Periprosthetic Anesthetic for Postoperative Pain After Laparoscopic Ventral Hernia Repair: A Randomized Clinical Trial

Importance  Laparoscopic ventral hernia repair (LVHR) using mesh is a well-established intervention for ventral hernia, but pain control can be challenging.

Objective  To determine whether instillation of a long-acting local anesthetic between the mesh and the peritoneum after LVHR reduces pain or narcotic requirements.

Design, Setting, and Participants  A prospective, double-blind, randomized clinical trial with data collection during a brief hospital stay in a tertiary care, community teaching hospital over 3 years between December 15, 2011, and March 28, 2014. Of 120 screened patients undergoing LVHR in this intention-to-treat analysis, 99 eligible patients were randomized. Forty-two patients received the study drug, and 38 patients received placebo. Patients with a history of chronic narcotic use were excluded.

Intervention  After mesh placement, a long-acting local anesthetic (bupivacaine hydrochloride, 0.50%) or placebo (0.9% normal saline) was injected between the mesh and the peritoneum.

Main Outcomes and Measures  Postoperative pain (on a standard scale ranging from 0 to 10), and narcotic medication use (intravenous morphine equivalents). There were no adverse events.

Results  Baseline and operative characteristics were similar except that the treatment group was older (61.8 vs 52.3 years, P = .001). After surgery, pain scores in the recovery room (3.2 vs 4.7, P = .003), interval total narcotic use (6.7 vs 12.5 mg, P = .003 at <4 hours and 0 vs 2.7 mg, P = .01 at 8-12 hours), and total intravenous narcotic use (9.2 vs 17.2 mg of morphine sulfate equivalents, P = .03) were significantly less in the treatment group.

Conclusions and Relevance  Administration of a long-acting local anesthetic between the mesh and the peritoneum significantly reduces postoperative pain and narcotic use after LVHR.

Trial Registration  clinicaltrials.gov Identifier: NCT01530815

Laparoscopic ventral hernia repair (LVHR) has become a well-established intervention for ventral hernias. Advantages of laparoscopic repair include decreased hospital length of stay and fewer wound infections, with similar low rates of hernia recurrence compared with open repair.^1,2 However, postoperative pain may be equivalent to that of the open approach³ and continues to be a concern and a possible drawback to LVHR.

Intense pain immediately following LVHR is common and results in a 2-day hospitalization on average for pain management.² This pain has been attributed to the sutures and tacks used to fix the mesh to the anterior abdominal wall. Methods for controlling the pain, including lidocaine, 5%, patches applied to the abdomen⁴ and local anesthetic injected during surgery to suture sites,⁵ have been evaluated, with varying success demonstrated. Other studies have evaluated local anesthetic delivered via catheter to the surgical site⁶ and local anesthetic injected to specific sites of pain after surgery,⁷ with mixed results and with concern about increased cost plus risk of infection at the surgical site.

To address the need for improved pain control after LVHR, we proposed that intraoperative delivery of a long-acting local analgesic, bupivacaine hydrochloride, 0.50%, to the operative site under direct visualization would improve postoperative pain control. Therefore, narcotic requirements would be reduced, without the safety concerns or cost associated with more invasive delivery methods of pain medication.

After Cottage Health System Institutional Review Board approval, all patients undergoing LVHR at Santa Barbara Cottage Hospital between December 15, 2011, and March 28, 2014, were screened for trial eligibility. Individuals were eligible if they were 18 years or older, undergoing LVHR with no other planned procedure, and agreeable to completing a postoperative pain journal. Patients were excluded from the trial if they had an allergy or sensitivity to bupivacaine, had a history of chronic pain requiring daily opiates, were cognitively unable to complete the postoperative pain journal, or underwent an additional unplanned procedure.

After written informed consent was obtained, patients’ medical records were reviewed for their medical and surgical history. Patients were prospectively randomized via a generated randomization sheet⁸ in a double-blind fashion to receive the study drug (bupivacaine hydrochloride, 0.50%, in the treatment group) or placebo (0.9% normal saline in the control group). Participants, surgeons (M.Z. and D.S.T.), and caregivers assessing outcomes were all blinded to the intervention. Laparoscopic ventral hernia repair was performed by a standardized approach,⁹ including the use of preoperative antibiotics, reduction of hernia sac contents, and complete adhesiolysis. All port sites were infiltrated with 4 mL of 0.50% bupivacaine before their placement. A permanent mesh treated to decrease visceral adhesions was chosen to allow for at least a 4-cm overlap of mesh and fascia circumferentially. In most cases, the fascial defect was closed with permanent suture before mesh placement. Mesh fixation was achieved by a combination of tacks and transfascial sutures as determined by surgeon preference. On conclusion of the operation and before desufflation, the study drug (bupivacaine or placebo) was instilled under direct visualization into the space between the mesh and the parietal peritoneum to coat the fixation points and layer the study drug on top of the mesh. A proportional dose of study drug was determined by the size of the mesh: each patient received 1 mL of study drug per centimeter of mesh in its greatest dimension. If patients had multiple hernia defects, the total hernia size was calculated as the sum of the area of both hernias.

In the postanesthesia care unit (PACU), pain was assessed and recorded by nursing staff using a 7-point or 10-point visual analog scale. The PACU pain score was calculated as the mean of all pain scores obtained in the PACU. When participants were transferred to the ward, they self-reported pain using a standard 10-point numerical rating scale every 30 minutes for the first 90 minutes, every hour for the next 24 hours, and then every 4 hours until discharge. At each of the intervals, study participants rated their pain, recorded the value, and noted if they were resting, sitting, walking, or coughing. Ward pain scores were calculated as the mean of pain scores obtained during the first 90 minutes after leaving the PACU and then over each subsequent 3-hour interval.

In the PACU, patients received intravenous (IV) narcotic medication, limited to combinations of fentanyl citrate, morphine sulfate, and hydromorphone hydrochloride. If meperidine hydrochloride was administered for shivering, its narcotic dose was also included. When patients tolerated oral intake, they were transitioned to oral analgesia in the form of hydrocodone bitartrate–acetaminophen. While surgical patients often also receive and benefit from nonsteroidal anti-inflammatory drugs, their administration in this study was prohibited to avoid confounding variables. Doses of IV and oral narcotics were converted using a validated conversion equation to IV morphine equivalents (IVMEs).¹⁰ The mean pain medication requirements were determined from the total of IVMEs used over 4-hour intervals.

Sample size estimation showed that 40 patients were needed in each group to detect a difference of 20% in pain medication use in the first 4 hours after surgery between the treatment and control groups, with a power of .95 and a level of significance of .05. Continuous data were summarized as the mean (SD) or the median (interquartile range). Discrete data were summarized using the number in the group (percentage of the group). Two-sided unpaired t test, Mann-Whitney test, or χ² test was used where appropriate. Statistical significance was set at P < .05. Analysis was performed in a statistical computing environment (R; The R Foundation). The study protocol can be found in the trial protocol in the Supplement.

During the study period, 120 patients were assessed for study eligibility (Figure). Of those assessed for the trial, 99 patients were eligible, provided informed consent for the study, and were prospectively randomized. During the course of the study, 19 patients were excluded from the final analysis, most frequently because they did not receive the study drug (n = 5) or because more than 1 procedure was performed (n = 5). In total, 80 patients with incisional, umbilical, and epigastric hernias completed the study and were included in the final analysis as the control group (38 receiving 9% normal saline) and the treatment group (42 receiving bupivacaine, 0.50%). Although 5 surgeons participated in the study, 74% (59 of 80) of the patients were operated on by 2 of us. Each surgeon had cases distributed evenly between the treatment and control groups. The trial ended when the predetermined enrollment goal was reached.

The patients in the control group were significantly younger (52.3 vs 61.8 years, P = .001), but they otherwise had characteristics similar to those of the treatment group, including sex and type of hernia repaired (Table 1). Most hernias were reported as incisional type (n = 39), with fewer being umbilical (n = 33) or epigastric (n = 19). The median operative details for the control vs treatment groups were also similar, including surgery length (58.5 vs 58.0 minutes, P = .64), PACU length of stay (117 vs 113 minutes, P = .91), number of tacks used (30.7 vs 32.5, P = .52), hernia size (4.0 vs 4.2 cm², P = .65), mesh size (176.7 vs 150.0 cm², P = .60), and volume of study drug (15 vs 15 mL, P = .97). There were no significant differences among the 5 surgeons’ patients regarding hernia size or mesh size. There was a significant difference in the number of tacks used by the different surgeons, ranging from 19.6 to 33.9 (P = .02).

The study outcomes were postoperative pain and narcotic medication use. These results are summarized in Table 2.

The 7-point and 10-point visual analog scales used in the PACU were normalized to a 10-point scale to allow for overall comparison of the PACU and ward pain scores. Patients who received the study drug had significantly lower mean pain scores during their PACU stay (3.2 vs 4.7, P = .003); pain scores at later intervals were not significantly different. Although the patients in the wards noted their level of activity during pain assessment (resting, sitting, walking, or coughing), there were insufficient data to evaluate pain during activity separately. Given that the treatment group was older than the control group, the relationship between age and postoperative pain was evaluated, and there was no correlation found between age and pain in the treatment group or the control group. Although the surgeons used significantly different numbers of tacks, an evaluation of PACU pain scores stratified by individual surgeon showed no significant difference across the 5 groups.

Patients in the treatment group required significantly less IV narcotic medication than patients in the control group at less than 4 hours after surgery (6.7 vs 11.6 mg of IVMEs, P = .004). The treatment group also required significantly less of all narcotics during the 8 to 12 hours after surgery (0 vs 2.7 mg of IVMEs, P = .01). With oral medication included in the analysis of the first 4 hours, the total narcotic medication requirement (IV plus oral) remained significantly less in the treatment group (6.7 vs 12.5 mg of IVMEs, P = .003). Beyond the initial 4 hours after surgery, narcotic use was similar between the 2 groups. The total volume of IV narcotics used during the hospital stay was significantly less in the treatment group (9.2 vs 17.2 mg of IVMEs, P = .03). There was no difference in the total narcotic use (IV plus oral) during the hospital stay (19.5 vs 25.2 mg of IVMEs, P = .15). In the control group, pain medication used in the first 4 hours represented half of all the total narcotics used (12.5 of 25.5 mg of IVMEs). Again, because individual surgeons used different numbers of tacks, the amount of pain medications used per each of the surgeon’s patients was examined, and no significant difference was found between the treatment and control groups. There were no adverse events or unintended effects.

This study demonstrated that a simple perioperative injection of bupivacaine between the mesh and the parietal peritoneum in LVHR significantly decreases early postoperative pain and total (IV plus oral) narcotic medications during the first 4 hours after surgery. This is consistent with the immediate but transient effect of bupivacaine, with known peak levels by 30 to 45 minutes and a half-life of 2.7 hours.¹¹

Because IV pain medications are used before the tolerance of oral intake and narcotic pills, the finding that the treatment group required less IV narcotics during the hospital stay supports the conclusion that the intervention’s greatest value is in the early postoperative period. Almost half of the control group’s total number of narcotics were administered in the first 4 hours after surgery. Because the distribution of narcotic medication requirements is skewed toward the immediate postoperative period, this interval likely offers the greatest opportunity for an intervention.

Rather than addressing palliation, many earlier studies sought to identify the etiology of pain after LVHR by evaluating methods of mesh fixation as a source of significant postoperative pain. Helical tack placement may cause irritation of muscle and peritoneum, and transfascial sutures may lead to tissue pressure and trauma.⁵ Variations in these mesh fixation methods have been widely evaluated, demonstrating increased pain with helical tacks¹² and with transfascial sutures¹³ and no differences in outcome measures between the 2 systems.^14,15 Similarly, the number of tacks used in mesh fixation has not consistently correlated with the magnitude of postoperative pain.¹⁶ A recent meta-analysis confirmed that no fixation method is clearly superior to another.¹⁷ Given that variability in mesh fixation techniques is unlikely to yield clinically significant results, we chose to focus on a potentially palliative intervention.

The most common treatment for perioperative pain is opiate analgesics. However, their well-documented adverse effects include respiratory depression, nausea, pruritus, constipation, and delayed return of bowel function. While it would be difficult to eliminate postoperative narcotic use entirely, a nonnarcotic adjunct that decreases pain and minimizes opiate requirements is worthy of study.

Previous methods of delivering a long-acting local anesthetic in LVHR have led to promising results, but an ideal system remains to be determined. A randomized but nonblinded trial of lidocaine patches in LVHR demonstrated postoperative pain reduction.⁴ Decreased pain was also noted after perioperative local anesthetic injections through the abdominal wall down to the peritoneum at the site of transfascial sutures.⁵ However, helical tacking sites were not addressed, and narcotic requirements were not assessed.

Local anesthetic catheter-directed infusion has successfully reduced pain after other procedures, including inguinal hernia repair,¹⁸ open colorectal procedures,¹⁹ and nephrectomies,²⁰ compared with standard narcotic strategies. However, an elastomeric pain pump device delivering local anesthetic between the mesh and the peritoneum after LVHR similar to the system used herein was not found to decrease postoperative pain scores or narcotic requirements.⁶ That pain pump device allows for directed delivery of narcotics over time but also may introduce infection to a surgical site containing mesh, with additional concerns regarding cost, difficulties of wound care, need for equipment, leakage of medication, and consistency of delivery.^5,6 We chose to avoid these potential drawbacks by using a single dose of bupivacaine delivered throughout the entire area of mesh and to all sites of abdominal wall fixation.

Laparoscopic ventral hernia repair has become an accepted if not standard approach for the repair of a variety of ventral hernias, with low reported rates of complications and recurrences,²¹ which are comparable if not decreased compared with the traditional open surgery.^1,22 Although LVHR is safe and effective, many patients report considerable pain in the hospital and after surgery for up to 3 months, which can substantially affect recovery and quality of life.^7,16,23 Aggressive pain control must be addressed as an essential component of a successful LVHR, but much of the previous literature on postoperative pain control is focused on open surgery. While some strategies from the large body of literature addressing pain control after laparotomy (eg, continuous wound infusion¹⁹) may have a role in laparoscopic surgery, tailoring the measure of pain control after minimally invasive procedures seems appropriate. To contribute to that effort, this trial studied immediate postoperative pain, and our results confirm that a simple perioperative intervention can decrease pain and narcotic requirements. During the course of this study, the US Food and Drug Administration approved liposomal bupivacaine, a slow-release formulation providing analgesia for up to 72 hours.²⁴ This may present an even better option although it is now very expensive.

There are limitations and caveats to our findings. One particularly worrisome aspect of pain after LVHR is that it can persist past the acute perioperative period for up to several months.²⁵ The present study evaluated pain only until the time of discharge, which was uniformly brief, so we were unable to address whether early, effective pain control affected subsequent pain after discharge. Patients were aware that their pain was being observed, so a Hawthorne effect must be considered. However, because patients and all caregivers were blinded to their group status, any potential influence would be the same in both groups and would not have contributed to the differences observed. Finally, while our treatment group was significantly older than our control group, statistical analysis failed to show a correlation between age and pain or narcotic use in this study.

Laparoscopic ventral hernia repair using mesh is a well-established intervention, but postoperative pain remains a challenge. After mesh placement, a long-acting local anesthetic (bupivacaine) injected between the mesh and the peritoneum significantly reduces early postoperative pain and narcotic use after LVHR.

Accepted for Publication: March 25, 2015.

Corresponding Author: Aimee E. Gough, MD, Department of Surgery, Santa Barbara Cottage Hospital, 400 W Pueblo St, Santa Barbara, CA 93105 (agough@sbch.org).

Published Online: July 8, 2015. doi:10.1001/jamasurg.2015.1530.

Author Contributions: Dr Thoman had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Chang, Thoman.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Gough, Ferrigno, Thoman.

Critical revision of the manuscript for important intellectual content: Thoman.

Statistical analysis: Grotts.

Obtained funding: Chang, Thoman.

Administrative, technical, or material support: All authors.

Study supervision: Thoman.

Conflict of Interest Disclosures: Dr Thoman reported being a paid consultant to Covidien. No other disclosures were reported.

Funding/Support: This research project was supported by a grant from Santa Barbara Cottage Hospital.

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Previous Presentation: This study was presented at the 86th Annual Meeting of the Pacific Coast Surgical Association; February 20, 2015; Monterey, California.

Additional Contributions: Santa Barbara Cottage Hospital physicians Farida Bounoua, MD, Jeffrey Gauvin, MD, and Stephen Kaminski, MD, provided access to their surgical patients. No financial compensation was provided.