Elective abortion is one of the most common outpatient surgical procedures with an estimated 43.8 million performed annually worldwide.1 The majority of patients undergoing a surgical first-trimester abortion in the United States receive oral ibuprofen and local anesthesia through a paracervical block.2 Despite a variety of analgesic approaches, many patients find surgical abortion extremely uncomfortable; 78–97% report at least moderate procedural pain.3–5
Only 4.4% of abortion providers in North America reported routinely using oral opioids in 2002. However, it is likely that this percentage has increased along with the documented rise of prescription and illicit opioid use.6 At our study site, Planned Parenthood-Columbia Willamette, patients often request opioids by name, and 80% of patients choose this for premedication when offered (personal communication with Planned Parenthood-Columbia Willamette, 2010).
Gestational age has been shown to influence pain experience. Post hoc analysis of a prospective randomized clinical trial estimating the benefit of intrauterine lidocaine demonstrated increased abortion-related pain at a gestational age of 8 weeks or more compared with 7 6/7 weeks or less.7,8 This is in contrast to the results of an observational study.9
A recent Cochrane review highlighted the absence of data to support the use of oral opioids such as hydrocodone–acetaminophen in abortion procedures.4 Although this medication is rarely used as the sole analgesic for abortions, hydrocodone–acetaminophen is increasingly added to the regimen of ibuprofen and paracervical block and is recommended by the World Health Organization as a pain control option.2,10 The objective of this study was to evaluate the efficacy of hydrocodone–acetaminophen for pain control during first-trimester surgical abortion.
MATERIALS AND METHODS
This trial was conducted at Planned Parenthood-Columbia Willamette in Portland, Oregon. Study procedures were initiated on February 22, 2011, after approval by the institutional review boards at Oregon Health & Science University and Planned Parenthood Federation of America. Enrollment was completed on October 15, 2011. A randomized, double-blinded, placebo-controlled design was incorporated to address the hypothesis that oral hydrocodone–acetaminophen (two tablets of 5–325 mg) given 45–90 minutes before the procedure along with a standard medication regimen of ibuprofen, lorazepam, and paracervical block would decrease pain related to first-trimester surgical abortion in patients who were at less than 11 weeks of gestation. Because gestational age may confound the relationship of pain, randomization was stratified by gestational age (less than 8 0/7 weeks and 8 0/7 weeks to 10 6/7 weeks).
Eligible women were recruited by research personnel at Planned Parenthood-Columbia Willamette until the required sample size was obtained. Potential study participants were approached after the decision to undergo a surgical abortion was made and gestational age was determined by ultrasonography. Participants in this randomized controlled trial were English-speaking or Spanish-speaking women aged 18 years or older at less than 11 weeks of gestation by ultrasonographic dating, voluntarily requesting surgical pregnancy termination at Planned Parenthood-Columbia Willamette. The upper gestational age limit was selected because misoprostol is used for cervical ripening starting at 11 weeks of gestation. All patients had an intrauterine gestational sac on ultrasonogram, were eligible for suction curettage, in good general health, and willing and able to give informed consent and agree to the terms of the study. Women were excluded if they had symptomatic early pregnancy failure, had received premedication with misoprostol, had used any opioid within the previous 7 days or requested an opioid medication or intravenous sedation during the procedure, or if they had contraindications to any of the medications used in the study protocol. Patients were also excluded if they refused ibuprofen or lorazepam or if they had evidence of pelvic inflammatory disease or known hepatic disease.
Participants signed an Oregon Health & Science University institutional review board–approved written consent available in English and Spanish. A recruitment log tracked patients who were excluded at any point throughout recruitment of the study or who declined entry. Participants completed a detailed questionnaire inquiring about demographic information, previous use of hydrocodone–acetaminophen and other opioids (prescription or illicit), health and pregnancy history, baseline pain, anxiety, nausea, and expected pain. Participants were then randomized to treatment group using a predetermined computer-generated blocked randomization (block size of six) and were allocated using sequentially numbered, opaque, sealed envelopes. The randomization scheme and block size were provided to the primary investigator after enrollment and data entry were completed.
The study coordinator presented the nurse administering the preoperative medications with the allocation sealed envelope containing either hydrocodone–acetaminophen or placebo. All patients received premedication with 800 mg oral ibuprofen and 2 mg oral lorazepam along with either two tablets of 5–325 mg hydrocodone–acetaminophen (a total of 10 mg hydrocodone and 650 mg acetaminophen) or two tablets of placebo. Each hydrocodone–acetaminophen or placebo (methylcellulose powder) tablet was placed inside an identical separate capsule to ensure blinding. The participant, the patient advocate who was present for the procedure, the physician performing the procedure, and the study coordinator assessing outcomes were blinded to participants' allocation status. Premedication occurred between 45 and 90 minutes before the procedure.
The procedures were performed in accordance with standard clinical procedure. Research personnel collected patient pain scores on the visual analog scale immediately after each step. After placement of the speculum, patients received a standard paracervical block based on a common technique shown to be effective in a recent study.11 The cervix was then dilated to the number of millimeters equivalent to or 1 mm less than the patient's gestational age in weeks and a corresponding size suction cannula was used. Up to a gestational age of 7 6/7 weeks, manual vacuum aspiration generally was used; beyond 8 weeks of gestation, electric vacuum aspiration was selected. To minimize variability in procedures, all procedures were performed by experienced health care providers (three attending physicians, three family planning fellows, and three midlevel health care providers).
Research personnel collected postprocedure information from participants 30 minutes after speculum removal, including vital signs (heart rate and blood pressure), pain, side effects (nausea, sleepiness, and pruritus), and satisfaction. Participants were asked to indicate whether they believed they were randomized to receive active drug or placebo. Health care providers were also given a questionnaire to report any unusual or adverse events. The primary outcome was patient perception of pain with uterine aspiration reported on a 100-mm visual analog scale. Secondary outcomes included reported pain at time points before, during, and after the procedure, side effects, anxiety and nervousness as well as participant satisfaction with pain control and overall procedure.
Data were analyzed with Stata 11.2. Graphics were created with Stata and Prism 5. All variables were analyzed using an intention-to-treat approach. The two gestational age strata were analyzed separately, because it was of interest to look at effects for each group. We also tested the interaction between gestational age stratum and treatment group and if differences between treatment groups did not differ by gestational age, we planned to analyze the two gestational age strata together for an overall effect. Two-sample t tests or the Mann-Whitney U test were used to compare the means of procedural characteristics and outcomes on the visual analog scale as appropriate. We used a linear mixed effects model to test whether the difference between the treatment groups changed throughout the procedure using an interaction term between procedural time point and treatment group. The identification variable for participants was entered into the model as the random effect to account for correlation among observations at multiple time points for each participant. Pearson's χ2 and Fisher's exact test were used to compare adverse events and health care provider type between groups. Univariable and multivariable linear regression was performed to explore predictors of pain and satisfaction using data from both gestational age groups.
Based on previous data, a 13-mm to 20-mm (or equivalent to 30%) difference in pain on a 100-mm visual analog scale has been considered clinically meaningful.12–14 The mean standard deviation using the visual analog scale to evaluate pain during first-trimester abortion was 26 mm.4 To achieve 80% power at a significance level of 5% (two-sided α=0.05), a total of 54 participants (27 in each arm) were required to detect a 20-mm difference or greater on a 100-mm visual analog scale for each gestational age strata (a total of 108 patients for both strata). Combining the two gestational age strata allowed detection of a 15-mm difference in pain. Adding 10% more patients to compensate for possible withdrawal of study participants resulted in a target of 120 participants. Detection of the minimal clinically meaningful difference of 13 mm would have required a total of 126 participants (63 per group).
Participant flow is shown in Figure 1. A total of 911 patients were assessed for eligibility to enroll 121 patients. The most common reason for exclusion was gestational age. Enrollment of the early gestational age stratum was completed first; 109 early gestational age patients were excluded, whereas recruitment of the late stratum continued. Other common reasons for exclusion were request for intravenous sedation or oral opioids and requesting only ibuprofen for the procedure.
Two participants randomized to the placebo group did not complete the study. One participant in the late gestational age stratum vomited immediately after ingesting the study medication and declined further participation in the study after completing the initial questionnaire. A participant at 10 weeks of gestation experienced a uterine perforation necessitating transfer to a hospital for laparoscopy; she did not complete any study procedures after the visual analog scale for pain with cervical dilation. The existing data for these two participants were included in the analysis. One additional patient beyond the planned 60 was recruited in the late gestational age group on the day the previous participant vomited and withdrew from the study.
Demographic characteristics of participants according to study group and gestational age stratum are shown in Table 1. Participants enrolled in the study were in their mid-20s, on average, and most identified as white. Approximately half of participants were nulliparous. Average gestational age was 7.2 weeks for the placebo group and 7.4 weeks for the hydrocodone–acetaminophen group. Within the early gestational age stratum, participants randomized to the hydrocodone–acetaminophen group had a higher mean body mass index (BMI [ calculated as weight (kg)/[height (m)]2] 27.1 compared with 22.9), and a greater proportion had at least one prior surgical abortion (46.7% compared with 20.0%). However, patient characteristics were similar between participants receiving placebo and hydrocodone–acetaminophen when the gestational age strata were combined. Most participants described their level of menstrual symptoms as easy or mild cramping. Most participants (83%) reported use of one or more prescription opioid medications in the past; none reported use of heroin.
Procedural characteristics for the study groups are shown in Table 2. There were no significant differences in health care provider type or postprocedure vital signs among participants receiving placebo or hydrocodone–acetaminophen. There was a nonsignificant increase in procedure time among the hydrocodone–acetaminophen group for the early gestational age stratum (P=.05), likely related to higher BMI among these participants. There was no difference in medication-related adverse events. One participant in the hydrocodone–acetaminophen group underwent reaspiration for hematometra shortly after the procedure. Three participants experienced vomiting while in the recovery room; two had received hydrocodone–acetaminophen and one received placebo.
Mean visual analog scale scores throughout the abortion procedure for participants receiving placebo compared with hydrocodone–acetaminophen are shown for the early and late gestational age strata in Figure 2 A–B. Based on the linear mixed effects model, there were no significant differences between participants receiving placebo or active drug in visual analog scale scores during the procedure for the early gestational age stratum (P=.58) or late gestational age stratum (P=.50). Results were similar when combining gestational age strata (P=.43), and there was no significant interaction between gestational age and treatment group. The mean visual analog scale score for the primary outcome, pain with uterine aspiration, was 65.9 mm for the active treatment group and 63.2 mm for the placebo group (P=.59).
There were no statistically significant differences in postoperative side effects including nausea, sleepiness, and pruritus among study groups. In the late gestational age stratum, median nausea among participants who received hydrocodone–acetaminophen was 6 mm compared with 1 mm among those who received placebo (P=.05). Baseline nausea was significantly greater in the placebo group; in regression analysis controlling for baseline nausea, the hydrocodone–acetaminophen group had significantly greater postoperative nausea (P=.03). Sleepiness scores on the visual analog scale were high for all participants. Mean sleepiness was 54.1 mm in the placebo group compared with 63.2 mm in the hydrocodone–acetaminophen group (P=.06).
Participant satisfaction with pain control and with the overall procedure was high for both study groups (overall procedure mean 84.9 mm in the placebo group, 88.5 mm in the hydrocodone–acetaminophen group, P=.23). For both the early and late gestational age strata, the two satisfaction visual analog scale scores were somewhat higher in the hydrocodone–acetaminophen group, but the differences were not statistically significant in either gestational age stratum or in the combined analysis.
In univariable linear regression analysis of pain during aspiration, Hispanic race, nervousness about the procedure and pain, and expected pain were found to be significantly associated with this outcome (Table 3). In multivariable analysis, Hispanic race and nervousness about pain remained significant when controlling for study group and gestational age (Table 4). Previous use of opioids was not significantly associated with pain outcomes, and there was no interaction with study group. In the univariable regression analysis of satisfaction with pain control, procedural pain (speculum insertion, paracervical block placement, dilation, and aspiration), postoperative pain, and participant belief of randomization to the placebo group were significantly associated with decreased satisfaction with pain control (Table 5). However, only pain during dilation and aspiration were significant predictors of satisfaction with pain control in multivariable analysis controlling for study group and gestational age (Table 4). In this analysis, study group was a significant predictor of satisfaction with pain control (P=.05), indicating that despite no difference in procedural pain, participants in the active treatment group reported slightly higher satisfaction.
In multivariable analysis, study group (hydrocodone–acetaminophen or placebo) was not found to be a significant predictor of pain at any procedural time point even when controlling for expected pain as a continuous or dichotomous variable. Similarly, it was not significantly associated with pain outcomes when controlling for BMI and prior surgical abortion, the two demographic variables that were significantly different between study groups in the early gestational age stratum.
Our study did not find that hydrocodone–acetaminophen decreases pain during first-trimester abortion. A benefit was not observed for participants less than 8 weeks of gestation or for those between 8 and 11 weeks of gestation. For the primary outcome, pain during aspiration, the mean difference on the 100-mm visual analog scale was 2.6 mm with slightly higher pain in the active hydrocodone–acetaminophen group. Although we designed our study to demonstrate a clinical benefit, the 95% confidence interval for this mean difference in pain and, for the differences in pain at each procedural time point, was within ±15 mm, suggesting that hydrocodone–acetaminophen is clinically equivalent to placebo for pain control during surgical abortion. Importantly, although we showed no benefit with use of hydrocodone–acetaminophen, there was a suggestion of harm with increased postoperative nausea compared with placebo.
Oral opioids have not been well studied in gynecologic outpatient settings. One previous randomized controlled trial compared 50 mg tramadol with 800 mg ibuprofen administered 1 hour before surgical abortion. Ibuprofen was more effective than tramadol at reducing pain 30 minutes after surgical abortion.15 In another randomized controlled trial, 10 mg oral oxycodone with 1 mg lorazepam was compared with an intravenous regimen of 100 µg fentanyl and 2 mg midazolam.16 All participants received a standard paracervical block. The patients in the intravenous group had significantly lower intraoperative pain scores on a 100-mm scale. Pain scores of participants in the oral oxycodone arm were comparable to published results for local anesthesia and nonsteroidal anti-inflammatory analgesics, suggesting that the oral opioid is no better than local anesthesia alone.
A 2011 systematic review and meta-analysis of pain management for office gynecologic procedures (hysteroscopy, hysterosalpingography, sonohysterography, and endometrial ablation) identified only one placebo-controlled randomized controlled trial evaluating an opioid medication.17 This study of sublingual buprenorphine for hysteroscopy concluded that this medication did not decrease pain but substantially increased side effects including nausea, vomiting, and drowsiness.18
One limitation of our study design is that all participants received 800 mg oral ibuprofen and 2 mg oral lorazepam for analgesia and anxiolysis. In addition, all participants received a standard paracervical block. It is possible that hydrocodone–acetaminophen, if given alone, would control pain during abortion. However, the addition of hydrocodone–acetaminophen was hypothesized to decrease patient perception of pain when added to the standard regimen because hydrocodone and acetaminophen target different pain pathways. A multimodal approach to pain management, with different medications targeting different pathways, has been shown to be effective in other surgical settings.19,20 Moreover, in the case of cervical and uterine pain, direct conduction blockade with a paracervical block and prevention of prostaglandin production in the periphery with ibuprofen likely are more effective strategies.
Our participants were less than 11 weeks of gestation, because patients at this gestational age or greater receive misoprostol cervical priming at our study site. This may limit the generalizability of the results to patients in the later first trimester. However, given that no significant difference was observed between study groups at any procedural time point, and that there were no trends of decreased pain among participants at any gestational age in the hydrocodone–acetaminophen group, it is unlikely that participants between 11 and 14 weeks of gestation would benefit from this medication for procedural pain, although there might be benefit for managing pain related to misoprostol.
It is possible that the opioid dose in our study is too low to affect acute procedural pain. However, 10 mg hydrocodone is a standard oral dose, and higher doses would be more likely to cause excess sedation and other side effects. Although the biological mechanism is unknown, opioids in general may be less effective than other modalities for acute uterine and cervical pain. Women experience significant pain during surgical abortion, and few oral medications have been shown to be effective. Other pain management strategies must be studied to improve patient care in this setting.
1. Sedgh G, Singh S, Shah IH, Ahman E, Henshaw SK, Bankole A. Induced abortion: incidence and trends worldwide from 1995 to 2008. Lancet 2012;379:625–32.
2. O'Connell K, Jones HE, Simon M, Saporta V, Paul M, Lichtenberg ES; National Abortion Federation Members. First-trimester surgical abortion practices: a survey of National Abortion Federation members. Contraception 2009;79:385–92.
3. Kan AS, Ng EH, Ho PC. The role and comparison of two techniques of paracervical block for pain relief during suction evacuation for first-trimester pregnancy termination. Contraception 2004;70:159–63.
4. Renner RM, Jensen JT, Nichols MD, Edelman A. Pain control in first trimester surgical abortion. The Cochrane Database of Systematic Reviews 2009, Issue 2. Art. No.: CD006712. DOI: 10.1002/14651858.CD006712.pub2.
5. Allen RH, Kumar D, Fitzmaurice G, Lifford KL, Goldberg AB. Pain management of first-trimester surgical abortion: effects of selection of local anesthesia with and without lorazepam or intravenous sedation. Contraception 2006;74:407–13.
6. Gomes T, Juurlink DN, Dhalla IA, Mailis-Gagnon A, Paterson JM, Mamdani MM. Trends in opioid use and dosing among socio-economically disadvantaged patients. Open Med 2011;5:e13–22.
7. Edelman A, Nichols MD, Leclair C, Astley S, Shy K, Jensen JT. Intrauterine lidocaine infusion for pain management in first-trimester abortions. Obstet Gynecol 2004;103:1267–72.
8. Edelman A, Nichols MD, Leclair C, Jensen JT. Four percent intrauterine lidocaine infusion for pain management in first-trimester abortions. Obstet Gynecol 2006;107:269–75.
9. Belanger E, Melzack R, Lauzon P. Pain of first-trimester abortion: a study of psychosocial and medical predictors. Pain 1989;36:339–50.
10. World Health Organization. Safe abortion: technical and policy guidance for health systems. Geneva (Switzerland): World Health Organization; 2003.
11. Renner RM, Nichols MD, Jensen JT, Li H, Edelman AB. Paracervical block for pain control in first-trimester surgical abortion: a randomized controlled trial. Obstet Gynecol 2012;119:1030–7.
12. Jensen MP, Chen C, Brugger AM. Interpretation of visual analog scale ratings and change scores: a reanalysis of two clinical trials of postoperative pain. J Pain 2003;4:407–14.
13. Todd KH, Funk KG, Funk JP, Bonacci R. Clinical significance of reported changes in pain severity. Ann Emerg Med 1996;27:485–9.
14. Rowbotham MC. What is a ‘clinically meaningful’ reduction in pain? Pain 2001;94:131–2.
15. Romero I, Turok D, Gilliam M. A randomized trial of tramadol versus ibuprofen as an adjunct to pain control during vacuum aspiration abortion. Contraception 2008;77:56–9.
16. Allen RH, Fitzmaurice G, Lifford KL, Lasic M, Goldberg AB. Oral compared with intravenous sedation for first-trimester surgical abortion: a randomized controlled trial. Obstet Gynecol 2009;113:276–83.
17. Ahmad G, Attarbashi S, O'Flynn H, Watson AJ. Pain relief in office gynaecology: a systematic review and meta-analysis. Eur J Obstet Gynecol Reprod Biol 2011;155:3–13.
18. Lin YH, Hwang JL, Huang LW, Chen HJ. Use of sublingual buprenorphine for pain relief in office hysteroscopy. J Minim Invasive Gynecol 2005;12:347–50.
19. Bauchat JR, Higgins N, Wojciechowski KG, McCarthy RJ, Toledo P, Wong CA. Low-dose ketamine with multimodal postcesarean delivery analgesia: a randomized controlled trial. Int J Obstet Anesth 2011;20:3–9.
© 2012 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
20. Costantini R, Affaitati G, Fabrizio A, Giamberardino MA. Controlling pain in the post-operative setting. Int J Clin Pharmacol Ther 2011;49:116–27.