Patient Complications after Interscalene Block: A Retrospective Comparison of Liposomal Bupivacaine to Nonliposomal Bupivacaine

Background The purpose of this study was to investigate if the addition of liposome bupivacaine (LB) to an interscalene block (ISB) had an effect on the number of patients with surgical- or block-related complications. Methods This was a single-center retrospective chart view performed by identifying patients who received an ISB from January 1, 2014, through April 26, 2018, at the University of Minnesota. 1,518 patients were identified who received an ISB (LB = 784, nonliposomal bupivacaine = 734). Patients were divided into two groups those who did receive liposome bupivacaine in their ISB and those who did not receive liposome bupivacaine in their ISB. Medical records were individually reviewed for surgical procedure, block medications, complications related to the block or surgical procedure, phone calls to the healthcare system for issues related to opioids or pain within 3 and within 30 days, readmissions within 30 days, and emergency room visits for complications within 3 and 30 days. Results There was no significant difference in the number of patients with surgical or anesthetic complications. Only phone calls for pain within 3 days were significantly different. The LB group had 3.2% of patients call compared to 5.6% in the nonliposomal bupivacaine group (aOR = 1.71 (95% CI: 1.04–2.87), p=0.036). We found no significant difference in any of the other secondary outcomes. Conclusions The use of LB in an ISB demonstrated no significant difference compared to nonliposomal bupivacaine in numbers of complications, emergency room visits, and readmissions.


Background
Patients commonly experience moderate-to-severe postoperative pain after undergoing shoulder surgery [1]. Traditionally, postoperative pain had been managed with opioids because of their effectiveness in controlling moderate-to-severe pain; however, adverse effects such as nausea, vomiting, and constipation are common [2]. Opioids can lead to persistent use and dependence, and they are a direct cause of respiratory depression, which can lead to overdose and death [3]. Patients are becoming increasingly concerned about their postoperative pain management and the use of opioids [4]. As a result, the use of regional anesthesia techniques such as interscalene blocks (ISBs) is increasing.
Analgesia for shoulder procedures can be safely and effectively achieved through the use of regional anesthesia [5,6]. ISBs are commonly performed in shoulder surgery, which aim to block the proximal trunk of the brachial plexus. ISBs have been shown to be a safe and effective method for analgesia during shoulder procedures [5][6][7][8]. When compared to general anesthesia alone, ISBs have been shown to decrease recovery room stay, nonsurgical intraoperative time, and intraoperative narcotics [9,10]. ISBs have also been shown to result in improved early postoperative pain control, decreased unplanned hospital admissions secondary to severe pain, decreased nausea/vomiting, and decreased duration of hospital stay [8,10,11]. Complications of ISBs include brachial plexus injury, nerve damage, local anesthetic systemic toxicity, Horner's syndrome, and phrenic nerve paralysis [12][13][14][15].
More recently, liposomal bupivacaine (LB) (EXPAREL ® ; Pacira Pharmaceuticals, Inc; Parsippany, NJ, USA) has been developed to increase the duration of analgesia for both infiltration and regional anesthesia. Compared to standard bupivacaine, LB used in ISBs for rotator cuff repair (RCR) or total shoulder arthroplasty (TSA) was associated with no difference in adverse events, decreased pain, and higher patient satisfaction in the first week following surgery [16]. Sun et al. performed a systematic review and found that LB was associated with a lower complication rate when compared to other local anesthetics used in an ISB [17]. e purpose of this study was to determine if there were any differences in the number of patients with surgical-or block-related complications with and without the addition of LB. Additionally, we wanted to understand if the use of liposomal bupivacaine in an ISB impacted the number of readmissions, phone calls, or emergency department (ED) visits for patients undergoing shoulder surgery. We hypothesized that there would be no difference in postoperative complications, pain phone calls, and ED visits between blocks performed with added LB versus those without.

Methods
is was an Institutional Review Board (IRB) exempt study (IRB ID STUDY00003305) as it involved research involving the collection or study of existing data, documents, and records in which the information recorded by the research team was such that the subjects could not be identified. e requirement for written informed consent was waived by the IRB.
is manuscript adheres to the applicable STROBE guidelines. is was a single-center retrospective chart review of patients who received interscalene brachial plexus blocks at the University of Minnesota from January 1, 2014, to April 28, 2018. Inclusion criteria included American Society of Anesthesiologists (ASA) I-III patients 18 years of age or older who received a preoperatively placed interscalene single shot block for postoperative pain. Patients were excluded if they declined access to their medical charts for research purposes when asked on admission, received an interscalene catheter, or had a confirmed diagnosis of chronic pain in their medical record. Electronic medical charts were individually reviewed for surgical procedure, block medications used, complications related to the block or surgical procedure, phone calls to the healthcare system for issues related to opioids or pain within 3 days and within 30 days, readmissions within 30 days, and emergency room visits for pain, opioids, or block-related complications within 3 days and within 30 days. All notes in the patient chart from the day of surgery until 30 days after surgery were reviewed. Opioid-related issues included calls or visits due to side effects from opioids such as constipation, nausea/ vomiting, urinary retention, intolerance of the medication, opioid overdose, and need for more opioid medication were recorded. Block-related complications included but were not limited to shortness of breath, prolonged numbness in the affected extremity, or Horner's syndrome. Patients presenting with pain to the emergency room or reporting pain control as an issue during a phone call were recorded as uncontrolled pain. Demographic data such as age, sex, weight, and ASA status was also documented. Patients were then divided into liposomal bupivacaine and nonliposomal bupivacaine groups for statistical analysis. ose patients in whom the medications used for the block were recorded but volumes of medications used were not recorded were included in the study. ose patients in whom there was no mention of what medication was used for their ISB were excluded in the final analysis. e primary outcome of this study was the number of patients with postoperative block and surgical complications seen in each group. Secondary outcomes were rates of readmission, rates of emergency room visit related and unrelated to the ISB, pain, and opioids, and phone calls to the healthcare system related and unrelated to the ISB, pain, and opioids.

Statistical Analysis.
Continuous measures were summarized as mean (standard deviation), and categorical measures were summarized as count (%) with a two-sample t-test or chi-squared test to detect any significant differences between the liposomal bupivacaine group and nonliposomal bupivacaine groups, respectively. As all outcomes were binary, Firth's penalized logistic regression was used to compare groups to take into account the small sample sizes of events and potential bias of parameter estimates, adjusting for any significant baseline measures (p < 0.05). e p values are not adjusted for multiple comparisons and should be interpreted cautiously. All analyses completed with R version 3.5.1 [18], and Firth's logistic regression used the logistf package [19].

Results
1,869 ISBs were identified to have occurred from January 1, 2014, through April 26, 2018. After applying the inclusion and exclusion criteria, a total of 1,518 patients were included in the analysis. 246 patients were excluded due to not allowing their charts for research, and 105 were excluded due to being under age 18, ASA IV, having a continuous catheter, incomplete data regarding what medication was used for the block, and having a diagnosis of chronic pain in their medical record. 784 patients received liposomal bupivacaine, and 734 patients received a nonliposomal bupivacaine block. e nonliposomal bupivacaine group had significantly more male patients than the liposomal bupivacaine group (p < 0.001). e liposomal bupivacaine group had significantly more ASA 3 and less ASA 1 compared to the nonliposomal bupivacaine group (p � 0.010) All other baseline characteristics were similar (Table 1).
ere was a wide variation in the combination of local anesthetics used in the two groups. In the liposomal bupivacaine group, the combination of liposomal bupivacaine along with 0.25% bupivacaine + epinephrine (N � 229) and liposomal bupivacaine + 0.5% bupivacaine + epinephrine (N � 225) were the most common. e median volume of total local anesthetic used was 20 mL for the liposomal bupivacaine group. e median volume of liposomal bupivacaine used was 10 mL, and median volume of nonliposomal bupivacaine used was 10 mL (Table 2). e patients in the nonliposomal bupivacaine group received either bupivacaine or ropivacaine as the primary agent of their block. e most common combination in this group was 0.5% ropivacaine with clonidine and epinephrine (n � 387). e median volume of the interscalene was 25 mL in this group (Table 3). Table 4 shows the occurrence of outcomes and the odds ratios estimated from Firth logistic regression results for comparisons between the two groups, adjusting for sex and ASA status. ere were 25 patients in the liposomal bupivacaine group with complications and 22 in the nonliposomal bupivacaine group (aOR � 0.95 (95% CI: 0.53-1.70), p � 0.873). ere were 5 patients with cardiopulmonary complications observed in the nonliposomal bupivacaine group (postoperative atrial fibrillation, intraoperative hemodynamic instability, bradycardia, aspiration, ST-segment elevation, and pulmonary embolus). ere were seven patients with postoperative numbness, three postoperative infections, two with uncontrolled nausea/vomiting, one postoperative bleed, one with axillary nerve paresthesias, one patient with postoperative shortness of breath, and one with postoperative delirium.
In the liposomal bupivacaine group, there were eight patients with prolonged postoperative numbness, seven patients with infection, two patients with uncontrolled nausea/vomiting, two patients with shortness of breath, two with retear needing repair, one postoperative bleed, one with a pulmonary embolus, one with acute lower extremity weakness, and one with bilateral hand swelling and weakness.
Of the secondary outcomes evaluated, only calls for pain within 3 days were significant. e liposomal bupivacaine group had 3.2% of patients call compared to 5.6% in the nonliposomal bupivacaine group (aOR � 1.71 (95% CI: 1.04-2.87), p � 0.036). Tables 5 and 6 describe the reasons for emergency room  visits for opioid or pain-related issues within 3 and 30 days,  while Tables 7 and 8 report reason for phone calls over the same time frame. Among both groups, arthroscopic rotator cuff repair (LB 36.0% vs. non-LB 38.6%), total shoulder arthroplasty (LB 15.6% vs. non-LB 14.6%), and reverse total shoulder arthroplasty (LB 11.4% vs. non-LB 9.3%) were the most common procedures performed (Table 9).
Additionally, we explored three sets of subgroup analyses for the results of Table 4 for the block types of bupivacaine 0.25%, bupivacaine 0.25% with epinephrine 1 : 200000, and bupivacaine 0.50% with epinephrine 1 : 200000. Given the focus on each specific combination, the total volume administered was also included in Firth's logistic regression models in addition to sex and ASA status. Given the smaller sample sizes within each subgroup, we generally observed insignificant results for comparing liposomal bupivacaine and the nonliposomal bupivacaine group, with the exception of calls for pain within 30 days for the bupivacaine 0.25% with epinephrine 1 : 200000 subgroup, where the liposomal bupivacaine group had 23.8% of patients call compared to 9.8% in the nonliposomal bupivacaine group (aOR � 0.43 (95% CI: 0.16-0.99), p � 0.047).

Discussion
Compared to nonliposomal bupivacaine single shot ISBs, there was no difference in the number of patients with block-or surgical-related complications. Of the block-related complications in both groups, all patients with postoperative numbness resolved within 30 days, and the SOB resolved by postoperative day 2. Our study results are comparable to data from several large database studies and meta-analysis on nonliposomal bupivacaine interscalene blocks. e meta-analysis by Moore et al. looked at 6243 interscalene blocks and found a major complication rate of 0.35% and a minor complication rate of 11.32% [20]. Shin et al. performed a large database study on 27,072 shoulder arthroscopies and found a surgical complication rate of 7.9% and block-related complication rate of 0.59% [21]. Koh et al. evaluated 11,450 shoulder arthroplasty patients finding a complication rate of 7.6% [22]. While the rate of complications seen in our study is similar to the large database meta-analyses [20][21][22], all are still lower than described by Vandepitte et al. who compared liposomal bupivacaine added to bupivacaine vs. bupivacaine interscalene blocks [16]. ey found that 17 out of 26 patients in the liposomal bupivacaine group reported adverse events compared to 10 out of 24 in the bupivacaine group. All adverse events occurred within the first 24 hours and reported no further adverse events. ere was 0% in the liposomal bupivacaine group with shortness of breath compared to 16% in the bupivacaine group. While these rates of adverse events are much higher than the number of patients with complications that we saw in our study and in the previous large database studies, this is likely due to the retrospective nature of our study and differences in documentation of adverse events.
Postoperative pain-related emergency room visits are important to surgical teams in patient satisfaction and cost to the hospital. e most common reason for visiting the     visits within 7 days after outpatient rotator cuff repair. Of those 90, 34 (or 2.6%) were related to pain, which is similar to our findings [23]. ere was also a similarity in hospital readmissions between the liposomal bupivacaine and nonliposomal bupivacaine group, with 2.4% and 2.9% readmissions, respectively. Additionally, there were 19 readmissions in the liposomal bupivacaine group, which indicates that liposomal bupivacaine is not masking serious complications that require hospital attention. ese readmission rates are in line with previously reported readmissions rates following shoulder surgery. Both Shin et al. and Kosinski et al. found a 1% 30-day readmission rate for patients undergoing outpatient rotator cuff repairs, whereas Koh et al. and Cvetanovich et al. found a 30-day readmission rate of 2.3% and 2.56%, respectively, in patients undergoing shoulder arthroplasty [21,[23][24][25].
Postoperative phone calls from surgical patients may indicate inadequate pain control, which can ultimately require increased time demand from surgeons. Our study demonstrates that there were significantly fewer patient calls in the liposomal bupivacaine group compared to the nonliposomal bupivacaine group within three days after surgery. e most common reason for phone calls within the liposomal bupivacaine group was for pain medication refills, indicating a continued need for pain control. e number of calls for pain medication refills between the two groups was similar (13 in the liposomal bupivacaine group and 14 in the nonliposomal bupivacaine group). In contrast, the most common reason for phone calls in the nonliposomal bupivacaine group was uncontrolled pain. Phone calls reporting symptoms such as constipation or numbness were scarce and similar in both groups. e similarity in the number of phone calls within 30 days of surgery is expected since the effect of both blocks has worn off.
It is important to note that the baseline characteristics of the two groups were not similar. e nonliposome group had significantly more males than the liposomal bupivacaine group, and the liposomal bupivacaine group had significantly more ASA 3 and significantly less ASA 1 than the nonliposomal bupivacaine group.
ese may impact the results as having more males may increase the risk of nerve injury as well as having increased ASA level [26,27].
Limitations of this retrospective study include the wide variety of medication combinations used, including medication type and volume, in both the liposomal bupivacaine and nonliposomal bupivacaine groups. While the wide variety of procedures in this study was beneficial in indicating medication safety, they could impact statistical analysis as variables. Postoperative complications were followed for 30 days and thus may not provide a complete assessment between rates of complications between the two study groups. Additionally, any out-of-network phone calls, ER visits, or admissions were not captured by this study.

Conclusions
e use of liposomal bupivacaine for interscalene blocks demonstrated no significant difference in the number of patients with surgical or block complications when compared to nonliposomal bupivacaine. While there was a significant difference in the number of patients that made a phone call about their pain by postoperative day 3, there was no difference in any of the other secondary outcomes. us, further prospective studies are warranted to demonstrate the safety and potential efficacy of liposomal bupivacaine in ISB for shoulder operations.    Data Availability e data sets generated and analysed during the current study are not publicly available due to the IRB but are available from the corresponding author on reasonable request.

Ethical Approval
is was a University of Minnesota IRB exempt study (IRB ID STUDY00003305).

Consent
All patients enrolled gave their consent allowing their data to be used for research.

Disclosure
Interim data from this manuscript was presented at the 2018 European Society of Regional Anesthesia meeting in Dublin, Ireland.

Conflicts of Interest
J. Hutchins is a speaker, consultant, and has received research funds from Pacira Pharmaceuticals; he is a consultant and owns stock with Insitu Biologics, is a consultant, speaker, and owns stock for Acel RX, a speaker for Sonosite, a consultant for Worrell, a consultant for Johnson and Johnson, and a consultant, speaker, and has received research funds from Avanos. AB is a consultant for Avanos and Pacira Pharmaceuticals. e other authors declare no conflicts of interest.