New minimally invasive technologies have allowed traditional operative procedures to be moved to an office setting, which offer advantages to physicians and patients. For physicians, these include more efficient use of time by avoiding operating room (OR) delays and scheduling cases into regular office hours, increased productivity, overall less costs to insurance companies, and potentially greater patient satisfaction.1–3 Patients benefit by avoiding extra preoperative testing and registration time often needed in standard ORs. Nevertheless, the OR offers a user-friendly environment for procedures with more technologically advanced equipment, available staff, and better resources for management of complications. Access to anesthesia is a critical advantage because a common patient concern in the office setting is adequate pain control.4
Paracervical block has been a mainstay of gynecologic local anesthesia, with published reports dating to 1917.5 It functions by blocking the innervation to the cervix and uterus that travels in the paracervical region. Although multiple studies have supported6 and refuted7–10 its efficacy in transcervical procedures and abortions, it may be ineffective with fundal and cornual manipulation as additional neural pathways external to the paracervial supply innervate these regions.10
The Essure (Conceptus, Mountain View, CA) method of hysteroscopic sterilization can be easily performed in an office setting and is well-tolerated, with high rates of patient satisfaction.11–13 Because of the consistency of this procedure and its increased office use in the United States, we thought that it would be an ideal procedure to evaluate the efficacy of a paracervical block for operative office hysteroscopic procedures. Although several studies have indicated that paracervical block may be unnecessary or ineffective in diagnostic hysteroscopic procedures,7,8 few have investigated operative hysteroscopy. As such, we sought to estimate the efficacy of paracervical block in a randomized, double-blind, placebo-controlled fashion during office hysteroscopic sterilization.
MATERIALS AND METHODS
This study was conducted at the Montefiore Medical Center and Albert Einstein College of Medicine faculty practice offices. Approval was obtained from the Montefiore Medical Center institutional review board. This study was designed as a double-blind, randomized, placebo-control trial in accordance with Consolidated Standards of Reporting Trials (CONSORT) guidelines.14,15 All patients seeking hysteroscopic sterilization at the Montefiore Medical Center faculty practice who were eligible for a sterilization procedure from March 2007 to June 2008 were offered participation in this study. Patients were excluded from participation if 1) they had a planned secondary procedure, 2) they had a known allergy to lidocaine, 3) they were having a second attempt at placement of the device after an initial failure, or 4) they were having a planned unilateral placement.
The risks and benefits of the study and the general procedure were explained to all participants at the time of presentation, and a written informed consent was obtained by the investigators. All patients volunteered participation, and no incentives were provided to encourage participation. After consent, demographic data were collected on each patient as described in Box 1. All demographic data and procedural data were entered and stored in a password-protected Microsoft Access database (Microsoft Corporation, Redmond, WA).
After enrollment, participants were assigned a sequential patient number corresponding to a sequential study vial kept in the institutional pharmacy. The vial contents were randomized by a computer program that selected a permuted block randomization scheme with five per block. The vials contained either 11 mL of 1% lidocaine or 11 mL of normal saline. All vials were identical in appearance, prepared by the pharmacy in a blinded fashion, and delivered to the investigator's office in sealed brown bags labeled only with sequential numbers on the bags and the corresponding vials. All participants and investigators were blinded to randomization assignments. The corresponding vial for the patient number was selected and drawn into a syringe immediately before the procedure.
The procedure was performed in a similar fashion for all patients by one of two gynecologic surgeons to minimize interobserver variations. The steps for performing the procedure were based on the manufacturer's instructions (Box 1). At nine points during the procedure, patients were asked to rate the pain on a 10-cm visual analog scale (VAS). We chose the visual analog scale (VAS) as the instrument for pain measurement because the VAS is robust, is able to provide precise and reliable estimates, and has been regularly used in similar trials by multiple specialties for different procedures.16–19 Pain was assessed at each step that could potentially elicit pain, specifically when there was direct physical manipulation (Box 1). The intention of recording pain at each step was to decrease recall bias. Before the performance of each step, participants were advised that they would be required to score the pain for that step, and were then directed by the nurse to a clipboard that listed a VAS for each procedural step to be scored. There was a brief pause at the conclusion of each step to allow participants to mark their score. The procedure proceeded after the participant verbally confirmed marking the VAS for that step and reported that she was ready to proceed with the next step. Participants were also asked to report their overall pain and most pain at the conclusion of the procedure to allow for comparison to other studies that evaluated pain with hysteroscopic sterilization in this fashion.
Procedural information was also recorded at the conclusion of the procedure, including 1) procedural time from hysteroscope insertion to removal, 2) success in device placement, 3) number of trailing coils, 4) amount of distention medium used (milliliters), 5) amount of fluid deficit (milliliters), 6) adverse events, and 7) unusual specific findings. Patients and physicians were queried regarding what treatment group they thought the patient was assigned to, and what led them to that conclusion. Successful placement for both groups was defined as the ability to place a coil in the tubal ostium with two to eight coils trailing in the uterus. All patients were monitored throughout the procedure and for approximately 20 minutes after the procedure for any serious adverse events. These events included lightheadedness, dizziness, chest pain, shortness of breath, loss of consciousness, hemorrhage, uterine perforation, cervical laceration, and intolerable pain. All adverse events were reported to the institutional review board as per institutional policy. If patients were unable to tolerate the procedure secondary to pain, the procedure would be terminated.
A sample size calculation and power analysis was conducted using PS Power and Sample Size Calculations 2.1.30 (Vanderbilt University, Department of Biostatistics, Nashville, TN) to determine necessary sample size. For the purposes of power analysis and sample size calculations, we determined a 0.9-cm difference on the VAS to be a clinically significant difference in pain scores as is standard in prior studies.17,20–25 The reported range of clinically significant differences was from 0.9 to 2.5 cm, and we chose the most conservative end of the range to minimize a potential type II error. Studies have also indicated that differences in the VAS is clinically significant throughout the range of the scale (eg, in low, medium, or high ends of the scale).22,23 The overall calculated mean pain score was the primary outcome measure to power and analyze our results. A prior, unpublished pilot study of 209 patients conducted at our institution found the overall calculated mean pain score pain to be 2.6 cm and the standard deviation for mean pain during the procedure to be 1.41 cm on the VAS. Using 0.9 cm as the clinically significant difference in pain between the lidocaine and placebo groups, we calculated that we would need 40 patients in each group, assuming an α of .05 and a power of 80% and a 1:1 allocation ratio of controls to experimental patients. Although the choice of using a more conservative effect size resulted in an increase in the required number of patients for each arm, we thought that even small differences in pain might be clinically relevant depending on the situation.
To account for potential interpatient pain threshold variability, standardized pain scores were established. Standardized scores were created by taking the raw pain score and adding 1 to it, and subsequently dividing it by the reference pain score. The purpose of adding 1 to the scores was to transform and shift the scale and prevent zero scores in the denominator, without affecting statistical significance because this is a linear transformation. Standardized pain scores were established using two measures: pain with ketorolac injection was used as an objective measure because all patients received an injection before the procedure, and menstrual pain was used as a subjective measure because this is a universal experience for all of the patients, and when providing counseling to patients about the procedure it is common for providers to describe the procedural pain in reference to menstrual pain.
Statistical analysis was performed using Stata/IC 10.1 (Stata Corporation LP, College Station, TX). Determination of normal data distribution was conducted by assessing skewness and kurtosis, the Shapiro-Wilk normality test, and evaluation of plotted histograms. After determination of data distribution, all normally distributed continuous variables were evaluated with a Student t test, with correction if necessary for unequal variances. Nonnormally distributed, continuous variables were evaluated using a Mann-Whitney test. Categorical variables were evaluated with the χ2 test or Fisher exact test, as appropriate. Statistical significance was predetermined at a value of .05. For pain scores that involve procedural laterality (eg, placement in the right and left tube), if no statistical differences were observed between of the two sides as determined by a paired t test, a composite average score was used for the two steps in the analysis.
A multiple linear regression analysis was planned to evaluate the impact of potentially clinically relevant covariates such as age, race, socioeconomic status, body mass index, and procedural time variables on pain scores. Exploratory analysis with multivariable linear regression using a forward stepwise approach was planned to evaluate the potential impact of other demographic and procedural variables on pain scores, particularly if a univariable association, as determined by Pearson correlation or Spearman correlation as appropriate, of P<.10 was found. For the initial univariable exploratory analysis, several of the categorical variables were dichotomized, such as race (Hispanic or non-Hispanic) and education (greater than high school education or high school diploma or less). Additionally, depending on data distribution, either an analysis of variance was performed or the Kruskal-Wallis test was used to look for associations. Indicator variables were created to assess for relationships of pain scores to categorical variables with multiple options in the regression model. Model diagnostics were conducted on the regression analysis to confirm model fit. Unblinding of treatment groups occurred after all statistical analyses were complete.
A total of 103 women presented for hysteroscopic sterilization during the period of the trial; 89 were eligible for participation, and ultimately 80 women were randomized, with 40 women assigned to each group and available for analysis (Fig. 1). Randomization was successful for all demographics (Table 1) except for age, with an average age±standard deviation of 36.4±4.1 (range 29–43) years in the placebo group and 34.2±4.5 (range 24–43) years in the lidocaine group (P=.03). Thirty-seven of the 40 (93%) in each group had successful bilateral placement. There were no complications in either treatment group. The average procedural time for the placebo (8.6±4.1 minutes, 95% confidence interval [CI] 7.3–9.9 minutes) was similar (P=.87) to that of the lidocaine group (8.4±4.8 minutes, 95% CI 6.9–10.0 minutes). No patient required mechanical cervical dilation or chemical cervical preparation to achieve entry into the uterus with the hysteroscope. The variability in procedural times is primarily a reflection of difficulty with accessing and cannulating the tubes. The median fluid deficit was 0 in both groups (P=.26), with a median of 250 mL of saline (range 50–800 mL) used in the placebo group and 300 mL of saline (range 50–1,200 mL) used in the lidocaine group.
As demonstrated in Table 2, statistically significant lower raw pain scores were detected in the lidocaine group for application of the tenaculum (P<.001), delivery of the paracervical block (P<.001), and traversing the external os (P<.001) and the internal os (P<.001). However, no significant difference was observed in the placement of the Essure device (P=.33). The largest reported procedure pain when viewing all the patients in aggregate was placement of the paracervical block itself (3.53±2.30). This observation persisted in the individual treatment groups as well for the placement of the paracervical block in the placebo group (4.49±2.38). The largest reported pain in the lidocaine group was placement of the Essure device (3.15±2.69). The differences in the standardized pain scores (Table 3) decreased in magnitude and statistical significance during ketorolac and even less with the standardized pain score for menses, with no significant difference observed with delivery of the paracervical block (P=.08) and traversing the internal os (P=.14) in the standardized pain score for menses. When comparing the overall calculated mean pain scores (Table 3) for the entire procedure between the treatment groups, there was significantly less pain in the lidocaine group (P<.001), which persisted in the ketorolac standardized pain score (P=.003) but was not present in the menses standardized pain score (P=.11).
When patients were asked to guess which treatment group they were assigned to, 31 of 40 (78%) correctly guessed they were in the lidocaine group and 26 of 40 (65%) correctly guessed they were in the saline group (P<.01). When physicians were asked, 24 of 40 (60%) correctly identified the lidocaine group patients and 28 of 40 (70%) correctly guessed the saline group patients (P<.01). Predictions for group assignment were found to be similar among the physicians and patients (P=.21). Evaluation of pain scores based on patient assignment guesses demonstrated that those who guessed they were assigned to the lidocaine group had self-reported overall pain scores (2.4±1.63, 95% CI 1.92–2.90) that were 1.1 points lower than those who believed they received saline (3.5±1.99, 95% CI 2.77–4.18). A similar finding was observed among the physician guesses, with self-reported overall pain scores of 3.5±1.93 (95% CI 2.91–4.11) and 2.1±1.46 (95% CI 1.61–2.60) in the saline and lidocaine groups, respectively.
In examining the relationship of demographic and procedural factors to overall calculated mean pain score, univariable analysis revealed statistically significant associations of higher parity (r=−0.26, P=.02), prior vaginal delivery (r=−0.26, P=.02), and placement status (r=−0.48, P<.001) with lower pain scores, whereas longer procedural time (r=0.22, P=.05), higher levels of education (r=0.2429, P=.04), and prior surgical history (r=0.34, P=.003) were associated with higher pain scores. No significant difference was determined when examining procedural variables or pain scores between the two surgeons who performed the procedure. When controlling for these factors against the calculated raw average pain scores and treatment group assignment in a multivariable linear regression analysis (Table 4), only treatment group and procedural time were associated with increased pain scores (r2=0.27, P<.01). When examining the model using the ketorolac standardized pain score calculated average pain score, procedural time was not statistically significant, suggesting that procedural time is not directly associated with pain experienced (r2=0.22, P=.016). In the menses standardized pain score model (r2=0.40, P<.001), procedural time (P=.006) and placement status (P=.003) demonstrated a significant association, suggesting that procedural time is truly a substitute for procedural difficulty, and that those who did not have successful placement were more likely to have difficult procedures.
Examination of isolated Essure device placement scores revealed similar univariable associations as the calculated average scores. When a multivariable linear regression analysis was performed looking solely at the pain score associated with placement of the devices, the treatment group was no longer significant (P=.80), whereas procedural time continued to be significant (P=.009), suggesting that paracervical block was ineffective in relieving the pain associated with placement of the devices.
Our results indicate that paracervical block with 1% lidocaine is effective at reducing pain compared with normal saline during the Essure procedure; however, the block has little effect on the pain experienced during placement of the device in the tubal ostia or the ability to successfully place the devices. One study specifically looking at paracervical block with hysteroscopic sterilization had significant methodologic concerns relating to patient selection and pain assessment.2,3,7 Contrary to those prior studies,9,10,26 which have indicated that the paracervical block has little or no effect on cervical manipulation, our results indicate that the paracervical block was effective at reducing pain associated with cervical manipulation. This was supported not only by the statistically significant differences in pain scores between the treatment groups, but also the identical magnitude of pain reduction seen between the groups when traversing both the external and internal os. Additionally, the injection of the anterior lip of the cervix with lidocaine was effective in reducing pain associated with tenaculum placement, something that is frequently assumed to be relatively painless or elicit minimal discomfort that can be managed with distraction.
Close examination of the pain scores does not necessarily support the use of paracervical block in all cases. Because the most painful component of the procedure was the injection, perhaps avoiding the injection altogether would provide a significant reduction in pain scores. This may be due to a “priming the pump” effect, whereby the initial painful stimulus creates an expectation for pain, sensitizing patients to pain, and ultimately leading to higher pain reports. This may be supported physiologically by the initiation of the prostaglandin production cycle, which is more difficult to stop when activated but is easier to interrupt before its initiation. If the placebo group did not have an initial painful stimulus, ie, the placebo injection, it is possible that they would have rated the subsequent scores lower. Although during the designing of this study we considered adding an additional arm that received no injection, we thought that it would be difficult to have an honest comparison between the groups, because neither the patients nor the physicians would be blinded, something that may have a significant bias on pain perception and standardization of procedural performance by the physicians, because physicians may be more inclined to modify their approach to the patients.
Another observation was the fact that pain scores were associated with procedural time. A likely explanation for this is that procedural time is a marker for difficulty of the procedure. In general, the more difficult the Essure placement is, the longer it will take to achieve correct placement, and frequently additional manipulations are needed to assist in appropriate placement. As our results demonstrated, paracervical block does not afford significant pain relief in upper uterine manipulations. As such, the patient is likely to have more pain associated with the procedures.
An important finding is that the largest difference in observed pain scores was 2.3 on the VAS. Even although for the purposes of this study we used a relatively conservative difference of 0.9 on the VAS to be clinically significant to prevent underpowering, some studies indicate that the clinically significant difference in the VAS is around 2.5.20–24 Therefore, although a difference was observed and found to be statistically significant, this may not represent a clinically significant difference.
When examining the placebo group, we observed that pain was not significantly greater than reported menstrual pain. This is critical when counseling patients regarding the pain from the Essure procedure, as well as likely other office hysteroscopy procedures that require less manipulation than Essure. Because pain is one of the most common patient concerns when choosing to undergo an office procedure, the ability to tell a patient that the pain will be similar or less than a typical menstrual period can be very reassuring for many patients.
Although some advocate for using larger doses of lidocaine in the delivery of the paracervical block, our dosage choice was based on the manufacturer's recommendations for paracervical block. In general, those who advocate for using a larger dose of lidocaine do so to allow for greater tissue area dispersion. However, this serves to highlight one of the deficiencies in paracervical block, mainly potential inaccuracy in the delivery along with aberrant neural pathways, because accurate delivery of the block should be effective with an even smaller dose. As such, delivery of a larger bolus of lidocaine may serve to overcome these inaccuracies, but this does not come without risk. The larger doses are more likely to result in intravascular injection and potential side effects. Additionally, some have observed7,26 that the primary analgesic effect may be related to neural compression rather than pharmacologic neural blockade. Therefore, larger volumes of solution, independent of the pharmacologic activity of those solutions, may provide greater relief. However, our results are similar to those of other studies27,28 that do support the pharmacologic activity of lidocaine in producing an analgesic effect rather than a pure mass effect. Nevertheless, future studies evaluating the optimal dosage of lidocaine for paracervical block may provide additional information and influence the effectiveness of this modality.
As our study indicates, overall pain is decreased for the procedure when paracervical block with 1% lidocaine is administered compared with saline injection. This persists in both the calculated average pain and the reported average pain for the procedure in all of the pain score standards we looked at. Given these findings, we would recommend the continuation of paracervical block for uncomplicated Essure procedures. However, additional methods for pain reduction in procedures performed higher in the uterus at the time of hysteroscopy should be sought, because paracervical block offers little improvement in these symptoms. Additionally, further studies designed to observe no injection compared with paracervical block should be considered.
1. Soll D, Snyder DP, Tidwell C. In-office procedures: emerging trends and practical issues. OBG Manage 2007;19 suppl:1–12.
2. Syed R, Levy J, Childers ME. Pain associated with hysteroscopic sterilization. JSLS 2007;11:63–5.
3. Lopes P, Gibon E, Linet T, Philippe HJ. Hysteroscopic tubal sterilization with Essure intratubal devices: a case-control prospective with inert local anesthesia or without anesthesia. Eur J Obstet Gynecol Reprod Biol 2008;138:199–203.
4. Pasternak R. Risk assessment in ambulatory surgery: challenges and new trends. Can J Anesth 2004;51(suppl 1):R1–5.
5. Palmer CB. Obstetrical anesthesia. Cal State J Med 1918;16:175–9.
6. Readman E, Maher PJ. Pain relief and outpatient hysteroscopy: a literature review. J Am Assoc Gynecol Laparosc 2004;11:315–9.
7. Lau WC, Lo WK, Tam WH, Yuen PM. Paracervical anaesthesia in outpatient hysteroscopy: a randomised double-blind placebo-controlled trial. Br J Obstet Gynaecol 1999;106:356–9.
8. Yang J, Vollenhoven B. Pain control in outpatient hysteroscopy. Obstet Gynecol Surv 2002;57:693–702.
9. Renner RM, Jensen JT, Nichols MD, Edelman A. Pain control in first trimester surgical abortion. Cochrane Database of Systematic Reviews 2009, Issue 2. Art. No.: CD006712. DOI: 10.1002/14651858.CD006712.pub2.
10. Tangsiriwatthana T, Sangkomkamhang US, Lumbiganon P, Laopaiboon M. Paracervical local anaesthesia for cervical dilatation and uterine intervention. Cochrane Database of Systematic Reviews 2009, Issue 1. Art No.: CD005056. DOI: 10.1002/14651858.CD005056.pub2.
11. Arjona JE, Mino M, Cordon J, Povedano B, Pelegrin B, Castelo-Branco C. Satisfaction and tolerance with office hysteroscopic tubal sterilization. Fertil Steril 2008;90:1182–6.
12. Sinha D, Kalathy V, Gupta JK, Clark TJ. The feasibility, success and patient satisfaction associated with outpatient hysteroscopic sterilisation. BJOG 2007;114:676–83.
13. Levie MD, Chudnoff SG. Prospective analysis of office-based hysteroscopic sterilization. J Minim Invasive Gynecol 2006;13:98–101.
14. Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al; CONSORT Group (Consolidated Standards of Reporting Trials). The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663–94.
15. Moher D, Schulz KF, Altman DG. The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials. Lancet 2001;357:1191–4.
16. Berthier F, Potel G, Leconte P, Touze MD, Baron D. Comparative study of methods of measuring acute pain intensity in an ED. Am J Emerg Med 1998;16:132–6.
17. Bijur PE, Silver W, Gallagher EJ. Reliability of the visual analog scale for measurement of acute pain. Acad Emerg Med 2001;8:1153–7.
18. Gallagher EJ, Bijur PE, Latimer C, Silver W. Reliability and validity of a visual analog scale for acute abdominal pain in the ED. Am J Emerg Med 2002;20:287–90.
19. Williamson A, Hoggart B. Pain: a review of three commonly used pain rating scales. J Clin Nurs 2005;14:798–804.
20. Bird SB, Dickson EW. Clinically significant changes in pain along the visual analog scale. Ann Emerg Med 2001;38:639–43.
21. Gallagher EJ, Liebman M, Bijur PE. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001;38:633–8.
22. Kelly AM. Does the clinically significant difference in visual analog scale pain scores vary with gender, age, or cause of pain? Acad Emerg Med 1998;5:1086–90.
23. Kelly AM. The minimum clinically significant difference in visual analogue scale pain score does not differ with severity of pain. Emerg Med J 2001;18:205–7.
24. Kendrick DB, Strout TD. The minimum clinically significant difference in patient-assigned numeric scores for pain. Am J Emerg Med 2005;23:828–32.
25. Salaffi F, Stancati A, Silvestri CA, Ciapetti A, Grassi W. Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale. Eur J Pain 2004;8:283–91.
26. Miller L, Jensen MP, Stenchever MA. A double-blind randomized comparison of lidocaine and saline for cervical anesthesia. Obstet Gynecol 1996;87:600–4.
27. Chanrachakul B, Likittanasombut P, O-Prasertsawat P, Herabutya Y. Lidocaine versus plain saline for pain relief in fractional curettage: a randomized controlled trial. Obstet Gynecol 2001;98:592–5.
28. Titapant V, Chawanpaiboon S, Boonpektrakul K. Double-blind randomized comparison of Xylocaine and saline in paracervical block for diagnostic fractional curettage. J Med Assoc Thai 2003;86:131–5.