Secondary Logo

Journal Logo

Paravertebral Block Does Not Reduce Cancer Recurrence, but Is Related to Higher Overall Survival in Lung Cancer Surgery: A Retrospective Cohort Study

Lee, Eun Kyung MD*; Ahn, Hyun Joo MD, PhD*; Zo, Jae Ill MD, PhD; Kim, Kyunga PhD; Jung, Dae Myung MD*; Park, Joo Hyun MD*

doi: 10.1213/ANE.0000000000002342
Cancer and Supportive Care: Original Clinical Research Report

BACKGROUND: Postoperative analgesic methods are suggested to have an impact on long-term prognosis after cancer surgery through opioid-induced immune suppression. We hypothesized that regional analgesia that reduces the systemic opioid requirement would be related to lower cancer recurrence and higher overall survival compared to intravenous patient-controlled analgesia (PCA) for lung cancer surgery.

METHODS: Records for all patients who underwent open thoracotomy for curative resection of primary lung cancer between 2009 and 2013 in a tertiary care hospital were retrospectively analyzed. Patients were divided by postoperative analgesic methods: PCA (n = 574), thoracic epidural analgesia (TEA, n = 619), or paravertebral block (PVB, n = 536). Overall and recurrence-free survivals were compared among 3 analgesic methods via a multivariable Cox proportional hazard model and a log-rank test after adjusting confounding factors using propensity score matching (PSM).

RESULTS: Analgesic method was associated with overall survival (P= .0015; hazard ratio against TEA [95% confidence intervals]: 0.58 [0.39–0.87] for PCA, 0.60 [0.45–0.79] for PVB). After confounder adjustment using PSM, PVB showed higher overall survival than PCA (log-rank P= .0229) and TEA (log-rank P= .0063) while PCA and TEA showed no difference (log-rank P= .6). Hazard ratio for PVB was 0.66 [0.46–0.94] against PCA and 0.65 [0.48–0.89] against TEA after PSM. However, there was no significant association between the analgesic methods and recurrence-free survival (P= .5; log-rank P with PSM = .5 between PCA and TEA, .5 between PCA and PVB, .1 between TEA and PVB).

CONCLUSIONS: Pain-control methods are not related to cancer recurrence. However, PVB may have a beneficial effect on overall survival of patients with lung cancer.

From the *Department of Anesthesiology and Pain Medicine and Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea; and Statistics and Data Center, Research Institute for Future Medicine, Samsung Medical Center, Seoul, Korea.

Accepted for publication June 5, 2017.

The authors declare no conflicts of interest.

Funding: None.

Reprints will not be available from the authors.

Address correspondence to Hyun Joo Ahn, MD, PhD, Department of Anesthesiology and Pain Medicine, Samsung Medical Centre, 50 Ilwon-dong, Kangnam-gu, Seoul 135–710, South Korea. Address e-mail to

Postoperative analgesic methods have been suggested to influence cancer recurrence and long-term prognosis after cancer surgery. Opioids trigger immune suppression through substantial impairment of innate immunity, altered antigen presentation, and a predominant balance in favor of protumor cytokines. As a result, they can impair immune surveillance against cancer.1–6 If a patient’s immune surveillance is suppressed during the immediate postoperative period, remnant malignant cells or micrometastases may grow, establish, and spread early after surgery.4,7,8 Regional analgesia can reduce the need for perioperative opioids. Early studies on breast and prostate cancer reported that using regional analgesia had beneficial outcomes in cancer recurrence and survival time.9,10 However, subsequent studies showed controversial results,11–13 and few studies included paravertebral block (PVB), which uses local anesthetics only.

Therefore, in this large retrospective cohort study, we evaluated the association of regional analgesia, especially of PVB, on the prognosis of lung cancer surgery. We hypothesized that thoracic epidural analgesia (TEA) and PVB would be related to lower cancer recurrence and higher overall survival compared to intravenous patient-controlled analgesia (PCA) for lung cancer surgery. We chose to study lung cancer because large doses of opioids are administered to alleviate severe thoracotomy pain and since alternatives such as TEA and PVB exist for pain control.

Back to Top | Article Outline


Patient Records

Our institutional review board approved this retrospective cohort study (institutional review board no. 2016-02-025) and the need for informed consent was waived because of the retrospective design. Electronic records for all patients who underwent curative resection by open thoracotomy for biopsy-proven primary non–small-cell lung cancer between 2009 and 2013 in a tertiary care university hospital were obtained (n = 1741). The study was limited to patients who underwent surgery no later than December 2013, so that at least 2 to 6 years of follow-up data were available for each patient. Information collected from patient records was postoperative analgesic methods, age, sex, body mass index, comorbidities, history of smoking and alcohol use, American Society of Anesthesiologists (ASA) physical status, cancer stage (tumor-lymph nodes-metastasis classification system), duration and type of surgery, adjuvant treatments (chemotherapy and/or radiotherapy [CCRT]), and perioperative blood transfusions. Data on the number of hypotensive events and its treatment, and opioid consumption until postoperative 5 days (POD5) were collected. Total dose of opioid consumption was calculated by opioid conversion ratio and converted to the dose of morphine.14 In-hospital complications, duration of intensive care unit (ICU) and hospital stay, time to cancer recurrence, and date of and reasons for death were collected from the tumor registry of the Department of Thoracic Surgery.

Recurrence-free survival was from date of surgery to date of recurrence, or to the last date the patient was seen without recurrence according to patient medical records. Overall survival was from date of surgery to the date of death or to the last date that the patient was seen alive according to the medical records.

Back to Top | Article Outline

Anesthesia and Analgesia Techniques

Patients were assigned to 1 of 3 groups according to analgesia received postoperatively: PCA (n = 574), TEA (n = 619), or PVB (n = 536). The analgesic methods were decided by each surgeon’s preference and existence of contraindications for regional analgesia. The opioid solution used for PCA was fentanyl (15 μg/mL). The mean duration of PCA was 3 days, and administration was at a basal rate of 1 mL/h, with 1 mL bolus and a 15-minute lockout interval. Patients in the TEA group had a thoracic epidural catheter placed between the T4 and T6 interspaces before surgery. Placement of the catheter was confirmed by fluoroscopy or a test dose. The epidural solution was a mixture of ropivacaine (0.15%) plus hydromorphone (8 μg/mL) and infused at a basal rate of 5 mL/h with 3 mL of bolus and a 15-minute lockout interval. PVB was done by surgeons in the operating field under direct visualization. A multihole catheter was threaded 10 cm into the paravertebral space near incisions. For PVB, 0.5% ropivacaine was used with continuous infusion at 5 mL/h (On-Q PainBuster; B Braun, Irvine, CA). In all 3 groups, analgesics were usually started during thoracotomy closure. Epidural and paravertebral catheters were left in place for an average of 3 days.

During surgery, most patients received balanced general anesthesia, which was a combination of a volatile anesthetic agent (sevoflurane or desflurane) with an end-tidal concentration of <1 minimum alveolar concentration, nondepolarizing muscle relaxants (rocuronium, vecuronium, or cisatracurium), and intravenous opioids with repetitive injections of fentanyl or continuous infusion of remifentanil. All patients were transferred to the ICU after operation.

Back to Top | Article Outline

Statistical Analysis

Patient demographic and clinical data were summarized as frequency (percentage) for categorical variables and mean (standard deviation) or median (interquartile range) for continuous variables. Analysis of variance or the Kruskal-Wallis test was used as appropriate to compare distributions of continuous variables among 3 analgesic groups. Similarly χ2 or Fisher exact test was used as appropriate for comparison of categorical variables among 3 analgesic groups. The recurrence-free and overall survival rates were estimated by Kaplan-Meier method, and multivariable Cox proportional hazards models were applied to assess the effect of the analgesic method on recurrence-free and overall survivals with adjustment of possible confounders. Variables with P< .2 in univariable analyses were entered into a multivariable analysis.

Additionally, we conducted 1-to-1 propensity score matching (PSM) to reduce possible bias due to confounders in comparison of recurrence-free and overall survivals among the 3 analgesic groups. Propensity scores were estimated via multivariable logistic regression with matching variables including age, sex, smoking, ASA classification, hypertension, adjuvant therapy, transfusion, duration and type of operation, and follow-up time. Based on the standard deviation of the logit of the estimated propensity scores, 1-to-1 matching was performed using the nearest-neighbor method with a caliper width of 0.2 in pairwise manner. The matched data included n = 384 each for comparison between TEA and PCA groups; n = 476 each for comparison between TEA and PVB groups; n = 362 each for comparison between PCA and PVB groups. Hazard ratios with 95% confidence intervals (CIs) were estimated by Cox proportional hazards models with the matched data.

The incidence of composite complications among groups was compared by multivariable logistic regression.

For sample size justification, the authors calculated the maximum hazard ratio for the PVB group against the other analgesic methods, which can be detected with 90% power and with the given sample size of 1729 when multivariable Cox model is used. Two-sided test for H0: “no effect of analgesic method on overall survival” was considered at the significance level of 0.05. The maximum hazard ratio for PVB group against the other analgesic method was 0.6 with the observed standard deviation of the group variable and by assuming R2 of the analgesic group variable and other covariates is 0.3. (nQuery+nTerim version 4.0 was used for calculation.)

For all analyses, 2-sided P values <5% were considered statistically significant. P values were corrected by Bonferroni’s method in cases of multiple tests (denominator = 3 if not specified). Data were analyzed using SAS 9.4 software (SAS Institute, Inc, Cary, NC) and R package “Matching” (Jasjeet S. Sekhon, UC Berkeley) for PSM.

Back to Top | Article Outline


Patient Characteristics

A total of 1741 patients received open thoracotomy for biopsy-proven primary non–small-cell lung cancer between 2009 and 2013. Patients who were younger than 18 years (n = 1) or with incomplete data in their records (n = 11) were excluded. The final analysis included 1729 patients. Median follow-up time was 774 days (range, 2–2254 days). Patients were classified according to analgesic method: PCA, TEA, or PVB. Final numbers were 574 patients (33%) for PCA, 619 (36%) for TEA, and 536 (31%) for PVB.

Demographic and clinical characteristics by group are in Table 1. The PCA group had significantly fewer men, smokers, and people with CCRT, and less extensive and shorter duration surgery than the other groups. Incidence of CCRT, duration, and extent of operation were not different between the TEA and PVB groups.

Table 1.

Table 1.

Opioid consumption until POD5 was higher in the order of the PCA, TEA, and PVB groups (morphine equivalent dose 255 ± 73, 174 ± 122, 132 ± 70 mg, respectively, P= .0000, post hoc comparisons between 2 groups were all different after Bonferroni correction, Table 2). All patients were prescribed with the same analgesic protocol for the next 2 to 3 weeks which comprises oral analgesics and/or fentanyl patch (data are not shown).

Table 2.

Table 2.

Number of patients who experienced hypotensive events (systolic blood pressure <90 mm Hg) was higher in the TEA group compared to the other groups with no difference between the PCA and the PVB groups (POD0: 199 [32%], 104 [18%], and 103 [19%], P< .0001; POD1: 231 [37%], 123 [21%], and 104 [19%], P< .0001; POD2: 170 [27%], 130 [23%], and 107 [20%], P= .0038 in the TEA, PCA, and PVB groups, respectively). Number of patients who received vasopressor/inotrope was higher in the TEA group compared to the other groups with no difference between the PCA and PVB groups (POD0: 257 [42%], 57 [10%], and 36 [7%], P< .0001; POD1: 29 [5%], 12 [2%], and 4 [1%], P< .0001; POD2: 14 [2%], 9 [2%], and 4 [1%], P= .2 in the TEA, PCA, and PVB groups, respectively, Table 2).

Back to Top | Article Outline

Recurrence-Free Survival

Recurrence rates during the observation period by group were 32% for PCA, 34% for TEA, and 30% for PVB. Cancer recurrence was the cause of death in 84% of patients in the PCA group, 88% in TEA, and 87% in PVB. Kaplan-Meier survival curves (univariable analysis) showed no difference in recurrence-free survival among 3 analgesic groups (log-rank test P= .7; P= 1 between TEA and PVB, P= 1 between PCA and PVB, P= 1 between PCA and TEA, Bonferroni correction, Figure A).



Multivariable Cox proportional hazards analysis showed that the analgesic method was not related to the recurrence-free survival (P= .5; hazard ratio against the TEA group [95% CI]: 1.01 [0.82–1.24] for the PCA group, 0.90 [0.73–1.10] for the PVB group). Related variables were cancer stage and extent of operation (Table 3).

Table 3.

Table 3.

Additionally, we performed PSM analysis to adjust for confounding factors (Table 5). There was no difference in the recurrence-free survival among 3 groups after PSM (log-rank P value with PSM = .5 between PCA and TEA; .5 between PCA and PVB; .1 between TEA and PVB; PVB group hazard ratio [95% CI]: 0.92 [0.76–1.19] against PCA group, 0.84 [0.67–1.05] against TEA).

Back to Top | Article Outline

Overall Survival

Overall survival rates were 80% in PCA, 75% in TEA, and 86% in PVB. Kaplan-Meier survival curves (univariable analysis) showed higher overall survival in the PVB than the TEA group (log-rank test P= .0016; P= .001 between TEA and PVB, P= .3 between PCA and PVB, P= .2 between PCA and TEA, Bonferroni correction, Figure B).

Multivariable Cox proportional hazards modeling demonstrated that analgesic methods were associated with overall survival. The PVB and PCA groups showed higher overall survival than the TEA group (P= .0015; hazard ratio against the TEA group [95% CI]: 0.58 [0.39–0.87] for the PCA group, 0.60 [0.45–0.79] for the PVB group). However, the initial higher overall survival in the PCA group was not sustained, and decreased over time (PCA × time in Table 4). Other variables related to overall survival were age, male sex, cancer stage, transfusion, and duration and extent of operation (Table 4).

Table 4.

Table 4.

Table 5.

Table 5.

After confounder adjustment using PSM, the PVB group showed higher overall survival than the PCA (log-rank P value = .0229) and TEA groups (log-rank P value = .0063), while the PCA and TEA groups showed no significant difference in overall survival (log-rank P value = .6). Hazard ratio for PVB group was 0.66 [0.46–0.94] against the PCA group and 0.65 [0.48–0.89] against the TEA group (Table 5).

Back to Top | Article Outline

Postoperative Complications

Composite complications were the combined incidence of acute lung injury, pneumonia, delirium, and arrhythmia during the hospital stay. The incidences of composite complications by group were 42% for PCA, 56% for TEA, and 49% for PVB. Analgesic methods were related to composite complications by multivariable logistic regression analysis (odds ratio against TEA group [95% CI]: PCA 0.53 [0.41–0.69], PVB 0.77 [0.60–0.99]; P< .0001). Other related variables were age (odds ratio [95% CI]: 1.03 [1.02–1.05], per year; P< .0001), male sex (odds ratio [95% CI]: 2.14 [1.60–2.86]; P< .0001), ASA physical status (odds ratio [95% CI]: 1.30 [1.04–1.62]; P= .0004), transfusion (odds ratio [95% CI]: 1.94 [1.51–2.49]; P< .0001), duration (odds ratio [95% CI]: 1.18 [1.07–1.20]; P= .0004), and extent of operation (odds ratio [95% CI]: sleeve lobectomy 1.46 [1.03–2.05], P= .0326, pneumonectomy 1.54 [1.11–2.14], P= .0096).

Length of ICU stay (median [interquartile]: 1 [1–2], 1 [1–2], 1 [1–2]; P= .3) and hospital stay (median [interquartile]: 9 [7–12], 9 [7–12], 8 [7–12]; P= .2) did not vary among the PCA, TEA, and PVB groups.

The incidence of in-hospital death was lower in the PCA group than the TEA group (1.0% for PCA, 3.4% for TEA, and 2.6% for PVB; P= .023 between TEA and PCA, P= .2 between PVB and PCA, P= 1.0 between TEA and PVB, Bonferroni correction).

Back to Top | Article Outline


In this study, analgesic methods were not related to recurrence of lung cancer. This finding was also true for PVB, with a regimen that included local anesthetic only.

Previously, regional analgesia was related to reduced risk of cancer recurrence in breast and prostate cancer surgery.9,10 However, a series of subsequent retrospective studies, especially on other types of surgeries, showed that epidural analgesia did not improve oncologic outcomes such as recurrence-free survival.11–13 However, these retrospective studies included a relatively small number of patients (99–509). The only randomized clinical trial reported similar recurrence-free survival between epidural and control groups (epidural group hazard ratio [95% CI]: 0.95 [0.76–1.17]; P= .6). However, the study included various abdominal surgeries (esophagogastric, hepatobiliary, bowel, renal, prostate, etc, n = 446) that have different morbidities and mortalities.14 Our study was a large cohort study on surgery for lung cancer and included 2 types of regional analgesia. Patients in the PCA group received the highest dose of opioid and the patient in the PVB group received the lowest dose of opioid. Our results support no difference in cancer recurrence between systemic opioid use and regional analgesia including PVB in lung cancer surgery. Perioperative systemic opioids may inhibit immune surveillance; however, their effect and duration of immune suppression seem to be clinically insignificant for lung cancer surgery.

We found higher overall survival in the PVB and PCA groups compared to the TEA group in multivariable Cox proportional hazards analysis. The PCA group had more favorable patient and surgical characteristics than the other groups. This may have resulted in lower composite complications (P< .0001) and in-hospital mortality (P= .023), and led to a higher overall survival during the early postoperative period in the PCA group. This possibility is supported by the findings that the initial higher overall survival was not sustained and decreased over time and that the higher overall survival disappeared after PSM in the PCA group. The PVB group showed higher overall survival than the PCA and TEA groups after PSM. Several studies reported a better overall outcome for patients with poor physical status when using regional analgesia compared to systemic opioids. Holler et al15 reported a greater benefit from regional analgesia for patients with ASA 3 or 4. Regional analgesia showed survival benefits for patients older than 64 years undergoing colorectal cancer surgery.11 Therefore, better management of high-risk patients may be the reason for the longer overall survival in the PVB group compared with the PCA group.

Between the 2 types of regional analgesia, the PVB group showed higher overall survival than the TEA group. The reasons are not clear but may be related to a high failure rate and unstable hemodynamics frequently seen with TEA. In previous studies, TEA failed more frequently than is generally recognized16 and had failure rates of up to 31%.17–19 Hemodynamic compromise is also frequently observed in TEA patients, especially during the perioperative period of fluid-restricted thoracic surgery.20 If hypotension occurs, vasopressors/inotropes and fluid should be infused, and TEA infusion should be decreased or stopped to establish stable hemodynamics. This process may complicate patient management. In contrast to TEA, PVB was performed under direct visualization and does not generally lead to significant hemodynamic changes.20,21 In our study, opioid consumption until POD2 was higher in the TEA group than the PVB groups which may indicate a higher failure rate in the TEA group. Patients who received TEA showed a higher incidence of hypotension and received vasopressor/inotrope more frequently than the PVB group until POD2. This may reflect unstable hemodynamics in the TEA group. Accordingly, the PVB group showed a lower incidence of composite complications than the TEA group (odds ratio [95% CI]: 0.77 [0.60–0.99]; P= .0427). Therefore, a higher success rate and more stable hemodynamics in the PVB group may have resulted in fewer complications and better postoperative management and overall survival in the PVB group compared to the TEA group. Consumption of opioid was lower and incidence of hypotension was comparable to the PCA group in the PVB group.

The limitations of this study include its retrospective nature, which may have introduced confounding or unmeasured factors that affected the results. We tried to adjust for potential confounding factors by using multivariable analysis and PSM. Second, time to recurrence and overall survival are related data. Therefore, hazard ratios comparing groups on time to recurrence while censoring deaths may be biased by the computing risk of death. Third, we did not provide data on failure rate to support our assumption on TEA. Fourth, analgesics infusion was started near the end of surgery when a patient was weaning from anesthetics in all 3 groups. In terms of preventing surgical stress, start of infusion before surgical incision may be more preferable. Finally, our results represent the practice of a single center and may not reflect current practice at other institutions.

In conclusion, the type of analgesia used (systemic versus regional) during the postoperative period for patients with lung cancer was not related to cancer recurrence. Better long-term survival was observed in PVB than in PCA and TEA. Prospective randomized controlled trials using 3 analgesic methods are required to validate the effects of postoperative analgesia, especially PVB, on surgical patients.

Back to Top | Article Outline


Name: Eun Kyung Lee, MD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Name: Hyun Joo Ahn, MD, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Name: Jae Ill Zo, MD, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Name: Kyunga Kim, PhD.

Contribution: This author helped conduct the study and analyze the data.

Name: Dae Myung Jung, MD.

Contribution: This author helped analyze the data and prepare the submitted manuscript.

Name: Joo Hyun Park, MD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

This manuscript was handled by: Scott M. Fishman, MD.

Back to Top | Article Outline


1. Beagles K, Wellstein A, Bayer B. Systemic morphine administration suppresses genes involved in antigen presentation. Mol Pharmacol. 2004;65:437–442.
2. Beilin B, Shavit Y, Hart J, et al. Effects of anesthesia based on large versus small doses of fentanyl on natural killer cell cytotoxicity in the perioperative period. Anesth Analg. 1996;82:492–497.
3. Buckley A, McQuaid S, Johnson P, Buggy DJ. Effect of anaesthetic technique on the natural killer cell anti-tumour activity of serum from women undergoing breast cancer surgery: a pilot study. Br J Anaesth. 2014;113suppl 1i56–i62.
4. Mojadadi S, Jamali A, Khansarinejad B, Soleimanjahi H, Bamdad T. Acute morphine administration reduces cell-mediated immunity and induces reactivation of latent herpes simplex virus type 1 in BALB/c mice. Cell Mol Immunol. 2009;6:111–116.
5. Roy S, Chapin RB, Cain KJ, Charboneau RG, Ramakrishnan S, Barke RA. Morphine inhibits transcriptional activation of IL-2 in mouse thymocytes. Cell Immunol. 1997;179:1–9.
6. Sacerdote P, Bianchi M, Gaspani L, et al. The effects of tramadol and morphine on immune responses and pain after surgery in cancer patients. Anesth Analg. 2000;90:1411–1414.
7. Koebel CM, Vermi W, Swann JB, et al. Adaptive immunity maintains occult cancer in an equilibrium state. Nature. 2007;450:903–907.
8. Udagawa T. Tumor dormancy of primary and secondary cancers. APMIS. 2008;116:615–628.
9. Biki B, Mascha E, Moriarty DC, Fitzpatrick JM, Sessler DI, Buggy DJ. Anesthetic technique for radical prostatectomy surgery affects cancer recurrence: a retrospective analysis. Anesthesiology. 2008;109:180–187.
10. Exadaktylos AK, Buggy DJ, Moriarty DC, Mascha E, Sessler DI. Can anesthetic technique for primary breast cancer surgery affect recurrence or metastasis? Anesthesiology. 2006;105:660–664.
11. Gottschalk A, Ford JG, Regelin CC, et al. Association between epidural analgesia and cancer recurrence after colorectal cancer surgery. Anesthesiology. 2010;113:27–34.
12. Tsui BC, Rashiq S, Schopflocher D, et al. Epidural anesthesia and cancer recurrence rates after radical prostatectomy. Can J Anaesth. 2010;57:107–112.
13. Cata JP, Gottumukkala V, Thakar D, Keerty D, Gebhardt R, Liu DD. Effects of postoperative epidural analgesia on recurrence-free and overall survival in patients with nonsmall cell lung cancer. J Clin Anesth. 2014;26:3–17.
14. Myles PS, Peyton P, Silbert B, Hunt J, Rigg JR, Sessler DI; ANZCA Trials Group Investigators. Perioperative epidural analgesia for major abdominal surgery for cancer and recurrence-free survival: randomised trial. BMJ. 2011;342:d1491.
15. Holler JP, Ahlbrandt J, Burkhardt E, et al. Peridural analgesia may affect long-term survival in patients with colorectal cancer after surgery (PACO-RAS-Study): an analysis of a cancer registry. Ann Surg. 2013;258:989–993.
16. Hermanides J, Hollmann MW, Stevens MF, Lirk P. Failed epidural: causes and management. Br J Anaesth. 2012;109:144–154.
17. Alagoz A, Sazak H, Tunc M, et al. Teaching practices of thoracic epidural catheterizations in different grade of anesthesia residents. Braz J Anesthesiol. 2016;66:1–6.
18. Desai A, Alemayehu H, Weesner KA, St Peter SD. Impact of epidural failures on the results of a prospective, randomized trial. Eur J Pediatr Surg. 2016;26:160–163.
19. Pöpping DM, Zahn PK, Van Aken HK, Dasch B, Boche R, Pogatzki-Zahn EM. Effectiveness and safety of postoperative pain management: a survey of 18 925 consecutive patients between 1998 and 2006 (2nd revision): a database analysis of prospectively raised data. Br J Anaesth. 2008;101:832–840.
20. Pintaric TS, Potocnik I, Hadzic A, Stupnik T, Pintaric M, Novak Jankovic V. Comparison of continuous thoracic epidural with paravertebral block on perioperative analgesia and hemodynamic stability in patients having open lung surgery. Reg Anesth Pain Med. 2011;36:256–260.
21. Kashiwagi Y, Iida T, Kunisawa T, Iwasaki H. [Efficacy of ultrasound-guided thoracic paravertebral block compared with the epidural analgesia in patients undergoing video-assisted thoracoscopic surgery]. Masui. 2015;64:1010–1014.
Copyright © 2017 International Anesthesia Research Society