Pain can be a barrier to the acceptance of intrauterine devices (IUDs). Despite the brevity of the procedure, pain with IUD insertion can be moderate to severe, especially for nulliparous women. A recent trial reported a median pain score of 50 mm for women receiving placebo medication and as high as 80 mm for the subset of nulliparous women (on a 100-mm visual analog scale [VAS]).1 Currently, there is no standard pain regimen for women undergoing IUD insertion.
Over-the-counter analgesics may be offered for pain management and have several advantages, including patient-controlled administration, low cost, and no interruptions in clinic flow when taken before the visit. The most commonly studied over-the-counter analgesics are nonsteroidal anti-inflammatory drugs (NSAIDs). Unfortunately, the popular NSAID ibuprofen has not been shown to reduce pain scores with IUD insertion in four randomized trials.2–6
Naproxen sodium, another oral NSAID, has been associated with a reduction in pain during insertion of the Multiload-Cu 375 IUD and after insertion of the Dalkon Shield.7,8 However, it is not clear whether these results can be extrapolated to current IUDs in the United States, because neither is T-shaped and pain with insertion might differ by IUD type, shape, and mechanism of deployment. This trial tests our hypothesis that naproxen sodium similarly reduces pain with insertion of T-shaped IUDs (primary outcome) and tenaculum placement, uterine sounding, and 5 and 15 minutes after insertion (secondary outcomes).
MATERIALS AND METHODS
This was a double-blind, randomized, placebo-controlled trial. It was approved by the Partners institutional review board. Women were recruited from the Planned Parenthood League of Massachusetts Health Center in Boston, Massachusetts, between May 11, 2015, and March 25, 2016. A research assistant approached women either over the phone in advance of their scheduled IUD insertion appointment or in person on the day of the procedure. Women who expressed interest in the study were prescreened using a standardized prescreening form. For those ineligible or declining participation, the reason was documented and nonidentifiable basic demographic information was extracted from the medical records.
Women were eligible if they were 18 years or older, premenopausal, presenting for insertion of any IUD type available in the clinic (CuT380A [Paragard], LNG20 [Mirena or Liletta]), and English-speaking or non–English-speaking with an appropriate translator available. Initially, we recruited parous and nulliparous women. However, after 5 months of recruitment, only three parous women were enrolled. Therefore, starting October 2015 recruitment was limited to nulliparous women. Women were excluded from the trial if they were currently pregnant or had been pregnant within the previous 4 weeks, were not eligible for IUD insertion per Planned Parenthood League of Massachusetts clinical protocols, were presenting for IUD removal and reinsertion, had taken any pain medications within 12 hours of enrollment or long-acting narcotics (ie, methadone, suboxone) within 48 hours of enrollment, had taken misoprostol within 24 hours of enrollment, or had any known allergy or contraindication to NSAIDs (including active renal disease, active hepatic disease, gastric ulcer disease or gastritis, and bleeding disorders). All women who were eligible and agreed to proceed with study participation provided written informed consent. Once consent was obtained, the clinician placing the IUD reviewed each woman's medical history and medical record to confirm eligibility for IUD insertion and study participation.
The randomization scheme was computer-generated using varying block sizes of 6, 8, and 12. A person unaffiliated with the study prepared opaque randomization envelopes, which were sequentially numbered with study IDs in advance of the study start. Once prepared, all opaque randomization envelopes were sealed and kept in a secure location in the research department at Planned Parenthood League of Massachusetts before and throughout the study. Once a participant was enrolled, the research assistant gave the next sequentially numbered concealed randomization envelope to a nurse who was not directly involved in the participant’s procedure and not otherwise involved in the study. The envelope contained either the naproxen sodium (550 mg) or placebo (100 mg vitamin B6) in a labeled blister pack. Because the blister packs were labeled, the nurse was not blinded. In a private location remote from the participant, the nurse removed the pill from the blister pack and placed it in a medicine cup. In another private space, she administered it to the patient without revealing study group allocation to the participant, health care provider, or research assistant. The pills were similar but not identical in appearance. All participants waited at least 1 hour and no more than 1.5 hours before undergoing their procedures, because the onset of analgesia for naproxen sodium starts at 30 minutes and reaches peak concentration in 1–2 hours.9
Baseline demographics and relevant medical history were collected, including age; race; ethnicity; primary language; education; height; weight; parity; history of cervical procedures including loop electrosurgical excision procedure; history of depression, anxiety, or other psychiatric disorders; history of a chronic pain condition; and drug and alcohol use in the past 3 months. We also administered a baseline assessment of depression, anxiety, stress, and pain tolerance and measured anticipated pain before study drug administration. For nonparticipants, we collected information on ethnicity, race, primary language, marital status, and highest level of education.
Our primary outcome was pain score with IUD insertion as measured on a 100-mm VAS immediately after IUD placement. Participants were asked to mark the VAS to indicate their level of pain (anchors: 0 mm=no pain, 100 mm=worst pain ever). Secondary outcomes included VAS pain scores for tenaculum placement, uterine sounding, and 5 and 15 minutes postinsertion. Participants were assessed for subjective pain on a scale of none, mild, moderate, or severe immediately after IUD insertion and at 15 minutes postprocedure. Participants were also assessed 15 minutes postprocedure for adverse effects (nausea, vomiting, heartburn, dizziness, and drowsiness) and overall satisfaction with pain control and the procedure. Participants were offered acetaminophen at 15 minutes and uptake was recorded. Health care providers completed a questionnaire, which recorded their level of training, type of IUD placed, purpose of the IUD placement (ie, for contraception or other indication), uterine position, need for cervical dilation, and any acute complications.
Our sample size was calculated to detect a difference of 15 mm on a 100-mm VAS. A clinically significant difference in pain scores has been defined as a minimum of 9–14 mm.10–13 We assumed 80% power, a 5% α error rate, and allowed for 10% participant discontinuation, missing data, or both, and assumed a nonnormal distribution. We initially planned to include parous and nulliparous participants given differences in expected pain scores by parity. We completed separate power calculations to ensure adequate power to detect an effect in each of these groups. Based on recent available data, a standard deviation of 22 mm was used for nulliparous women and 21 mm was used for parous women.1,7 Our initial calculation gave a sample size of 84 nulliparous participants and 76 parous participants. After slow enrollment in the parous arm (three participants over 5 months), we discontinued recruitment of parous women and expanded nulliparous recruitment to allow for other analyses. Our final sample size included 116 nulliparous participants, which increased our power to 93% for our primary outcome among nulliparous women.
We performed statistical analyses with parity strata combined using intention-to-treat analysis. We additionally performed statistical analyses for the primary and secondary VAS pain scores for nulliparous women only. Demographics and survey data were compared using the t test or Wilcoxon two-sample test for continuous variables and χ2 or Fisher exact test for categorical variables. The VAS pain scores were not normally distributed; therefore, the Wilcoxon two-sample test was used to compare median pain scores. Jonckheere-Terpstra tests were used to compare subjective pain. We performed a post hoc stratified analysis of VAS pain scores by marijuana use. In addition, we performed a secondary analysis for nulliparous women, comparing pain scores by IUD type using the Wilcoxon two-sample test and weighted least square regression. The data were analyzed using SAS 9.4.
A total of 119 women were enrolled (59 in the naproxen sodium arm, 60 in the placebo arm, 97% nulliparous). One woman was excluded from the naproxen sodium arm because she did not undergo IUD insertion; therefore, a total of 118 participants were analyzed (Fig. 1). There were no differences in baseline demographics, reproductive characteristics, alcohol use, or history of narcotic use. No participants reported prior cervical procedures. Baseline depression, anxiety, stress, and pain measures including anticipated pain were not significantly different between the two groups. There was more reported marijuana use in the past 3 months in the naproxen arm (55% naproxen compared with 30% placebo, P=.006). Baseline characteristics are shown in Table 1. There were no differences between study participants and screened nonparticipants in terms of primary language, marital status, or education, but there was a difference in the distribution of race (African American 1.7% compared with 9.6%, white 78.8% compared with 71.2%, Asian 10.2% compared with 7.8%, and multiracial 4.2% compared with 0.9% for participants compared with nonparticipants, respectively, P=.02).
There was no difference in the primary outcome of VAS pain score with IUD insertion between the naproxen sodium arm compared with the placebo arm (median VAS 69 mm compared with 66 mm, P=.89). Secondary VAS pain scores were also not different between groups for tenaculum placement (37.3 compared with 31.8 mm, P=.97) and uterine sounding (59.8 compared with 58 mm, P=.66). Visual analog scale pain scores were significantly lower at 5 minutes (16.5 compared with 26 mm, P=.01) and 15 minutes (12.8 compared with 24 mm, P=.01) after IUD insertion (Fig. 2; Table 2). Visual analog scale scores and the statistical significance of comparisons at all time points studied were similar in our analysis of nulliparous women only. Subjective pain ratings for IUD insertion were not different between the two groups (P=.94) with 63.8% (37/58) and 62.1% (36/58) reporting moderate pain and 27.6% (16/58) and 27.6% (16/58) reporting severe pain in the naproxen sodium and placebo arms, respectively. Acetaminophen uptake at 15 minutes was 36% (21/58) in the naproxen arm compared with 52% (31/60) in the placebo arm (P=.09).
There was no difference in the level of training of the health care provider, type of IUD placed, purpose of the IUD placement, uterine position, or need for cervical dilation between arms (Table 3). There were no differences in patient-reported adverse reactions or health care provider-reported acute complications (Table 4). Overall satisfaction for the IUD insertion procedure was high. Satisfaction VAS scores for pain control were no different between groups (Table 5). Of note, we were not powered to detect differences in the procedural data or the measures of adverse events and satisfaction.
We found that naproxen sodium did not reduce pain scores at the time of IUD insertion, although it did as soon as 5 minutes after insertion. These findings are consistent with the Massey et al trial, which found reduced pain after insertion of the Dalkon Shield among women premedicated with naproxen sodium (300 mg). The results are also consistent with trials of ibuprofen with no reduction in pain during insertion of modern T-shaped IUDs.2–5 None of the studies of ibuprofen evaluated pain immediately after insertion; therefore, it is unclear whether ibuprofen has a similar benefit to what we observed with naproxen sodium.
We believe that our study used appropriate timing and dosing for analgesic effect. Our wait time of at least 1 hour should have been adequate to achieve peak levels at the time of IUD insertion. In addition, naproxen sodium has up to a 12-hour duration of analgesia; therefore, we would expect postinsertion relief to be prolonged.9 We used prescription-strength dosing (550 mg) for this trial because that dose was found to be effective for insertion of the Multiload Cu 375 IUD. To increase generalizability to various clinical settings, patients may be instructed to take over-the-counter naproxen sodium (two tablets, 440 mg total) or offered clinic stocked naproxen sodium on arrival.7 Although it is possible oral NSAIDs do not have adequate potency to treat the degree of pain experienced with insertion of T-shaped IUDs, they can still reduce postinsertion pain. Additional pain relief with IUD insertion may be achieved by use of a local anesthetic.14
We had originally powered our sample size to evaluate nulliparous and parous women separately, because nulliparous women report higher pain scores.1,3,5 However, we were unable to meet recruitment goals for parous women at 5 months. With rapid enrollment in the nulliparous arm, we expanded enrollment of this arm, which resulted in greater power to detect differences in VAS pain scores and power to perform a secondary analysis comparing pain scores among the three IUD types used in this trial (to be published separately). Despite the increase in our sample size, we were likely underpowered for comparisons such as postprocedural acetaminophen uptake, for which we would have expected higher uptake in the placebo arm.
The results of the study are generalizable to nulliparous women. Limitations include the low proportion of parous women. Moreover, although study participants and nonparticipants had similar demographics, less African American women chose to participate in the study. African Americans may be less willing to participate in research related to contraception and reproduction for historic reasons; however, a review from 2006 did not find that African Americans were significantly less likely to participate in health care research in general.15,16 Our randomization was successful for all demographic, reproductive, and mental health parameters. However, randomization was not successful for use of marijuana over the past 3 months. Marijuana use could have an effect in either direction. It may be an indicator of lower pain tolerance for women who use it for chronic pain or anxiety. For recreational users, it may improve anxiety, which may help reduce reported pain. We performed a stratified analysis of VAS pain scores by recent marijuana use and did not find a difference; thus, it was not an actual confounder.
A randomized trial of ketorolac administered intramuscularly 30 minutes before IUD insertion showed similar results to this trial. However, a subgroup analysis of nulliparous women (n=16) showed a reduction in pain scores with IUD insertion. Given the small sample size of this subgroup, the statistical analysis was too unstable to make definitive recommendations for the use of ketorolac in nulliparous women. Although oral NSAIDs may not have the potency to reduce pain with IUD insertion, the pilot results from the ketorolac subgroup analysis suggest potential for intramuscular NSAIDs. This must be weighed against the pain associated with ketorolac injection, additional cost of the injection, and logistics of administration.1
For reducing immediate postinsertion pain in practice, health care providers can prescribe 550 mg naproxen sodium or instruct patients to take two over-the-counter-strength (220 mg) tablets (Aleve) before IUD insertion. Clinics might also consider stocking prescription-strength naproxen sodium for women who are unable to take the medication before arrival. Naproxen sodium is inexpensive and widely available in generic form, which allows ease of widespread use. The availability of an effective method of pain control may help reduce the fear of pain associated with IUD insertion and increase access to this highly effective method of contraception.
1. Ngo LL, Ward KK, Mody SK. Ketorolac for pain control with intrauterine device placement: a randomized controlled trial. Obstet Gynecol 2015;126:29–36.
2. Chor J, Bregand-White J, Golobof A, Harwood B, Cowett A. Ibuprofen prophylaxis for levonorgestrel-releasing intrauterine system insertion: a randomized controlled trial. Contraception 2012;85:558–62.
3. Bednarek PH, Creinin MD, Reeves MF, Cwiak C, Espey E, Jensen JT, et al. Prophylactic ibuprofen does not improve pain with IUD insertion: a randomized trial. Contraception 2015;91:193–7.
4. Jensen HH, Blaabjerg J, Lyndrup J. Prophylactic use of prostaglandin synthesis inhibitors in connection with IUD insertion [in Danish]. Ugeskr Laeger 1998;160:6958–61.
5. Hubacher D, Reyes V, Lillo S, Zepeda A, Chen PL, Croxatto H. Pain from copper intrauterine device insertion: randomized trial of prophylactic ibuprofen. Am J Obstet Gynecol 2006;195:1272–7.
6. Grimes DA, Hubacher D, Lopez LM, Schulz KF. Non-steroidal anti-inflammatory drugs for heavy bleeding or pain associated with intrauterine-device use. The Cochrane Database of Systematic Reviews 2006, Issue 4. Art. No.: CD006034. DOI: 10.1002/14651858.CD006034.pub2.
7. Karabayirli S, Ayrim AA, Muslu B. Comparison of the analgesic effects of oral tramadol and naproxen sodium on pain relief during IUD insertion. J Minim Invasive Gynecol 2012;19:581–4.
8. Massey SE, Varady JC, Henzl MR. Pain relief with naproxen following insertion of an intrauterine device. J Reprod Med 1974;13:226–31.
9. Naproxen. Micromedex, 2.0. Greenwood Village (CO): Thomson Micromedex. Retrieved October 20, 2014.
10. Rowbotham MC. What is a ‘clinically meaningful’ reduction in pain? Pain 2001;94:131–2.
11. Todd KH, Funk KG, Funk JP, Bonacci R. Clinical significance of reported changes in pain severity. Ann Emerg Med 1996;27:485–9.
12. 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.
13. 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.
14. Lopez LM, Bernholc A, Zeng Y, Allen RH, Bartz D, O'Brien PA, et al. Interventions for pain with intrauterine device insertion. The Cochrane Database of Systematic Reviews 2015, Issue 7. Art. No.: CD007373. DOI: 10.1002/14651858.CD007373.pub3.
15. Kennedy BR, Mathis CC, Woods AK. African Americans and their distrust of the health care system: healthcare for diverse populations. J Cult Divers 2007;14:56–60.
© 2016 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
16. Wendler D, Kington R, Madans J, Van Wye G, Christ-Schmidt H, Pratt LA, et al. Are racial and ethnic minorities less willing to participate in health research? PLoS Med 2006;3:e19.