12), to find a solution for pain during hysteroscopy; but besides not reaching consensus on their actual effectiveness (13), those studies also altered the outpatient nature of the procedure (14). Special attention has been paid to the role played by distension media
Received May 19, 2014; revised July 15, 2014; accepted July 25, 2014. I.M. has received nonfinancial support from Karl STORZ Endoscopy GmbH STORZ, Tuttlingen, Germany. A.F. has nothing to disclose. M.G. has nothing to disclose. S.H. has nothing to disclose. V.B. has nothing to disclose. G.C.D.R. has nothing to disclose. S.G. has nothing to disclose. Reprint requests: Sandro Gerli, M.D., Department of Obstetrics and Gynecology, University of Perugia, S. Andrea delle Fratte, 06156 Perugia, Italy (E-mail:
[email protected]). Fertility and Sterility® Vol. -, No. -, - 2014 0015-0282/$36.00 Copyright ©2014 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2014.07.1249 VOL. - NO. - / - 2014
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
during hysteroscopy (15), and despite the substantial literature available on the matter, results are often controversial (16–19) and even unreliable in some cases, owing to the limited number of enrolled patients (17, 20). Ideally, an investigation should consider blinding on multiple levels (including the operator); nevertheless, because of the nature of these trials, the only ones who can be blinded are those who collect the data and the women who participate (21). To investigate the relation between different variables and pain levels during diagnostic hysteroscopy, and considering that there are several variables with different natures among them, a multivariate approach is required: in this way, the distension medium, the pathologies, and the 1
ORIGINAL ARTICLE: GYNECOLOGY AND MENOPAUSE patients' characteristics can all be taken into consideration at the same time. To date, no report has been published with this methodology. To include the distension media in our analysis, we first created a randomized, controlled trial using carbon dioxide (CO2) or saline solution. At a later stage the same patients were analyzed according to the pain experience perceived during hysteroscopy, and a multivariate analysis was carried out. In this manner it was possible to include the distension media as an independent variable in the context of a multivariate logistic regression model.
MATERIALS AND METHODS Phase I: Randomized, Controlled Trial From June 2013 to January 2014 we conducted a prospective, randomized, controlled study of 603 patients who underwent diagnostic hysteroscopy at the ‘‘Arbor Vitae’’ Center for Endoscopic Gynecology (Clinica Nuova Villa Claudia, Rome, Italy). Institutional review board approval was obtained for data collection. The study protocol was registered in the Clinical Trials.gov Protocol Registration System (ClinicalTrials.gov identifier: NCT01873391). Informed consent was obtained from all the patients. All Italian-speaking patients were initially enrolled. Indications for a hysteroscopy were abnormal uterine bleeding, infertility, recurrent miscarriage, fibroids, polyps, uterine malformation, and endometrial thickening. Women with ongoing pregnancy, cervical carcinoma, pelvic inflammatory disease, or women who had previously undergone cervical surgery were not included. To exclude factors that could influence the perception of pain, patients who had previously undergone diagnostic hysteroscopy were ruled out. In addition, exclusion criteria included the presence of excessive uterine bleeding and the execution of an endometrial biopsy during the procedure. Patients allocated to the study were divided into two groups according to the distension media used for the procedure: carbone dioxide (CO2) (group A) or normal saline (group B), in accordance with a previously generated randomized sequence. Randomization was performed by a statistical consultant (V.B.) using computer-generated numbers (StatsDirect version 2.7.2; StatsDirect Ltd.). The distension medium was assigned before starting the diagnostic procedure by opening a sealed, opaque envelope containing the group allotment. To avoid significant bias concerning hysteroscope diameter and shape, the hysteroscopy was performed using a rigid telescope with a diameter of 4 mm and a 30 fore-oblique covered with a single-flow examination sheath with a diameter of 5.1 mm (Hamou II; Karl Storz). In group A an electronic hysteroflator (Hamou Hysteroflator; Karl Storz) was used at a flow rate of 30 mL/min and a maximum intrauterine pressure of 75 mm Hg. In group B normal saline was mechanically infused using a squeeze bag at a maximum pressure of 80 mm Hg. Patients were not allowed to follow the procedure on a monitor. Two expert surgeons with the same abilities and educational background (M.G. and S.H.) indifferently performed 2
all the procedures through vaginoscopy with both distension media according to the randomized allocation. The cervical canal was explored using a 30 fore-oblique view, and the image was placed at a 6 o'clock position on the screen. The surgeons never resorted to the use of a tenaculum to grasp the cervix, nor did they resort to cervical dilatation or to the administration of prostaglandins. Local or general anesthesia was not used, nor were analgesic or antispastic drugs administered before or after the procedure. In women of fertile age, the procedure was carried out during the proliferative stage of the menstrual cycle. Medical history data were collected, such as age, pre- or postmenopausal state, parity, the number of previous cesarean sections, and the indication for the procedure. The duration of the examination was measured from the moment the hysteroscope was inserted into the external orifice of the uterus until its complete removal. In the case of unbearable pain or unsatisfactory vision quality of the uterine cavity, the procedure was stopped and the perceived level of pain was collected. The main characteristics of the cervix were recorded, so as to describe its morphology (anteversion, retroversion, laterodeviation, normoversion, and mixed pattern) and the presence of synechiae, endocervical polyps, stenosis of the external orifice of the uterus, and isthmocele. An individual's pain threshold is somewhat subjective: therefore, to better compare the groups and to obtain a reference point in the evaluation of pain, before starting the procedure patients were asked to quantify their menstrual discomfort on a visual analogue scale (VAS) from 0 to 10, where 0 corresponds to no pain and 10 to the worst pain imaginable. Using the same method, they were asked to express the pain they expected to perceive during the hysteroscopy (22). To avoid shoulder tip pain, at the end of hysteroscopy we invited patients to remain in the supine position for a few minutes while taking deep breaths, to absorb the gas from the abdominal cavity. Five minutes after the procedure, the patients were once again asked the same question, so as to quantify the pain actually experienced during the procedure, as well as the presence of symptoms like nausea, dizziness, or shoulder pain.
Phase II: Analysis According to the Pain Experience The studied population was subsequently divided into two new groups according to the pain experienced during the procedure, regardless of the distension medium used. To investigate which variables were responsible for pain during the hysteroscopy, we defined a painful experience as all the procedures in which the patients reported a VAS score >3. We divided the patients into a first group with a VAS score >3 and a second group with a VAS score %3. These criteria were based on the widely approved standard in the literature, which assesses >3 as the perception of significant pain (22–24). All the variables taken into consideration in phase I were analyzed through a bivariate analysis, to highlight the factors that could potentially and significantly influence the perception of pain in hysteroscopy. A multivariate analysis was VOL. - NO. - / - 2014
Fertility and Sterility® carried out to exclude confounding factors, using a VAS score >3 as the dependent variable and all factors potentially responsible for intense pain as independent variables.
Statistical Analysis We estimated that a sample size of at least 190 women per group would have been required for the primary outcome to detect a 1-VAS-point reduction, at a ¼ 0.01 and b ¼ 0.1. The Shapiro-Wilk test was used to assess the normal distribution of variables. A c2 test with Yates' correction for continuity and the Mann-Whitney U test were used for comparisons of categorical and continuous variables, respectively. A multivariate logistic regression model was fitted for the prediction of the presence of a VAS score >3, incorporating as explanatory variables all the variables that showed a P value %.25 in bivariate analysis (25). The goodness of fit of the logistic regression models was checked using the Hosmer and Lemeshow test. Odds ratios (ORs) with 95% confidence intervals (CIs) were also calculated. All calculations were carried out with Predictive Analytic Software, release 17.0.2 (SPSS Inc.). A two-sided P value of < .05 was considered significant.
RESULTS Of 603 women initially considered for inclusion, only 414 patients were eligible for the study and were allocated to the prospective study. The Consolidated Standards of Reporting Trials (CONSORT) flowchart of the study is illustrated in Figure 1. The data collected were blindly sent to the statistical consultant (V.B.), separated into two different groups.
The procedure was regularly completed in all cases, and the vision quality was always satisfactory. All patients completed the hysteroscopy with the same distension media initially used. The main characteristics of the population, patterns of the cervical canal, duration of the procedure, and the presence of endocavitary pathologies are summarized in Table 1. The intensity of menstrual discomfort reported by the patients was not significantly different between the two groups (P¼ .067). Conversely, a statistically significant difference was reported in the pain intensity in the two groups after the examination. The median pain score in group A was 2, whereas in group B it was 3 (P¼ .001). The median value of the pain perceived by the entire population was 3. Patients never complained of shoulder pain, nausea, or dizziness. With reference to phase II of the study, there were 170 patients who perceived a VAS score >3 (group VAS >3) and 222 women with a VAS score %3 (group VAS %3). The two groups did not show significant differences in age, menopause, or previous cesarean section. Significant correlation with a VAS score >3 was seen with the duration of the procedure (P¼ .01), nulliparity (OR 1.665; 95% CI 1.097– 2.527; P¼ .022), and the use of normal saline as a distension medium (OR 1.560; 95% CI 1.042–2.336; P¼ .039). A higher presence of cervical synechiae was observed in group VAS >3 (OR 2.193; 95% CI 1.292–3.721; P< .005) (Table 2). The multivariate analysis revealed an inverse correlation between parity and a VAS score >3 (OR 0.511; 95% CI 0.311– 0.841; P¼ .008), whereas the use of normal saline (OR 2.044; 95% CI 1.191–3.509; P¼ .009), the presence of synechiae in the cervical canal (OR 2.017; 95% CI 1.056–3.854; P¼ .034),
FIGURE 1
CONSORT flowchart showing recruitment of women into the study. Mazzon. Pain in diagnostic hysteroscopy. Fertil Steril 2014.
VOL. - NO. - / - 2014
3
ORIGINAL ARTICLE: GYNECOLOGY AND MENOPAUSE
TABLE 1 Characteristics of patients, cervical canal, procedure, and VAS score according to the two distension media. Characteristic
Carbon dioxide (n [ 197)
Normal saline (n [ 195)
P value
43.83 10.84 37 (18.78) 118 (59.89) 79 (40.1) 24 (12.18) 94.24 35.60 93 (47.2) 31 (15.73) 62 (31.48) 11 (5.59) 32 (17.02) 53 (27.74) 37 (18.97) 15 (7.89) 5 (0–10); 4.83 3.22 7 (0–10); 6.94 2.48 2 (0–10); 2.74 2.30
44.03 11.16 39 (20) 119 (61.02) 76 (38.97) 37 (19.07) 67.74 31.68 160 (82.07) 23 (11.79) 5 (2.56) 7 (3.58) 38 (20.54) 64 (34.22) 27 (13.84) 20 (10.69) 6 (0–10); 5.43 3.13 8 (0–10); 6.90 3 3 (0–10); 3.71 2.81
.934 .859 .901
Age, y Menopause Nulliparous Multiparous Previous CS Procedure time, s Anteverted cervix Retroverted cervix Normal cervix Mixed shape cervix Cervical synechiae Cervical polyps Stenosis of EUO Isthmocele Dysmenorrhea VAS before hysteroscopy (expected pain) VAS after hysteroscopy (percieved pain)
.082 .0001 .0001 .316 .0001 .25 .461 .211 .219 .448 .067 .547 .001
Note: Data are expressed as number (percentage), mean standard deviation, or median (range) and mean standard deviation. EUO ¼ external uterine orifice. Mazzon. Pain in diagnostic hysteroscopy. Fertil Steril 2014.
and the duration of the hysteroscopy (OR 1.011; 95% CI 1.004–1.019; P¼ .004) were all directly correlated to a VAS score >3 (Table 3).
DISCUSSION Pain in hysteroscopy is considered a major issue that makes the procedure not easily tolerated by patients. Many studies have focused on different variables that influence pain (4, 8, 12, 15, 19, 26–28), particularly on distension media. In a recent meta-analysis Craciunas et al. (21) reported that most of the included studies were of low methodologic quality
TABLE 2 Characteristics of patients, cervical canal, and procedure according to the pain experience. Characteristic Age, y Procedure time, s Carbon dioxide Normal saline Menopause Nulliparous Multiparous Previous CS Anteverted cervix Retroverted cervix Normal cervix Mixed shape cervix Cervical synechiae Cervical polyps Stenosis of EUO Isthmocele
VAS >3 (n [ 170)
VAS £3 (n [ 222)
43.89 11.83 86.16 37.56 75 (44.11) 95 (55.89) 38 (22.35) 115 (67.64) 55 (32.36) 30 (17.64) 117 (68.82) 16 (9.42) 29 (17.05) 8 (4.71) 41 (24.11) 44 (25.88) 34 (20) 17 (10)
43.97 10.73 76.30 34.40 122 (54.96) 100 (45.04) 38 (17.11) 123 (55.41) 99 (44.59) 31 (13.96) 135 (60.81) 39 (17.56) 38 (17.12) 10 (4.51) 29 (13.06) 73 (32.88) 30 (13.51) 18 (8.1)
P value
owing to unsuitableness of the randomization technique, allocation concealment, blinding, and power calculation. For the first time our study presents a multivariate analysis of the independent factors influencing perceived pain during diagnostic hysteroscopy, including the distension media, while respecting an adequate power analysis, the CONSORT guidelines, and blinding the only two levels possible (the data collector and the patients). For this reason we performed a double analysis, first to prospectively analyze the importance of the two distension media (phase I), and second to examine the role of each variable included according to the pain experience (phase II). The results of this study show that pain in hysteroscopy is significantly related to the presence of synechiae in the cervical canal, to the duration of the procedure, and to the use of normal saline as a distension medium. Conversely, parity seems to have a protective role in pain. Unlike other studies (1, 15, 17, 27, 29), we deemed it advisable to also exclude women who underwent an endometrial biopsy during the hysteroscopy. In 2007 de
.842 .01 .039a .248 .022b .385 .114 .031 1 1
TABLE 3 Multivariate analysis considering a VAS score >3 as the dependent variable. 95% CI Factor
B
P value
OR
Lower
Upper
0.715 0.011 0.671 0.146 0.403 0.702 0.454 0.073
.009 .004 .008 .653 .354 .034 .081 .819
2.044 1.011 0.511 1.157 0.668 2.017 0.635 1.076
1.191 1.004 0.311 0.613 0.285 1.056 0.381 0.575
3.509 1.019 0.841 2.187 1.566 3.854 1.058 2.012
Note: Data are expressed as number (percentage) or mean standard deviation. a OR 1.560 (1.042–2.336). b OR 1.665 (1.097–2.527). c OR 2.193 (1.292–3.721).
Normal saline Procedure time Parity Anteverted cervix Retroverted cervix Cervical synechiae Cervical polyps Stenosis of EUO
Mazzon. Pain in diagnostic hysteroscopy. Fertil Steril 2014.
Mazzon. Pain in diagnostic hysteroscopy. Fertil Steril 2014.
4
.005c .162 .115 .669
VOL. - NO. - / - 2014
Fertility and Sterility® Carvalho Schettini et al. (27) published a nonrandomized, observational study with a multivariate analysis examining the causes of pain in hysteroscopy, but the procedure was performed using two different hysteroscopes, the aid of a tenaculum, and with a biopsy. Moreover, women who had previously undergone cervical surgery were also excluded from the study, because they presented a higher risk of cervical synechiae (27). We considered patients in a ‘‘zone of analgesic success’’ if the VAS score was %3 (22–24). Some studies have described a VAS score
