Objective
The primary objectives of this study were to explore the pain experience after gynecologic laparoscopy that is performed for nonacute pain conditions and to determine whether preoperative psychologic tests and quantitative tests of sensitization can predict postoperative pain.
Study Design
Participants included 61 women who underwent laparoscopy for nonacute pain (n = 61). A second group of 16 women who had undergone tubal ligation was included to explore whether laparoscopy induced a painful postoperative response in women without preoperative pain. Subjective tests included numeric pain scale, pain catastrophizing scale, depression scale, global assessment of change, and the McGill Pain Questionnaire Short Form. Quantitative sensory testing included abdominal cutaneous allodynia, trigger points, and reduced pain thresholds. The nonacute pain sample had 80% power to detect a difference of 0.5 standard deviation in average pain levels. Analysis included parametric and nonparametric comparisons of groups and univariate and linear regression analysis of clinically relevant variables.
Results
In women who underwent tubal ligation, pain levels were low before and after the procedure. In women who underwent surgery for nonacute pain, pain levels at 6 months and all psychologic test scores were reduced significantly compared with baseline ( P < .001 and P = .001, respectively). Among those women with positive results on the quantitative pain tests of sensitization at baseline, average postoperative pain was also significantly reduced ( P < .001). Univariate analysis demonstrated only tests of sensitization that were correlated with change in average pain level ( P = .01). Regression analysis suggested that baseline pain, catastrophizing, and the presence of cutaneous allodynia significantly predicted pain levels 6 months after surgery ( P < .001).
Conclusion
Pain after laparoscopic surgery for nonacute painful conditions can be predicted by baseline pain, catastrophizing, and the presence of allodynia, which is a simple swab test that indicates sensitization.
Postoperative pain is a dilemma (5-40%) for a large number of surgical procedures. Pain has been associated with hernia, breast, thoracic, breast, and gallbladder surgeries, hysterectomy, and cesarean section delivery. The origin of pain has been reported to represent neural injury, central sensitization, and inflammatory injury. Pelvic surgery was not identified as having a high probability of neuropathic pain in an assessment of neuropathic probability.
Macrae’s definition of persistent postoperative pain requires that pain develop after a surgical procedure, that it is of at least 2 months duration, that other causes of pain have been excluded, and importantly the possibility that the pain is continuing from a preexisting condition must be explored and excluded. This last requirement is a challenge in pelvic surgery because the procedures commonly are undertaken for painful conditions and because the continuing pain is often attributed to persistence or recurrence of the disease. The term postoperative pain will be used in this study.
The few available randomized sham-controlled surgical trials on pelvic pain have focused on the excision of endometriosis to treat pain have shown both significant improvement to no difference from sham controls after the operation. None of the studies included objective pain testing preoperatively or psychologic contributions to the outcome. A structured review of the effect size of pelvic surgical interventions has shown that only 30–40% of women reported pain relief after short periods of follow-up observation and that the effect size decreased by as much as 50% at longer points of observation. A study from 1996-2007 of 24,000 women in Alberta, Canada, demonstrated extensive use of repeated operative laparoscopy procedures that supported the need for investigations into the physiology of postoperative pelvic pain.
Quantitative sensory testing has been investigated as a tool to identify those at risk and those who might benefit from preventative treatment. A systematic review of the available literature indicates that preoperative pain tests can identify 4-52% of the variance in postoperative pain experience.
At the time we designed our research, only one study examined risk factors for postoperative pain in women who had had laparoscopic tubal ligation. It was limited by a 10-day follow-up period and assessed pain in primarily nonpelvic locations. Heat pain perception and anxiety scores rendered the best predictive model during exercise ( R = 0.653; P < .001). Preoperative pain testing accounted for 29-43% of the total variance in postoperative movement-related pain.
The primary objective of this study was to explore the pain experience after gynecologic laparoscopy, which is indicated for women with nonacute pain. A second group of women who requested tubal ligation was included to explore whether laparoscopy induced a painful postoperative response in women without preoperative pain. Analysis of results of the nonacute pain group emphasized psychologic aspects and quantitative tests of sensitization to determine whether postoperative pain can be predicted.
Materials and Methods
Subjects
Our cohort study was performed in a tertiary care center in Calgary, Alberta, Canada. We recruited women 18-50 years old who were recommended to have laparoscopy to manage nonacute pain (pain of >1 month). A second group of women who sought a tubal ligation was recruited to an operative approach to explore whether laparoscopy induced postoperative pain. Women who had bilateral oophorectomy, who were menopausal, who were pregnant, or who had significant medical illness were excluded. Women were approached by their gynecologist and were interviewed by a study nurse who described the study and obtained informed consent. None of the patients who were approached declined the opportunity to participate. Eligibility was confirmed before the patient signed a consent form. Potential severe preoperative psychologic conditions were screened, but none occurred. Ethics approval was obtained from the University of Calgary Ethics Committee.
Questionnaires
Eligible women received questionnaires by mail that were to be completed before surgery. If a woman arrived for surgery without having completed the questionnaires, they were completed on the day of surgery. Five questionnaires were used: (1) the Center for Epidemiologic Studies Depression Scale, which includes 20 items that are each scored from 0–3 (possible range, 0–60) ; (4) the Pain Disability Index, which consists of 7 items and score range of 0 (no disability) to 10 (worst disability; possible range, 0–70) ; (5) the Short Form McGill Pain Questionnaire, which includes 5 subscales of continuous pain, intermittent pain, neuropathic pain, affective descriptors, and an overall Short Form McGill Pain Questionnaire score (each subscale and overall score had a possible range of 0–10) ; and Pain Catastrophizing Scale: 13 items which were each scored from 0 to 4 (possible range, 0–52). As well, women provided a subjective pain level rating of “average” pelvic pain that had been experienced over the past month (11-point numeric rating scale; range, 0 [no pain] to 10 [worst pain imaginable]). The total time for the completion of the questionnaires was approximately 30 minutes.
Preoperative tests
We selected 3 tests of sensitization that were found to be helpful clinically in the discrimination of visceral from somatic causes of chronic pelvic pain. The presence of abdominal cutaneous allodynia (positive test result) was identified with a cotton-tipped applicator test. The test involves passing a cotton-tipped applicator caudally from the mid clavicular line on the anterior abdominal wall inferiorly in line with the lateral aspect of the rectus abdominus. This detects changes close to regions of the anterior cutaneous nerves that emerge from the rectus fascia. The subject was asked to state clearly if and when the sensation changed suddenly and shifted from a normal touch to a painful sensation. Repeated measures were taken from side-to-side to delineate the regions of allodynia. Both sides of the abdomen were evaluated. The test of cutaneous allodynia is valid and reliable with a sensitivity of 73% (95% confidence interval [CI], 50–89%) and specificity of 100% (95% CI, 85–100%) in relation to a diagnosis or preexisting or concurrent visceral disease.
The presence of muscle tenderness (positive test result) was identified in the areas lateral to the rectus muscle in the anticipated regions of the anterior cutaneous spinal nerves.
Detection of myofascial trigger points followed the criteria that had been described in a recent review (the identification of the taut band and reproduction of the spontaneous pain complaint by pressure on a point of tenderness within the band). Recent findings have suggested that these areas are related closely to the anterior spinal nerves because they perforate the rectus fascia.
The presence of reduced pain thresholds (positive test result) was undertaken with the von Frey electroanesthesiometer (IITC Life Science, Woodland Hills, CA), which provides a measurement in grams. The instrument was placed on the lower abdomen in the region of muscle tenderness and the pressure was gradually increased. A positive test result was the onset of pain at <100 g. A negative test result was no pain at a pressure of 100 g. The total time for the testing for sensitization took approximately 5 minutes for all tests.
For each participating woman, patient characteristics were extracted from the hospital chart. Details of current surgery were also extracted from the hospital chart. Details of the anesthetic medications and pain experience in the recovery room and in the Day Surgical Unit at discharge were also collected.
Postoperative tests
All women were invited to return to the gynecology clinic 6 months after surgery when questionnaires and tests of sensitization were repeated. Every attempt was made to encourage women to return to the clinic for follow-up evaluation (by mail, telephone, and email reminders). Women who were unable to return to the clinic were invited to complete the questionnaires at home and mail them to the clinic.
Statistical analysis
All data were entered and managed with an Access database (Microsoft Office Professional; Microsoft Corporation, Redmond, WA) and were analyzed with SPSS statistical software (version 19; SPSS Inc, Chicago, IL). We estimated that a sample of 60 women with measurements before and after surgery would have >80% power to detect a medium effect size of 0.5 standard deviations of change in patient-reported pelvic pain level (assuming alpha = .05).
For both clinical indications (request for tubal ligation and the laparoscopic management of pelvic pain), patient characteristics are presented as number (percentage), median (interquartile range [IQR]), or mean (standard deviation) as appropriate. Pain testing and patient-reported pain levels were examined separately with descriptive statistics for the tubal ligation patients.
For patients who underwent laparoscopic treatment of nonacute pelvic pain, change from baseline for pain measurement tests and patient-reported pain was calculated with the use of paired statistical tests ( t -test for normally distributed change scores, Wilcoxon signed ranks test for nonnormally distributed change scores, and McNemar’s test for paired proportions), as appropriate. Patient-reported pain scores in hospital and 6 months after surgery were described for baseline pain testing variables with median (IQR).
Univariate analysis was undertaken to explore the relationship of baseline pain testing variables to the outcome measure of change in patient-reported average pain. Regression analysis was undertaken to examine the relationship of pain reports at 6 months in relation to the clinically relevant variables of baseline average pain scores and catastrophizing scores in addition to cutaneous allodynia.
The work is described according to the recommendations of the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. The study received ethics approval from the University of Calgary Conjoint Health Ethics Review Board (# E-23124).
Results
Patients and procedures
Sixty-one women underwent laparoscopic gynecologic surgery for painful conditions, and 16 women had a laparoscopic procedure for a nonpainful condition (tubal ligation for permanent sterilization). The characteristics of the women are shown in Table 1 . Use of narcotics and analgesics were found exclusively in the women with pain. More than one-half of these women reported previous abdominal surgery, which included laparoscopy, laparotomy, and adnexal surgical procedures. Indications for the current surgery and the specific surgical procedures are described in Table 1 . There were no exclusions on medical grounds.
| Variables | Laparoscopy | |
|---|---|---|
| Painful condition (n = 60) | Nonpainful condition: tubal ligation (n = 16) | |
| Age at time of operation, y a | 33.0 ± 7.5 (19–48) | 33.1 ± 6.6 (22–44) |
| Body mass index, kg/m 2 b | 23.1 (5.9; 17.0–45.4) | 26.6 (6.4; 19.8–32.0) |
| Current smoker, n (%) | 19 (31) | 6 (38) |
| Current prescription medications, n (%) | ||
| Narcotic | 22 (36) | 0 |
| Other analgesic | 14 (23) | 1 (6) |
| Antidepressant | 13 (21) | 2 (13) |
| Hormones | 23 (38) | 5 (31) |
| Other medications | 17 (28) | 0 |
| History of abdominal surgery, n (%) | ||
| Women with no previous surgery | 29 (48) | 6 (38) |
| Women with ≥1 previous surgery | 31 (51) | 10 (63) |
| Types of previous surgery, n (%) | ||
| Laparoscopy/laparotomy c | 21 (68) | 0 |
| Adnexal d | 11 (35) | 0 |
| Nongynecologic e | 9 (29) | 3 (30) |
| Other gynecologic f | 9 (29) | 3 (30) |
| Cesarean section delivery | 8 (26) | 7 (70) |
| Cautery/excision of endometriosis | 4 (13) | 0 |
| Painful indications for current surgery, n (%) | N/A | |
| Cyst/tumor | 19 (31) | |
| Endometriosis | 13 (21) | |
| Exploratory | 20 (33) | |
| Other | 19 (31) | |
| Index laparoscopic surgery, n (%) | ||
| Observation: no disease identified | 2 (3) | 0 |
| Endometriosis | 17 (28) | 0 |
| Adnexal | 43 (70) | 16 (100) |
| Other gynecologic | 30 (49) | 0 |
a Data are given as mean ± SD (range)
b Data are given as median (interquartile range; range)
c Any laparoscopy/laparotomy including for tubal ligation, cautery endometriosis, or excision endometriosis
d Tubal ligation, oophorectomy, or surgery for ectopic pregnancy
e Appendectomy, cholecystectomy, hernia repair, or abdominoplasty
f Endometrial ablation, hysterectomy, dilation, cervical loop electrosurgical excision procedure/laser treatment.
Pain measurement tests
Women who underwent tubal ligations had low frequencies of positive pain measurement tests at baseline (3/15 women had a reduced pain threshold; 2/16 women had cutaneous allodynia, and 1/15 women had trigger points), and none were positive at 6 months. The median average pelvic pain rating reported by these women was 0.5 at baseline (IQR, 3.3) and 0.0 (IQR, 2.0) at 6 months, with a median change in average pelvic pain of 0 (IQR, 0.5). Because they had low pain levels at baseline and 6 months, these patients were not included in the analysis that predicted pelvic pain after surgery.
The detailed pain experience of women who underwent laparoscopic surgery for pain is presented in Table 2 . The reported average pain overall decreased significantly, which was observed mainly in women who complained of continuous pain. Ten women had an increase in pain after surgery; 46 women noted a decrease in pain. There was no difference in the proportion of adnexal (oophorectomy, salpingo-oophorectomy) vs peritoneal surgery (cautery, excision, or adhesiolysis) among women with an increase in pain compared with women with a reduction in pain (70% peritoneal surgery for those with an increase in pain vs 52% for those with a reduction in pain; P = .485). All psychologic tests of depression, catastrophizing, disability and quality of life were improved significantly.
| Pain measurement | Baseline: median (interquartile range) or n (%) | 6-mo after operation: median (interquartile range) or n (%) | Mean change ± SD, median change (interquartile range), and/or difference in proportion (95% CI) | P value a |
|---|---|---|---|---|
| Patient reported pain | ||||
| Average pain rating over past month | 5 (3; range, 0–8.5; 2 missing) | 3 (4; range, 0–8.0; 3 missing) | –1.8 ± 2.6 | < .001 |
| (–2.5 to –1.1; 5 missing) | ||||
| Type | ||||
| Intermittent | 31 (52%) | 31 (53%) | –2% (–18% to 15%) | > .999 |
| Cyclic | 26 (43%) | 23 (39%) | –5% (–21% to 12%) | .678 |
| Sporadic | 28 (47%) | 23 (39%) | –9% (–26% to 10%) | .441 |
| Continuous | 22 (37%; 1 missing) | 9 (15%; 2 missing) | –19% (–22% to –6%; 3 missing) | .003 |
| Center for Epidemiologic Studies Depression Scale b | 19.5 (22; 1 missing) | 11 (13; 6 missing) | –5.5 (17) | < .001 |
| (–11 to –1; 7 missing) | ||||
| Pain Catastrophizing Scale c | 22 (25; 4 missing) | 13 (22; 10 missing) | –6 (15) | .001 |
| (–11 to –2; 12 missing) | ||||
| Pain Disability Index d | 36 (24; 2 missing) | 16.5 (29.8; 5 missing) | –16 (18) | < .001 |
| (–23 to –10; 7 missing) | ||||
| Short Form McGill Pain Questionnaire e | (3 missing) | (7 missing) | (9 missing) | |
| Continuous pain | 3.8 (3.6) | 1.8 (4.9) | –1.4 ± 2.1 | < .001 |
| (–1.9 to –0.8) | ||||
| Intermittent pain | 2.8 (3.9) | 1.4 (3.8) | –1.3 ± 2.8 | .001 |
| (–2.1 to –0.6) | ||||
| Neuropathic pain | 0.7 (2.3) | 0.4 (1.7) | –0.5 ± 1.8 | .064 |
| (–0.9 to 0.03) | ||||
| Affective | 3.0 (4.6) | 1.1 (3.8) | –1.3 ± 2.9 | .001 |
| (–2.1 to –0.5) | ||||
| Short Form McGill Pain Questionnaire Score e | 2.7 (3.2) | 1.4 (2.8) | –1.1 ± 1.9 | < .001 |
| (–1.6 to –0.6) | ||||
| Abdominal pain measurements f | ||||
| Reduced pain thresholds | 28 (48%; 3 missing) | 25 (46%; 7 missing) | –2% (–14% to 11%; 9 missing) | > .999 |
| Cutaneous allodynia | 31 (54%; 4 missing) | 25 (46%; 7 missing) | –6% (–21% to 11%; 8 missing) | .629 |
| Trigger points | 34 (60%; 4 missing) | 24 (44%; 7 missing) | –13% (–24% to 3%; 9 missing) | .118 |
a Number (and percentage) of women with positive test results
b 20 items, possible range, 0–60; higher scores indicate greater depressive symptoms
c 13 items, possible range, 0–52; higher scores indicate greater level of catastrophizing
d 7 items, possible range, 0–70; higher scores indicate greater disability
e 22 items, possible range, 0–10 for each subscale and overall score; higher scores indicate greater pain
f Paired t -test for normally distributed change scores, Wilcoxon signed ranks test for non-normally distributed change scores, or McNemar’s test for paired proportions.
To explore the experience of pain in relation to pain testing across the evaluation period, the pain scores are indicated at the time of release from the recovery room and from the discharge room and 6 months after the operation ( Table 3 ). No substantial changes were observed when the data were assessed by the presence or absence of positive pain measurement tests.
