Breast cancer surgery is associated with severe postoperative pain in nearly 40% of women undergoing this surgical procedure.1 Capitalizing on modern portable high-resolution ultrasound (US) guidance, pectoralis I2 and II3 and serratus4 fascial plane blocks have recently been described for analgesia after breast surgery, wherein the medial pectoral, lateral pectoral, and/or intercostal nerves are reportedly blocked by a large volume of local anesthetic injectate because these nerves traverse in between readily identifiable muscular fascial planes. However, evidence of the analgesic efficacy of these fascial plane blocks remains limited to a single randomized controlled trial that was neither double-blinded nor placebo-controlled and did not compare these blocks with each other.5 The promise of an easy-to-perform, single-injection, reliable, and safe regional anesthesia technique2–4 has spawned fervent interest and rapid uptake in clinical practice;5–9 yet, definitive evidence supporting the analgesic efficacy of these fascial plane blocks is lacking, and how these block techniques compare with each other is unknown.
In the absence of double-blind placebo-controlled randomized controlled trials, thoughtfully designed cohort studies with rigorous control for bias are considered reasonable substitutes that can offer comparable results to guide clinical practice.10 The US-guided pectoralis I2 and II3 and serratus4 blocks were introduced in tandem as analgesic options for breast cancer surgery at Women’s College Hospital (WCH), a university-affiliated free-standing ambulatory teaching hospital in Toronto, Ontario, Canada. As of July 2013, these blocks were offered preoperatively to breast cancer surgery patients as an analgesic adjunct to general anesthesia. The choice of fascial plane block, whether pectoralis I or II or serratus, was based on the preference of the attending anesthesiologist; however, performance of the pectoralis II block was promptly discontinued because of concerns regarding the disruption of the surgical tissue within the axilla.
Since their introduction into practice in July 2013, it has been our anecdotal clinical experience that both pectoralis I and serratus blocks are similarly effective in treating acute pain after breast tumor excision. Therefore, this cohort study aims to evaluate the early analgesic benefits of adding either a pectoralis I or serratus block to the care standard, as captured within hospital records, in women who underwent outpatient breast cancer surgery at WCH between July 2013 and May 2015, compared with women who did not receive these blocks during the same time period. For each of the pectoralis I and serratus blocks, compared with standard care, we tested the primary joint hypothesis that the block enhances postoperative analgesia by improving 2 important analgesic outcomes, namely both postoperative analgesic consumption and the risk of opioid-related side effects. To do so, we sought to sequentially demonstrate that the pectoralis I or serratus blocks: (1) reduce postoperative in-hospital (predischarge) opioid consumption; and (2) decrease the incidence of postoperative nausea and vomiting (PONV) compared with control. In addition, it has been our own anecdotal experience that pectoralis I block can provide similar analgesia to serratus blocks. Such an observation contests the original anatomical description of pectoralis I block2 as a block with clinical utility confined to the insertion of breast expanders and subpectoral prostheses. Therefore, as a secondary hypothesis, we also assessed whether pectoralis I was noninferior to the serratus block on the same 2 outcomes in an analogous joint hypothesis testing approach.
The authors adhered to the Strengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines for reporting observational studies11 in preparing the analysis plan and writing this article.
This propensity-matched retrospective cohort study was approved by the research ethics board at WCH, University of Toronto. The primary sources of data included in the study were the medical records of women who underwent ambulatory breast cancer surgery under general anesthesia between July 2013 and May 2015 at WCH. The requirement for written informed consent was waived because this retrospective review was limited to pre-existing data that had been collected prospectively as part of the care standard and documented by care providers (surgeons, anesthesiologists, and nurses). The primary sources of data included individual patient anesthesia charts, records from phases I and II of recovery, block room logbook, and other hospital records. The anesthesia chart and the block room logbook captured the complete details of all blocks performed. The data sets for all patients undergoing all types of breast cancer procedures during the specified period were retrieved, and relevant data were extracted and anonymized.
All female patients who had breast tumor resection under general anesthesia alone or with the addition of a pectoralis I or serratus fascial plane block performed preoperatively between July 2013 and May 2015 were considered. Women 18 to 75 years of age with a body mass index <40 kg/m2 and an American Society of Anesthesiologists (ASA) classification I or II undergoing unilateral or bilateral breast tumor excision were eligible. Both benign and neoplastic tumors were considered. Patients with a significant psychiatric history (major depression or generalized anxiety disorder), preplanned overnight hospital stay, pre-existing chronic pain (lasting for at least 3 months),12 or who were opioid dependent with a mean daily use of more than 30 mg of oxycodone or equivalent per day (as reported by patients) were excluded. Eligible surgical procedures included partial mastectomy with sentinel lymph node biopsy or axillary lymph node dissection, mastectomy alone, or with sentinel lymph node biopsy or axillary lymph node dissection performed by 1 of 4 breast oncology surgeons. Reconstructive breast procedures, whether simultaneous or subsequent, were excluded, and only the first surgical procedure was considered in cases of more than 1 occurrence during the period examined. Finally, patients who did not receive preoperative oral analgesics (see Care Standard) and/or those who received additional local anesthetic infiltration by the surgeons were excluded.
In-hospital analgesic management of patients undergoing breast cancer surgery at WCH is governed by a clinical pathway that includes 1 g of preoperative oral acetaminophen and 400 mg of celecoxib, as well as a preprinted order sheet for postoperative analgesia. Although intraoperative practices may vary slightly among anesthesiologists within the same institution, all patients undergoing breast cancer surgery at WCH routinely receive general anesthesia using sevoflurane or desflurane in an air-oxygen mixture with either an endotracheal tube or a laryngeal mask airway. Intraoperative analgesia is provided using fentanyl administered intravenously (IV) based on hemodynamic responses to surgical stimuli. For PONV prophylaxis, 8 mg of IV dexamethasone and 4 mg of ondansetron are routinely administered intraoperatively to all patients. Assessment of postoperative pain severity and administration of postoperative analgesics is performed by recovery room nurses. All patients are routinely discharged home on the same day from phase II recovery after nurses in this unit have decided that the institutional discharge criteria are met.13
In this cohort study, 3 analgesic modalities were considered. The choice of analgesic modality reflects one or more of patient-related, anesthesiologist-related, and surgeon-related preferences. The first modality reflects the institutional care standard, whereby patients requesting analgesia or reporting a visual analog scale score ≥4 postoperatively receive 25 to 50 µg of IV fentanyl every 5 minutes as needed up to 100 µg, followed by 2 to 4 mg of IV morphine every 5 minutes as needed up to 10 mg, or 0.2 to 0.4 mg of IV hydromorphone every 5 minutes as needed up to 2 mg, and/or 5 to 10 mg of oral oxycodone every 2 hours as needed. The second modality includes the same opioid-based analgesic regimen, in addition to a US-guided pectoralis I block performed preoperatively by injecting 15 to 20 mL of ropivacaine 0.33% to 0.5% in between the pectoralis major and minor muscles, as described previously.2 The third modality combines the same opioid-based analgesic regimen with a US-guided serratus block performed preoperatively by injecting 20 to 25 mL of ropivacaine 0.33% to 0.5% superficial or deep to the serratus anterior muscle, as described previously.4 Although the choice of local anesthetic volume and concentration was left to the individual anesthesiologists’ discretion, the total doses were similar for each of these blocks. Patients also received 1 to 3 mg of IV midazolam for anxiolysis before block administration, and all fascial plane blocks were performed by staff anesthesiologists or supervised regional anesthesia fellows.
Patients who received a pectoralis I block were matched to those who received a serratus block and to control patients (no block) on a 1:1:1 ratio using propensity score matching.14 This matching was used to obtain groups of patients corresponding to the 3 analgesic modalities that are balanced with regard to potential confounding baseline variables. Propensity scores were calculated using a nonparsimonious multivariable logistic regression model in which the mode of analgesia was used as a dependent variable. The independent variables included 5 confounders and/or risk factors for acute pain after breast cancer surgery, as reported in relevant literature15,16 or perceived by the authors, including: (1) age; (2) invasiveness (type of surgical procedure); (3) identity of the surgeon; (4) duration of surgical procedure; and (5) preoperative opioid consumption. Exact matching was used for the categorical confounders (type of surgical procedure and identity of the surgeon); and continuous confounders (age, duration of procedure, and preoperative opioid consumption) were modeled using restricted cubic splines. Propensity matching was performed using a 1:1:1 nearest neighbor algorithm; first, pectoralis I block patients were paired to control patients; then, pectoralis I block patients were paired to serratus block patients; finally, only pectoralis I block patients that were successfully matched to both serratus and control patients were selected, producing the matched triplets. We set the boundaries for pairs matching to a caliper size of 0.2 standard deviations of the logit of the propensity score, as suggested by Austin.17 Cases that were missing the primary outcomes or any of the prespecified acute pain risk factor data were excluded. To assess the adequacy of matching, we evaluated the balance on the 5 preidentified confounders after inverse weighting of these variables by the respective propensity scores; standardized differences in the matched sample that exceeded 0.1 were considered indicative of residual imbalance. All subsequent analysis was restricted to this set of propensity-matched patients.
The set of matched patients was compared for 2 primary outcomes: (1) cumulative postoperative in-hospital (predischarge) opioid consumption, as measured by oral morphine equivalents (mg)18; and (2) the proportion of patients experiencing PONV (yes/no) during their in-hospital (predischarge) stay. Both analgesic outcomes were considered mutually confirmatory evidence of the analgesic efficacy of the fascial blocks and were thus examined in conjunction. That is, we considered pectoralis I and serratus blocks to be effective analgesic interventions that improve patient care if they reduce cumulative postoperative opioid consumption and, accordingly, decrease the risk of opioid-related side effects. We therefore sequentially tested the primary joint hypothesis that the addition of a pectoralis I or serratus block to the care standard would reduce the postoperative oral morphine equivalent consumption and, consequently, also reduce the risk of PONV compared with control. Finally, we compared the pectoralis I block with the serratus block for noninferiority of both primary outcomes to test our secondary hypothesis.
Secondary outcomes included intraoperative fentanyl requirements (µg), pain severity at rest (visual analog scale score) at admission (initial) and discharge (final) to/from phase I and II recovery, as well as the lowest and highest scores recorded, time to first analgesic request (minutes), and durations of stay in phase I and phase II recovery (minutes). We also sought documented nerve block–related complications, including bleeding, hematoma formation, paresthesia, local anesthetic systemic toxicity, and pneumothorax. Other data captured included demographic data (age, weight, height, and ASA classification), comorbidities, preoperative medications, history of chronic opioid use, type and laterality of breast surgery, the surgeon performing the procedure, and the duration of surgery.
We used multivariable regression to model the 2 primary outcomes, postoperative in-hospital opioid consumption, and risk of PONV, within the propensity-matched cohort. The 2 primary outcomes were examined sequentially,19 using a serial gatekeeping approach,20 beginning with opioid consumption. The risk of PONV was analyzed for a particular block only after demonstrating that this block reduced opioid consumption. This serial testing approach is clinically sound because reduced PONV corroborates effective analgesia only if opioid consumption and/or pain scores are also reduced. This approach allowed the threshold of statistical significance for superiority testing to be maintained at .017 after a Bonferroni correction for the 2-tailed comparisons of the 3 groups over each outcome (ie, intervention versus intervention, and intervention versus control).
Subsequently, we compared the pectoralis I block with the serratus block for noninferiority of the 2 primary outcomes, sequentially, using a threshold of significance of .05 for the 1-sided test of noninferiority. We considered pectoralis I block noninferior to serratus block if it was noninferior for both outcomes, beginning with opioid consumption, and subsequently the risk of PONV, with noninferiority margins (Δ) of 10 mg of oral morphine equivalent consumption and 17.5% absolute difference in the risk of developing PONV. One-sided noninferiority testing for the primary outcomes was performed by comparing the 90% confidence interval (CI) of the difference between groups and their relative risk for the opioid consumption and PONV outcomes, respectively, with the predetermined Δ for these outcomes. The 90% CI was used because the noninferiority hypothesis of interest is unidirectional. Subsequently, and after establishing noninferiority over an outcome, we proceeded to test for superiority over this outcome as well, as part of the secondary hypothesis testing. Because superiority could be claimed if demonstrated for either of the 2 primary outcomes, we designated .025 as the threshold of statistical significance for superiority testing.
For intergroup superiority testing of the secondary outcomes, we used t tests, Fisher exact tests, χ2 tests, or Mann-Whitney U tests, as appropriate. We also used the Kaplan–Meier analysis with the log-rank test for the time-to-event outcomes. A Bonferroni–Holm correction for repeated measurements was used to set the threshold of significance at .025 for the measurement of oral morphine equivalent consumption in phase I and phase II recovery; and a similar correction was used to set the threshold of significance at .00625 for the 8 measurements of pain severity scores in phase I and phase II recovery. Another Bonferroni-Holm correction with a threshold of significance set at .007 was used to protect against the risk of a type I error posed by measurement of 7 secondary outcomes.
We used the SAS statistical software version 9.3 (SAS Institute, Cary, NC) and R statistical software version 2.15.2 (R Foundation for Statistical Computing, Vienna, Austria) in all of our statistical analyses and used SAS g-match macro for the propensity score matching.
Sample Size Calculation
Power analysis for this retrospective cohort was based on the noninferiority comparison of the risk of PONV between the pectoralis I and serratus groups because this outcome requires a larger sample size than a corresponding comparison using the opioid consumption. Based on a sample of 200 consecutive patients who received a pectoralis I (100 patients) or serratus (100 patients) block in 2013 for breast cancer surgery at WCH, we observed that 35 and 23 patients developed PONV, respectively. We therefore estimated the risk of PONV associated with pectoralis I and serratus blocks to be 35% and 23%, respectively. A 20% difference in the risk of PONV is generally considered as the least that is clinically meaningful. Therefore, for a 1-sided noninferiority test that designates a 17.5% absolute increase in PONV (or a relative risk of 1.76) as a noninferiority margin, type I error estimate of .05, and 80% power, we estimated that 72 patients per group were needed.
A 30-mg difference in oral morphine equivalent consumption (or a 10-mg difference in IV morphine consumption) is generally considered clinically meaningful; we therefore designated a 10-mg difference in oral morphine consumption as a noninferiority margin. The aforementioned sample size (72 patients per group) estimate provides 98% power to perform a 1-sided noninferiority test of the second primary outcome, cumulative in-hospital oral morphine equivalent consumption, considering an allowable difference of 0 mg, a noninferiority margin of 10 mg, a standard deviation of morphine consumption equivalent to 16 mg (or variance of 256 mg), and a type I error estimate of .05.
We screened all charts of breast cancer surgery patients during the time period of interest. Our retrospective review identified 787 potentially relevant records of breast cancer surgery patients who had their procedures performed between July 2013 and May 2015. Of these, 80 records did not meet the inclusion criteria (Figure 1), and another 42 records were excluded because of preplanned hospital admission, incomplete data, second same-site breast cancer surgery (reconstructive or re-excision), or surgery without inhalational general anesthesia. Of the 665 records that remained, 225 were successfully matched on a 1:1:1 basis based on predetermined confounders and baseline characteristics, including 75 patients for each of pectoralis I block, serratus block, and control groups. This number of patients was considered sufficient because it provided 81% power for the intended test of noninferiority. The sample also included 77% and 70% of eligible pectoralis I block and serratus block patients, respectively. Table 1 summarizes patient baseline characteristics after propensity score matching. Patients in the matched population were similar with respect to both matched and other baseline characteristics.
Both pectoralis I and serratus blocks were each associated with reduced postoperative in-hospital opioid requirements, with a mean (95% CI) cumulative oral morphine equivalent consumption of 20.1 mg (15.6, 24.6) and 24.3 mg (20.0, 28.7), respectively, compared with 43.2 mg (38.0, 48.4) in the control group (P < .0001); however, the difference between the pectoralis and the serratus groups (−4.2 mg [−10.5, 2.1]) was not statistically significant (Table 2). Both pectoralis I and serratus blocks were each associated with reduced in-hospital risk of PONV, with an incidence of 32.0% and 33.3%, respectively, compared with 88.7% in the control group (P = .00011); however, the difference between the pectoralis I and the serratus groups (1.02 [0.82, 1.27]) was not statistically significant.
The difference (90% CI) in cumulative oral morphine equivalent consumption during in-hospital stays between the pectoralis and serratus groups (pectoralis − serratus) was −4.2 mg (−10.4, 2.0). The upper CI of the difference was significantly smaller (P = .001) than the predetermined noninferiority margin (Δ = 10 mg; Figure 2). Consequently, the pectoralis I block was found to be noninferior but not superior to the serratus block in cumulative postoperative in-hospital opioid consumption after ambulatory breast cancer surgery. The relative risk (90% CI) of PONV of the pectoralis and serratus groups was 1.0 (.7, 1.4). The upper CI of the risk was significantly smaller (P = .002) than the predetermined noninferiority margin (Δ = 17.5%, or a relative risk of 1.76; Figure 3). Consequently, the pectoralis I block was found to be noninferior but not superior to the serratus block in preventing PONV in hospital after ambulatory breast cancer surgery.
Pectoralis I and serratus blocks were each associated with reduced intraoperative opioid requirements, with an intraoperative fentanyl consumption of 131.0 µg (111.6, 150.4) and 136.1 µg (118.7, 153.5), respectively, compared with 179.2 µg (157.7, 200.7) in the control group (P = .001); however, the difference between the pectoralis and serratus groups (−5.1 µg [−30.8, 20.6]) was not statistically significant (Table 3). The differences in pain severity scores between the 3 study groups were not significant at any of the time points, except for the worst pain scores recorded in phase II recovery, in which both pectoralis and serratus blocks appeared to provide better pain relief compared with control (P = .004); however, the difference between the pectoralis and serratus groups (−0.2 cm [−0.8, 0.4]) was not statistically significant (Table 3). Pectoralis and serratus blocks were also each associated with prolonged time to first postoperative analgesic request by 23.3 and 20.6 minutes, respectively, compared with the control group (P < .0001); however, the difference between the pectoralis and serratus groups (2.7 minutes [−2.2, 7.6]) was not statistically significant (Table 3). The proportion of patients requiring postoperative opioid analgesia during their in-hospital recovery was lower in both pectoralis and serratus groups, with 44.0% and 49.3% receiving opioids, respectively, compared with 74.5% in the control group (P = .0002); however, the difference between the pectoralis and serratus groups (0.8 [0.4, 1.5]) was not statistically significant. Finally, pectoralis I and serratus blocks each were associated with expedited discharge from phase I recovery by 18.4 and 17.8 minutes, respectively, compared with control (P < .0001); however, the difference between the pectoralis and serratus groups (−0.6 minutes [−4.1, 2.9]) was not statistically significant. Similarly, pectoralis I and serratus blocks were each associated with expedited discharge from phase II recovery by 23.3 and 30.9 minutes, respectively, compared with control (P < .0001); however, the serratus I block seemed to expedite discharge by 7.6 minutes (1.8, 13.4) (Table 3). No block-related complications were reported in any of the patients included in this review. None of the patients included in this cohort required unplanned extended hospital stays or admissions.
This cohort study addresses the need for evidence supporting the use of fascial plane blocks that have already found their way into clinical practice. It is the first to support the early analgesic benefits of each of the pectoralis I and the serratus fascial plane block for breast tumor excision. Our results suggest that both block techniques are similarly effective in reducing postoperative in-hospital opioid consumption and a classic opioid-related side effect, PONV, after ambulatory breast cancer surgery. Compared with control, each of these blocks is also associated with decreased intraoperative fentanyl consumption, prolonged time-to-first analgesic request, decreased likelihood for requiring postoperative opioid analgesics in hospital, and shortened duration of stay in phase I and phase II recovery. As such, each of these blocks can be considered as useful adjuncts in patients having ambulatory breast cancer surgery under general anesthesia.
We found the pectoralis I block to be noninferior to serratus block. This is a curious finding because the pectoralis I block is presumed to exert its analgesic effect primarily by blocking the medial and lateral pectoral nerves.2 In fact, the initial description of the pectoralis I block focused exclusively on pain relief for the insertion of breast expanders and subpectoral prostheses.2,3 However, our findings suggest that the applicability of this block may be far wider. Given the limited understanding of the mechanisms of action of these fascial plane blocks, several theories may explain the early analgesic efficacy observed herein. One possibility is that the reduction of postsurgical muscular spasm in the pectoralis major and minor muscles significantly improves postoperative pain relief.21,22 Another possibility is that the medial and lateral pectoral nerves may contain a more important sensory component than what traditional anatomical descriptions suggest.23 The latter may also help to explain some of the pain relief reported in the surgical literature with local anesthetics injected under direct vision by surgeons intraoperatively.24–26 These nerves may also communicate with and exert an analgesic effect through the anterior cutaneous branches of the intercostal nerves.27 In addition, our pectoralis I block technique may have resulted in lateral spread to the terminal branches of the intercostal nerves;28 this is particularly plausible considering that our anesthesiologists used larger local anesthetic volumes (15–20 mL) compared with the 10 mL volume described by Blanco et al.3 Finally, systemic absorption of local anesthetics, as observed with tumescent anesthesia,29,30 may offer another explanation as to why the pectoralis I block was found to be as effective as the serratus block.
There are several limitations to this retrospective study. Being an observational study, our work supports conclusions about potential associations and not causal relationships. Our conclusions may have also been limited by potential errors in data collection and documentation. Furthermore, although using propensity score matching may have reduced the risk of bias and improved the validity of our analysis, we could only account for identifiable confounders when relevant data were available. Although we could not account for anesthesiologists’ preferences and any potential tendency to assign more invasive surgical procedures to one block and not the other, matching by surgical procedure helped to mitigate this form of selection bias. It was not possible to examine certain acute pain risk factors, such as genetics, ethnicity, anxiety, and pain catastrophisation15 because of the lack of data. Moreover, our findings are confined to the specific settings in which the data were collected and are not necessarily generalizable to other patient subpopulations, breast surgeries, or fascial plane block techniques. Notably, the population examined was limited to ASA I and II patients in an ambulatory center, which may not be representative of the general population. Furthermore, the lack of difference in pain scores documented at admission and discharge from phase I and phase II recovery may reflect institutional policies that inform discharge readiness from these units. In addition, the matched cohort excluded numerous patients who had otherwise met the study inclusion criteria; this may limit the generalizability of our findings. Our outcomes were also restricted to early postoperative experience; data regarding the analgesic effects beyond hospital discharge were not captured. Likewise, our records did not capture other important outcomes, such as the incidence of undesirable partial or complete brachial plexus block following pectoralis I block, quality of postoperative recovery, patient satisfaction, and chronic pain. Finally, although our data set represents a modest-sized cohort, the CIs were relatively narrow, suggesting accuracy of treatment effect estimation.
In conclusion, our results suggest that the addition of a pectoralis I or serratus block is associated with early analgesic benefits, including a reduction in postoperative in-hospital opioid consumption and PONV, following ambulatory breast cancer surgery. This association supports the use of these blocks as adjuncts to general anesthesia in this patient population and provides important information for the development of robust randomized trials in the future.
Name: Faraj W. Abdallah, MD.
Contribution: This author helped design and conduct the study, collect and analyze the data, and prepare the final manuscript. This author is the archival author.
Name: David MacLean, MD.
Contribution: This author helped conduct the study, collect the data, and prepare the final manuscript.
Name: Caveh Madjdpour, MD.
Contribution: This author helped design the study and prepare the final manuscript.
Name: Tulin Cil, MD, Med, FRCSC.
Contribution: This author helped analyze the data and prepare the final manuscript.
Name: Anuj Bhatia, MBBS, MD, FRCA, FRCPC, FIPP, FFPMRCA, EDRA.
Contribution: This author helped analyze the data and prepare the final manuscript.
Name: Richard Brull, MD, FRCPC.
Contribution: This author helped design and conduct the study and prepare the final manuscript.
This manuscript was handled by: Honorio T. Benzon, MD.
1. Poleshuck EL, Katz J, Andrus CH, et al. Risk factors for chronic pain following breast cancer surgery: a prospective study. J Pain. 2006;7:626–634.
2. Blanco R. The ‘pecs block’: a novel technique for providing analgesia after breast surgery. Anaesthesia. 2011;66:847–848.
3. Blanco R, Fajardo M, Parras Maldonado T. Ultrasound description of Pecs II (modified Pecs I): a novel approach to breast surgery. Rev Esp Anestesiol Reanim. 2012;59:470–475.
4. Blanco R, Parras T, McDonnell JG, Prats-Galino A. Serratus plane block: a novel ultrasound-guided thoracic wall nerve block. Anaesthesia. 2013;68:1107–1113.
5. Bashandy GM, Abbas DN. Pectoral nerves I and II blocks in multimodal analgesia for breast cancer surgery: a randomized clinical trial. Reg Anesth Pain Med. 2015;40:68–74.
6. Desroches J, Grabs U, Grabs D. Selective ultrasound guided pectoral nerve targeting in breast augmentation: how to spare the brachial plexus cords? Clin Anat. 2013;26:49–55.
7. ELdeen HM. Ultrasound guided pectoral nerve blockade versus thoracic spinal blockade for conservative breast surgery in cancer breast: a randomized controlled trial. Eg J Anaesth. 2016;32:29–35.
8. Pérez MF, Miguel JG, de la Torre PA. A new approach to pectoralis block. Anaesthesia. 2013;68:430.
9. Wahba SS, Kamal SM. Thoracic paravertebral block versus pectoral nerve block for analgesia after breast surgery. Eg J Anaesth. 2014;30:129–135.
10. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med. 2000;342:1878–1886.
11. von Elm E, Altman DG, Egger M, et al; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Prev Med. 2007;45:247–251.
12. Merskey H, Bogduk N; International Association for the Study of Pain. Task Force on Taxonomy. Classification of Chronic Pain: Descriptions of Chronic Pain Syndromes and Definitions of Pain Terms. 1994.Seattle, WA: IASP Press.
13. Aldrete JA. Modifications to the postanesthesia score for use in ambulatory surgery. J Perianesth Nurs. 1998;13:148–155.
14. Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70:41–55.
15. Katz J, Poleshuck EL, Andrus CH, et al. Risk factors for acute pain and its persistence following breast cancer surgery. Pain. 2005;119:16–25.
16. Andersen KG, Kehlet H. Persistent pain after breast cancer treatment: a critical review of risk factors and strategies for prevention. J Pain. 2011;12:725–746.
17. Austin PC. Some methods of propensity-score matching had superior performance to others: results of an empirical investigation and Monte Carlo simulations. Biom J. 2009;51:171–184.
18. Canadian Pharmacists Association. Compendium of Pharmaceuticals and Specialties: The Canadian Drug Reference for Health Professionals. 2010.45th ed. Ottawa, ON, Canada: Canadian Pharmacists Assoc.
19. Dmitrienko A, Tamhane AC. Gatekeeping procedures with clinical trial applications. Pharm Stat. 2007;6:171–180.
20. Mascha EJ, Turan A. Joint hypothesis testing and gatekeeping procedures for studies with multiple endpoints. Anesth Analg. 2012;114:1304–1317.
21. Hoffman GW, Elliott LF. The anatomy of the pectoral nerves and its significance to the general and plastic surgeon. Ann Surg. 1987;205:504–507.
22. Goshievich A, Kirkham KR, Brull R, Brown MH. Novel approach to intractable pectoralis major muscles spasm following sub-muscular expander-implant breast reconstruction. Plast Surg Case Studies. 2015;1:1–3.
23. Brermner-Smith AT, Umvin AJ, Williams WW. Sensory pathways in the spinal accessory nerve. J Bone Joint Surg. 1999;81:226–228.
24. Parker WL, Charbonneau R. Large area local anesthesia (LALA) in submuscular breast augmentation. Aesthet Surg J. 2004;24:436–441.
25. Bell M. Office anesthesia for breast augmentation made easy. Can J Plast Surg. 2007;15:178.
26. McCarthy CM, Pusic AL, Hidalgo DA. Efficacy of pocket irrigation with bupivacaine and ketorolac in breast augmentation: a randomized controlled trial. Ann Plast Surg. 2009;62:15–17.
27. Akita K, Kawashima T, Shimokawa T, Sato K, Sato T. Cutaneous nerve to the subacromial region originating from the lateral pectoral nerve. Ann Anat. 2002;184:15–19.
28. Sleth JC. [Pecs block in breast surgery: in fact a simple intercostal block?]. Ann Fr Anesth Reanim. 2014;33:548.
29. Klein JA. Tumescent technique for local anesthesia improves safety in large-volume liposuction. Plast Reconstr Surg. 1993;92:1085–1098.
© 2017 International Anesthesia Research Society
30. Jabs D, Richards BG, Richards FD. Quantitative effects of tumescent infiltration and bupivicaine injection in decreasing postoperative pain in submuscular breast augmentation. Aesthet Surg J. 2008;28:528–533.