Cesarean delivery is the most common inpatient surgical procedure in the United States with 1.3 million cesarean deliveries performed in 2015.1 Prescription opioids are a mainstay of pain management after cesarean delivery. Although opioids are effective, significant side effects exist, including respiratory depression, nausea, vomiting, constipation as well as risks of substance use disorder and diversion. Given the national opioid epidemic, there is an urgent need to identify alternative postoperative pain management strategies to minimize opioid use while maintaining patient satisfaction.
Liposomal bupivacaine allows for controlled release of local anesthetic over time with a duration of action of 48–72 hours. Initially approved for hemorrhoidectomy and bunionectomy, the U.S. Food and Drug Administration broadened its indications to include postsurgical anesthesia in any surgical site.2 One retrospective study of women undergoing cesarean delivery with an incisional block of liposomal bupivacaine, compared with a historical cohort undergoing routine cesarean delivery, demonstrated less total postoperative opioid consumption.3 However, use of nonopioid analgesics was also higher in the treatment group, thus confounding the effect of the intervention.
We designed a single-blind, randomized controlled trial at a tertiary care institution to investigate whether liposomal bupivacaine reduces pain after cesarean delivery. We hypothesized that an incisional block of liposomal bupivacaine distributed in the skin and fascia would result in decreased pain scores with movement at 48 hours postoperatively, as the primary endpoint, and decreased opioid consumption and pain scores at all time points as secondary endpoints.
MATERIALS AND METHODS
This trial was approved by the Partners Human Research Committee. Women older than 18 years of age undergoing scheduled cesarean delivery through a Pfannenstiel incision with planned neuraxial anesthesia were eligible and sequentially identified from the labor and delivery operating room schedule. We excluded women with current or prior opioid use ever (methadone, buprenorphine, or other oral opioids) or with an allergy to local anesthetic. Women who initially received neuraxial anesthesia but required intraoperative conversion to general anesthesia were not excluded.
Primary obstetric care providers approached eligible women to inform them of the ongoing study. If interested in participation, study staff were notified and contacted the woman to describe the study in detail. Informed consent was obtained in the outpatient setting by a physician either over the telephone or in person, and sociodemographic information and responses to the Generalized Anxiety Disorder 7-item scale screening questionnaire were gathered before implementation of the intervention.4 All non–English-speaking patients had an in-person visit with an in-person interpreter for informed consent.
Liposomal bupivacaine is dispensed in a 20-cc vial of 266 mg liposomal bupivacaine, intended for expansion in normal saline, to create an adequate volume to cover the surgical field. The adequate amount of liposomal bupivacaine–normal saline mixture for a Pfannenstiel incision at the time of cesarean delivery is not known; manufacturer recommendations suggest 10 cc for every 1 cm of the incision (Pacira Inc, personal communication). In a prior trial of liposomal bupivacaine incisional block compared with transversus abdominus plane blocks with bupivacaine for abdominal hysterectomy through a Pfannenstiel incision, a total of 60 cc was infiltrated in the field.5 We opted to use 80 cc liposomal bupivacaine–normal saline solution, diluting 20 cc liposomal bupivacaine in 60 cc normal saline, to allow for a larger Pfannenstiel incision at cesarean delivery than at hysterectomy, but did not want to infiltrate 100 cc of fluid into the incision in a population at risk of wound infection. We used 80 cc of normal saline as the placebo solution.
Consenting women were randomized in a one-to-one ratio in blocks of four stratified by cesarean delivery type (primary or repeat) to an 80-cc liposomal bupivacaine (intervention) or 80-cc placebo incisional block. Randomization was performed by the institutional Clinical Trials Pharmacy, and no study investigator had access to the master randomization list. Study participants who presented in labor before their scheduled cesarean delivery date were not included in the study, whether they chose a trial of labor or had an unscheduled cesarean delivery, and their randomization number was used for the next enrolled patients.
For women who consented and were admitted for scheduled cesarean delivery, participation was reconfirmed and the study drug was then retrieved from the pharmacy by the study obstetrician and prepared for injection. Once the surgical team was ready for fascial closure, the study obstetrician infiltrated 40 cc of the study solution into the fascia and 40 cc into the skin (Fig. 1). The edges of the incised fascia were grasped with Kocher clamps. Using one 20-cc syringe, 3–5 cc of solution were injected intrafascially into one lateral corner of the fascial incision, where the abdominal wall innervation enters from, and the rest of the syringe contents was evenly distributed within the upper and lower aspects of the fascial incision, until the midline was reached. This procedure was repeated on the contralateral side with another 20-cc syringe. With the third 20-cc syringe, the skin on the superior aspect was infiltrated with study drug in an axis parallel to the axis of the incision.
All injections were performed by three study obstetricians (M.P., W.H.B., M.A.C.). One study obstetrician (M.P.) had prior experience with incisional injection of liposomal bupivacaine and taught the other two investigators the infiltration technique, confirmed to be in accordance with manufacturer recommendations, into the subfascial and subdermal spaces. Because the clarity of the intervention and placebo solutions is different, the institutional Clinical Trials Pharmacy was unable to match a placebo solution in similar opacity and viscosity. During the surgery, the contents of the syringe were never discussed in the room. However, given the inability to blind, the study obstetrician involved in the drug infiltration did not participate in any of the subsequent outcome assessments, including the pain scores with rest and activity at each 24-hour time point or follow-up phone calls.
During the trial, postcesarean pain management was standardized at our institution as follows: spinal or epidural morphine administered at the time of the cesarean delivery, scheduled 30 mg ketorolac every 6 hours for 24 hours after delivery followed by scheduled 600 mg ibuprofen every 6 hours for 24 hours and scheduled 650 mg acetaminophen every 6 hours for 48 hours after delivery. After 48 hours, 650 mg acetaminophen and 600 mg ibuprofen were administered every 6 hours on an as-needed basis. Oxycodone was the initial oral opioid of choice and was administered as needed throughout the postoperative period with a dose of 5–10 mg every 4 hours as needed. Implementation of this pain management strategy, escalation of pain medications, and withholding of nonopioid analgesics (nonsteroidal antiinflammatory drugs or acetaminophen) if contraindications such as transaminitis, renal dysfunction, or pre-existing intolerance or allergy were present was left to the discretion of the clinical team.
The primary outcome was pain score with movement at 48 hours postoperatively using an 11-point numeric rating scale. Pain with movement was ascertained by instructing patients to sit up from lying flat in bed to engage their abdominal wall. We chose this as the primary outcome because the duration of action for liposomal bupivacaine is between 48 and 72 hours, and movement that requires abdominal wall engagement is usually more painful postoperatively than pain at rest.
Secondary outcomes included pain scores at rest at 24, 48, and 72 hours postoperatively as well as pain scores with movement at 24 and 72 hours. All pain scores were attempted to be obtained within 1 hour of the calendar time that the surgery occurred on the postpartum unit. At 48 hours postoperatively, we administered a validated questionnaire, the International Pain Outcomes Questionnaire, focused on functional effect of surgical pain, opioid-related adverse effects, and satisfaction with surgical pain management.6 All assessments were performed in person using the services of an in-person or telephonic interpreter for non–English-speaking patients. The study obstetrician performing the incisional block and the outcomes assessor were always two different individuals; the study obstetrician performed no postoperative assessments.
Additional prespecified outcomes included inpatient opioid consumption, by 24-hour postoperative time periods and total for the hospitalization, dose of the discharge opioid prescription, hospital length of stay, and adverse reaction to the study drug. All opioid consumption was converted to oral morphine milligram equivalents.7 Charts were abstracted at 2 and 6 weeks postoperatively for wound complications, opioid refills, and postpartum readmissions. Women were also contacted at 6 weeks postoperatively to assess satisfaction with pain management, leftover opioids, and whether they believed they received liposomal bupivacaine or placebo. A standardized telephone interview was performed with a telephonic interpreter as applicable. If we were unable to reach the patient on the first attempt, at most nine additional attempts were made (one daily phone call per weekday for 2 weeks). As described previously, the study obstetrician performed no follow-up phone calls. The full protocol is available in Appendix 1, available online at http://links.lww.com/AOG/B99.
We estimated the expected pain scores at 48 hours postoperatively based on prior unpublished work among women undergoing cesarean deliveries at our institution. To demonstrate a decrease in pain score by 1.5 points from a mean pain score of 5 (with a SD of 2.1), we would need 31 women per group at a power of 80% with a two-sided type I error of 5%. We chose an effect size of 1.5 points because this is considered a clinically significant difference in prior studies evaluating pain scores.8 To account for possible loss to follow-up as well as a lower estimated sample size assuming normality of the primary outcome, we planned to enroll 40 women in each group.
All analyses were performed by intention to treat. Outcomes were compared either with a t test or Wilcoxon test, depending on normality of distributions, and categorical outcomes were compared with a χ2 test. No secondary analyses were prespecified and no interim analyses were performed. All data were securely stored in REDCap, and data analysis was performed in Stata 14.
The study was funded through an investigator-initiated grant (#2016-127) by Pacira, Inc, the pharmaceutical company that manufactures liposomal bupivacaine. Pacira, Inc, provided the study drug, salary support for study technical staff, and institutional overhead. Pacira, Inc, did not play a role in development of the study protocol, in the conduct of the study, or in the data analysis, and no individual patient-level data were shared with Pacira, Inc. Per our contract, representatives from Pacira, Inc, have reviewed the manuscript, but no alterations were requested.
Recruitment began on March 3, 2017, and ended on September 22, 2017. Of the 249 women screened for study eligibility, 228 were eligible, 103 enrolled in the study, and 80 were randomized (Fig. 2). Current or prior use of opioids was the most common reason for ineligibility (81.0%), and lack of interest in the study was the most common reason for not participating (58.7%).
Of those randomized, one patient in the liposomal bupivacaine group had a vertical midline skin incision, study drug was not infiltrated, and study outcomes were not collected. Of the 40 patients in the placebo group, no patients were discharged home before the 48-hour assessment, and four patients were discharged home before the 72-hour outcome assessment. Of the 39 patients in the liposomal bupivacaine group, two patients were discharged home before the 48-hour outcome assessment, and seven were discharged home before the 72-hour assessment. For the purposes of the pain score outcomes, these patients were considered lost to follow-up, although they were attempted to be contacted for responses to other questions.
The baseline characteristics of women in the placebo and liposomal bupivacaine groups were similar (Table 1). The mean (SD) ages were 34.7 (3.7) and 35.6 (4.2) years, respectively, and in both groups, the majority were white and privately insured, consistent with the patient population at this institution, and had similar distributions of comorbidities. Given stratified randomization, the number of primary and repeat cesarean deliveries was balanced across groups. Anesthetic type, administration of neuraxial morphine, operative time, and intraoperative complications were similar between groups.
Pain scores at 24 hours postoperatively were assessed within 1 hour of the time of surgery among 94.9% in the liposomal bupivacaine group and 90.0% in the placebo group. At 48 hours postoperatively, among patients not discharged, 94.6% in the liposomal bupivacaine group and 90.0% in the placebo group were assessed within 1 hour of the time of surgery. At 72 hours postoperatively, among patients not discharged, 93.8% in the liposomal bupivacaine group and 94.4% in the placebo group were assessed within 1 hour of the time of surgery.
The primary outcome, pain score with movement at 48 hours, was not statistically different between the liposomal bupivacaine and placebo groups with median (interquartile range) pain scores of 4 (2–5) and 3.5 (2–5.5) (P=.72), respectively (Table 2). No differences were seen in pain scores at rest or with movement at any time point. Median opioid consumption, in morphine milligram equivalents , was similar in both groups (Table 2). In the first 48 hours postoperatively, median (interquartile range) opioid consumption was 37.5 (7.5–60) morphine milligram equivalents in the liposomal bupivacaine group and 37.5 (15–75) morphine milligram equivalents in the placebo group (P=.44). For the entire postoperative stay, median (interquartile range) opioid consumption was 90 (15–127.5) morphine milligram equivalents in the liposomal bupivacaine group and 101.3 (41.3–165) morphine milligram equivalents in the placebo group (P=.61).
Satisfaction with pain management and postoperative length of stay was also similar between both groups. On discharge from the hospital, women in the liposomal bupivacaine group were discharged with a median of 187.5 (150–225) morphine milligram equivalents, and women in the placebo group were discharged with a median of 183.8 (75–193.8) morphine milligram equivalents, which represents the equivalent of 25 tablets of 5 mg oxycodone for both groups (P=.40). Eight (20.5%) women in the liposomal bupivacaine group and six (15.0%) women in the placebo group were not prescribed opioids on discharge. At the time of discharge, 61.5% of women in the liposomal bupivacaine group and 57.5% of women in the placebo group guessed that their intervention was liposomal bupivacaine, a difference that was not statistically significantly different (P=.93).
No events of local anesthetic toxicity or allergic reaction occurred in this study (Table 3). Opioid-related adverse events were common in both groups, including itching, drowsiness, dizziness, and nausea. Within 14 days of surgery, one patient in the placebo group was noted to have a wound seroma, two patients in the placebo group were noted to have a wound infection or cellulitis, and one patient in the liposomal bupivacaine group was noted to have a wound infection or cellulitis. No patients were readmitted with complications, and one patient in the liposomal bupivacaine group had received an opioid refill since discharge for incisional pain.
At the 6-week follow-up, 97.1% of patients in the placebo group and 91.4% of patients in the liposomal bupivacaine group were satisfied or very satisfied with postoperative pain control as part of this study (P=.59) (Table 4). No patients needed additional opioid prescriptions, and no additional wound complications were noted.
In this trial of a liposomal bupivacaine incisional block compared with placebo at the time of cesarean delivery, we found no differences in pain score with movement at 48 hours postoperatively. These findings are in contrast to prior studies using liposomal bupivacaine.
In a previously published retrospective study of women undergoing cesarean delivery with or without liposomal bupivacaine, Parikh et al3 used an injection technique similar to ours, although the volume of expansion was not reported. The results from their study are difficult to compare, because higher doses of intravenous ketorolac and intravenous acetaminophen were used among women also receiving liposomal bupivacaine. In addition, the retrospective nature of their study introduces the possibility of confounding by indication. More comparable is a randomized trial of women undergoing hysterectomy through a Pfannenstiel incision and receiving a liposomal bupivacaine incisional block or bilateral transversus abdominis plane block with bupivacaine.5 That study demonstrated decreased pain scores by approximately 1.5 points at rest and 2 points with coughing at 48 hours postoperatively and decreased opioid consumption by 8.3 morphine milligram equivalents between 24 and 48 hours postoperatively. In their study, liposomal bupivacaine was infiltrated in the subcutaneous and subfascial spaces, similar to our method, but also in the preperitoneal space, and their postoperative pain management was standardized. However, overall postoperative opioid consumption was much lower after hysterectomy than in our trial, perhaps as a result of differences between the procedures as well as the patients undergoing surgery.
Our trial has several limitations. First, the pain scores used to power this study, derived from institutional data, were higher than those observed in the placebo group of the current study. Whether this speaks to differences in eliciting pain scores as part of a study or a Hawthorne effect, improved multimodal nonopioid analgesia, or the power of the placebo effect is not clear. Moreover, we powered the study on differences in mean pain scores, because our baseline data were normally distributed; however, our data in this trial were not normally distributed, and thus medians were the appropriate summary statistic.
Another limitation was our inability to double-blind the trial. After discussions with the institutional Clinical Trials Pharmacy, there was no alternative, because no viscous, cloudy placebo solution could be created, and opaque syringes could not be prepared in a sterile fashion. For these reasons, the study obstetrician and the outcomes assessor were always two different individuals. Patients remained appropriately blinded, because the proportion of patients who guessed they received liposomal bupivacaine was similar in both groups. We also note a high degree of patient-reported opioid-associated side effects in both groups; however, our validated questionnaire (PAIN-OUT) does not assess severity or functional limitations associated with side effects. Finally, perhaps the placebo effect was more powerful than we anticipated, and a third arm with sham or no injection may prove beneficial to future studies. Alternatively, it is possible that our null findings are the result of inadequate volume expansion of liposomal bupivacaine for the length of the incision or perhaps another incisional technique would provide more pain relief.
Although our findings suggest no benefit to improving pain scores, given the scarcity of the investigation on this subject, future trials investigating the pain-relieving and opioid-sparing potential of liposomal bupivacaine remain important, and the pain score and opioid use data derived from this prospective patient population seeking out other pain- and opioid-reducing strategies will be valuable in powering future trials. We chose to power this trial on pain scores as a primary endpoint, because this is of primary interest to a patient during their postoperative course. Although we did study opioid consumption as a secondary outcome and saw no significant differences overall, this may be an area of specific study in larger trials, because the quartiles of overall morphine consumption appear different, and our study cannot exclude a small benefit. In addition, future studies are needed to evaluate the potential benefit of liposomal bupivacaine for patients with chronic pain or opioid use disorder or those on opioid oral replacement therapy.
We did not aim to decrease opioid-prescribing on discharge or the number of unused opioid tablets with our intervention. However, the wide variation and high quantity of both are also notable, because 18% of patients required no opioids on discharge and 51% had leftover opioids. Because leftover opioids are a source of diversion and possible development of substance use disorder, particularly among at-risk women, a focus on judicious opioid-prescribing and counseling on safe disposal is important.9–11
Our findings are generalizable to opioid-naïve women undergoing uncomplicated cesarean delivery at a tertiary care institution who were motivated to participate in a study to improve pain control and minimize need for opioid consumption. We did not have institutional review board approval to abstract the medical record for women who declined participation. This may limit the generalizability of our study findings.
In conclusion, a liposomal bupivacaine incisional block at the time of cesarean delivery did not decrease postoperative pain scores with movement at 48 hours among opioid-naïve women undergoing scheduled cesarean delivery in this single-blind, randomized controlled trial. Other strategies that might reduce postoperative opioid use while maintaining satisfaction with pain control need to be developed and evaluated in this patient population.
1. Martin JA, Hamilton BE, Osterman MJ, Driscoll AK, Matthews TJ. Births: final data for 2015. Natl Vital Stat Rep 2017;66:1.
2. Pacira. Liposomal bupivacaine. Highlights of prescribing information. Available at: https://www.accessdata.fda.gov/drugsatfda_docs/label/2011/022496s000lbl.pdf
. Retrieved May 1, 2016.
3. Parikh P, Sunesara I, Singh Multani S, Patterson B, Lutz E, Martin JN Jr. Intra-incisional liposomal bupivacaine and its impact on post cesarean analgesia: a retrospective study. J Matern Fetal Neonatal Med 2017:1–5.
4. Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006;166:1092–7.
5. Gasanova I, Alexander J, Ogunnaike B, Hamid C, Rogers D, Minhajuddin A, et al. Transversus abdominis plane block versus surgical site infiltration for pain management after open total abdominal hysterectomy. Anesth Analg 2015;121:1383–8.
6. Rothaug J, Zaslansky R, Schwenkglenks M, Komann M, Allvin R, Backström R, et al. Patients' perception of postoperative pain management: validation of the international pain outcomes (IPO) questionnaire. J Pain 2013;14:1361–70.
7. Von Korff M, Korff MV, Saunders K, Thomas Ray G, Boudreau D, Campbell C, et al. De facto long-term opioid therapy for noncancer pain. Clin J Pain 2008;24:521–7.
8. Olsen MF, Bjerre E, Hansen MD, Hilden J, Landler NE, Tendal B, et al. Pain relief that matters to patients: systematic review of empirical studies assessing the minimum clinically important difference in acute pain. BMC Med 2017;15:35.
9. Bateman BT, Cole NM, Maeda A, Burns SM, Houle TT, Huybrechts KF, et al. Patterns of opioid prescription and use after cesarean delivery. Obstet Gynecol 2017;130:29–35.
10. Osmundson SS, Schornack LA, Grasch JL, Zuckerwise LC, Young JL, Richardson MG. Postdischarge opioid use after cesarean delivery. Obstet Gynecol 2017;130:36–41.
11. Bateman BT, Franklin JM, Bykov K, Avorn J, Shrank WH, Brennan TA, et al. Persistent opioid use following cesarean delivery: patterns and predictors among opioid-naïve women. Am J Obstet Gynecol 2016;215:353.e1–18.