Secondary Logo

Journal Logo

Acute pain

Comparison between systemic analgesia, continuous wound catheter analgesia and continuous thoracic paravertebral block: a randomised, controlled trial of postthoracotomy pain management

Fortier, Simon; Hanna, Halim A.; Bernard, Alain; Girard, Claude

Author Information
European Journal of Anaesthesiology: November 2012 - Volume 29 - Issue 11 - p 524-530
doi: 10.1097/EJA.0b013e328357e5a1
  • Free



Thoracotomy is the most painful surgery in the postoperative phase and engenders the greatest demand for analgesia treatments in the first 48 h after intervention.1 This type of surgery is associated with a high mortality rate (3.8% overall, 2.4% after limited resection, 3% after lobectomy and 7.7% after pneumonectomy) and is responsible for persistent pain in up to 75% of patients.2,3 The pain is generally caused by the surgery technique itself, namely resection, or muscular and costal stretching, and can make breathing difficult and hamper physiotherapy. The wound (located at the 4th or 5th intercostal space) and the drainage holes (from the 8th to 11th intercostal space) are the main sources of pain. Generally, the pain is multimodal, with an extensive level of innervation (parietal, visceral and projected pain).

The aims of successful analgesia are manifold, namely to improve the patient's comfort by decreasing painful sensations, to contribute to improved clinical outcome after surgery, to decrease postoperative morbidity (by optimising patient ventilation, i.e. good chest expansion, efficient coughing and respiratory physiotherapy) and to shorten the durations of hospital stay and convalescence.

Systemic analgesia techniques have never been proven efficacious in pain management postthoracotomy. Even morphine, the gold standard for postoperative analgesia in most types of surgery, can lack efficacy during cough efforts. Moreover, high doses of morphine can engender complications such as sedation, respiratory depression, acute urinary retention, nausea, vomiting and pruritus. Among loco-regional analgesia techniques, only thoracic epidural has been shown to be superior and has become the gold standard in postthoracotomy pain care.1,3 However, with up to 30% of patients ineligible for thoracic epidural due to contraindications and treatment failures, alternative techniques are required.

First among these alternatives are continuous wound catheters (CWCs). In this technique, an analgesic agent (usually local anaesthetic) is infused in the vicinity of the wound (subcutaneous space, muscular aponeurosis, peritoneum).

An alternative for pain management therapy is thoracic paravertebral block (TPVB), which comprises infusion of a local anaesthetic into the paravertebral space, adjacent to the thoracic vertebra close to where the spinal nerves emerge from the intervertebral foramen. This results in ipsilateral somatic and sympathetic nerve blockade. TPVB is possible at any of the intervertebral spaces (from lumbar plexus block to deep cervical plexus block). The aim of our study was to evaluate the efficacy of CWC and TPVB in comparison with systemic morphine administration after thoracotomy.



Ethical approval for this study was initially provided by the Ethical Committee ‘Comité de Protection des Personnes Est I’, based in University Hospital Le Bocage, Dijon, France, (Chairperson Professor Bernard Blettery), on 18 January 2007 (number EudraCT 2006–0066242–34). Approval for an amendment to the protocol was obtained from the same committee on 16 April 2009, when we decided to allow patients to bypass the ICU and go straight to postoperative care in the surgical ward.

The present study is a randomised, controlled, open label trial. Inclusions began in April 2007 and were completed in February 2010. All patients gave written informed consent.

Inclusion criteria were as follows: patients aged 18 years and over, American Society of Anesthesiologists (ASA) physical status between II and III and scheduled to undergo pulmonary surgery by axillar thoracotomy or posterolateral thoracotomy (wedge resection, lobectomy, pneumonectomy or atypical resection), without associated pleurectomy.

Exclusion criteria were as follows: patients aged less than 18 years; pregnant women, or women of childbearing age not using contraception; patients undergoing emergency surgery; patients with a history of chronic pain; patients on morphine treatment; allergy to paracetamol, local anaesthetics or morphine; patients unable to comprehend the evaluation instruments used [Visual Analogue Scale (VAS) and Postoperative Nausea and Vomiting (PONV) Intensity Scale] and patients under legal guardianship.

Anaesthetic technique

Preoperative medication consisted of oral hydroxyzine (1 mg kg−1) given 1 h before anaesthesia. We used 25 mg tablets of hydroxyzine and the dose was rounded close to the patient's body weight. After standard monitoring, intravenous anaesthesia was induced with sufentanil (0.2 to 0.3 μg kg−1), propofol (1 to 3 mg kg−1) and cisatracurium (0.15 mg kg−1). Tracheal intubation with desflurane and 50% oxygen/air combination was used to maintain anaesthesia. Additional sufentanil and cisatracurium could be administered at the anaesthesiologist's discretion, depending on the duration of surgery and patient tolerance. One-lung or bipulmonary ventilation was used, depending on surgical conditions. Forced-air warming blankets were used on all patients during surgery to prevent hypothermia.

Surgical technique

With the patient in a lateral decubitus position, axillary or posterolateral thoracotomy was performed at the 5th intercostal space. One or two thoracic drains were placed between the 8th and 11th intercostal spaces at the end of the surgery, in patients who underwent limited pulmonary resection. No drains were placed when pneumonectomy was performed. At the end of the procedure, the surgeon inserted the CWC on both sides of the incision using Wheatley's technique, previously described elsewhere.4 After completion of the procedure, the placement of catheters involves introduction through a peel-away trocar 3 to 5 cm anterior and superior to the incision. One catheter is placed at the level of the pericostal sutures adjacent to the intercostal nerve bundle, and the other is placed above the fascia in the subcutaneous space. Alternatively, according to treatment allocation, the surgeon positioned the multiperforated TPVB catheter (Perifix Tip; B Braun, Boulogne-Billancourt, France) under visual control as per Sabanathan's technique.5 The original description involves reflecting the parietal pleura from the posterior wound margin on the vertebral bodies to create an extrapleural paravertebral pocket into which a percutaneously inserted catheter is placed against the angles of the exposed ribs.

Study design

Patients were randomised to one of three groups by random selection of envelopes performed in the operating theatre. The envelopes were prepared in advance and contained a computer-generated randomisation schedule indicating the technique to be used.

All patients received intravenous analgesia with paracetamol, initiated before the end of surgery, and maintained at a dose of 1 g per 6 h slow intravenous infusion, as well as morphine by patient-controlled analgesia (PCA), with a loading dose of 1 mg (bolus), a lockout interval of 7 min and a maximum dose of 34.2 mg per 4 h. (Gemstar Infusion Systems, Hospira, Meudon La Foret, France).

Patients in the PCA group received systemic analgesia only. Patients in the TPVB group underwent insertion of the paravertebral catheter by the surgeon. The catheter was flushed with ropivacaine (2 mg ml−1) just before the incision was closed, and then a bolus dose was infused before the patient woke up (1 mg kg−1 ropivacaine 7.5 mg ml−1). Analgesia was maintained by electric infusion pump (Pilot; Fresenius Vial, Brezins, France) at a constant speed of 0.3 mg kg−1 h−1 of ropivacaine solution 2 mg ml−1 for 48 h.

The CWC group underwent insertion of a double catheter (Multiholed catheter, ON-QPainBuster; B Braun; flow rate 2 × 2 ml h−1, 24 holes, length 12.5 cm), which was also flushed with ropivacaine (2 mg ml−1) just before the incision was closed. An induction dose of 5 ml ropivacaine 2 mg ml−1 in each branch of the catheter was administered before the patient woke up. The ON-Q Pain Buster pain relief system (I-Flow Corp, Lake Forest, California, USA), consisting of an elastomeric pump filled with ropivacaine 2 mg ml−1 was used to maintain analgesia at a constant flow rate of 2 ml h−1 in each branch of the catheter for 48 h.

In order to achieve pain control (VAS <3), intravenous morphine titration was performed at wake-up. In the recovery room, morphine PCA was provided. Doses were subsequently modified (bolus) by the anaesthetists according to the patient's VAS score during rounds in the ICU or conventional care unit. Patient pain levels were evaluated from wake-up, by assessing VAS at rest and after coughing. The morphine and PCA doses were recorded by the recovery room nurses. In order to optimise postoperative monitoring, the first 75 patients were hospitalised for at least 24 h in the surgical ICU. Patients were monitored by the anaesthetist during this time, and the nurse in charge collected the data. After obtaining new ethics committee approval, all further patients returned to the conventional surgical unit in which follow-up care was provided by the surgical team and the nurses on duty collected patient data. Catheters were removed after 48 h. The duration of PCA was 5 days.

Primary and secondary endpoints

The primary endpoint was VAS at rest, measured at 0, 1, 3, 6, 12, 24, 36 and 48 h by the nurse in charge of each patient.

Secondary endpoints were VAS during coughing recorded at 0, 1, 3, 6, 12, 24, 36 and 48 h, as well as total dose of morphine administered.

The 0 time point was defined as the time when the patient was extubated (in the recovery room).

Postoperative Nausea and Vomiting (PONV) Intensity Scale, pulmonary complications (secondary infection, atelectasis, severe pneumonia, severe hypoxia), cardiovascular complications (heart rhythm disorders), urological complications (urinary retention requiring catheterisation) and general complications (constipation, anaemia, disorientation, etc.) as well as any side effects (catheterisation-related, e.g. clinical signs of neurotoxicity or cardiotoxicity due to local anaesthetic, or clinical signs of local inflammation or infection, or morphine-related, e.g. postoperative nausea and/or vomiting) were recorded during the hospital stay. The duration of hospital stay was recorded at hospital discharge.

Data were collected in a standardised case report form (CRF). Baseline data were recorded at inclusion after allocation of the randomisation, and perioperative data were completed by nurses and/or the investigator and/or clinical research assistants.

Statistical analysis

The sample size was calculated in order to be able to detect a reduction of 30% in VAS, with an alpha risk of 5% and a beta risk of 20%, giving a total of 51 participants required per group.

Continuous variables are expressed as mean ± standard deviation and qualitative variables as number (percentage). Categorical variables were compared using the χ2 test and continuous variables by analysis of variance (ANOVA). Repeated measures ANOVA was used to compare variables with repeated measures over time, with correction using Huynt-Feldt's epsilon estimated at 0.29, to account for the exclusion of the measures from time point 0 and 1 h (excluded because median value was zero). Bonferroni correction was applied to account for multiple comparisons.

All analyses were performed using STAT 11 statistical software (StataCorp LP, College Station, Texas, USA).


The flowchart of patients included in the study is shown in Fig. 1.

Fig. 1
Fig. 1:
No captions available.

The baseline characteristics and surgical characteristics of the study population are shown in Table 1.

Table 1
Table 1:
Patient and surgical characteristics of the study population

Visual Analogue Scale at rest

There was a significant difference in VAS at rest among the three groups (Fig. 2; P < 0.0002 by repeated measures ANOVA). Individual intergroup comparisons at each time point are shown in Fig. 2 and identified five significant differences between treatment groups (P < 0.0026 for each). In the postoperative period, VAS at rest was lower at 0, 1, 3 and 6 h in the TPVB group than in the PCA group, and lower at 0 h in the TPVB group than in the CWC group. There was no significant difference between the PCA and CWC groups.

Fig. 2
Fig. 2:
No captions available.

Secondary endpoints

There was a significant decrease over time in VAS after coughing between the three groups (Fig. 3; P < 0.0001). Individual intergroup comparisons of the VAS values after coughing at different time points are shown in Fig. 3. By this analysis, nine comparisons showed significant differences (P < 0.003). In the postoperative period, cough VAS was lower in the TPVB group than in the PCA group at 0, 1, 3, 6 and 12 h, as well as at 0, 1, 6, 12 and 24 h between the TPVB and CWC groups. No significant differences were observed between the PCA and CWC groups.

Fig. 3
Fig. 3:
No captions available.

Morphine requirements were statistically different in the three groups over time (P < 0.02). Pair-wise comparisons showed a lower morphine consumption in the TPVB group than in the PCA morphine group and the CWC group only 24 h postoperatively (26.5 vs. 39.3 and 35.9 mg of morphine, P = 0.0036 and P = 0.016, respectively).

The data from 0 and 1 h were not included in the analysis as the median amount of morphine used was zero.

In the recovery room, titrated morphine doses were significantly lower in the TPVB group compared to the other two groups (TPVB group 4 ± 6 mg; PCA group 11 ± 6 mg; CWC group 10 ± 6 mg; P < 0.00001). However, total morphine doses (titration + PCA) were not different between groups at 48 h (PCA group 185 mg; TPVB group 143 mg; CWC group 173 mg).

The rate of complications did not differ between groups for PONV, pulmonary, cardiovascular, urological and general complications. No life-threatening technical complication related to the use of any of the three techniques occurred. There were no signs of neurological or cardiac toxicity related to the use of anaesthetics. Lastly, there was no significant difference in duration of hospital stay among groups.


In this study, VAS pain scores were lower at rest (at 0, 1, 3 and 6 h) and after coughing (at 0, 1, 3, 6 and 12 h) after thoracotomy, and there was a reduction in morphine titration dose when TPVB was used. In thoracic surgery, TPVB associated with intravenous morphine procures superior analgesia as compared to PCA alone. In the context of multimodal management, TPVB leads to significantly lower use of morphine, and consequently, results in a decrease in complications arising from its utilisation.6,7

A meta-analysis of 10 randomised studies reported that TPVB and thoracic epidural provided comparable pain relief after thoracic surgery.6 Furthermore, the results showed that TPVB was associated with a better side effect profile (hypotension, PONV, urine retention, pulmonary complications) and lower rates of pulmonary complications and block failure. Indeed, TPVB is currently recommended for thoracic surgery.8,9

In our study, we did not observe any complications related to catheter migration. We noted no spinal anaesthesia or complications related to epidural infusions, and no intravascular injection or signs of local toxicity were observed. Conversely, Naja and Lonnqvist10 reported an overall technical failure rate of 6.1% (performed in 620 adults and 42 children). Complications recorded included inadvertent vascular puncture in 6.8%, signs of epidural or intrathecal spread in 1.0% and pleural puncture in 0.8% (five patients), of whom three (0.5%) subsequently developed pneumothorax. An older study of the feasibility of TPVB reported an overall failure rate of 10.1%, with complications such as hypotension in 4.6%, vascular puncture in 3.8%, pleural puncture in 1.1% and pneumothorax in 0.5%.11 In most studies assessing TPVB, a percutaneous technique was used, whereas in our study, the catheters were positioned using the surgical technique described by Sabanathan.5 Our surgeon positioned the multiperforated TPVB catheter (Perifix Tip; B Braun) under visual control. The original description involves reflecting the parietal pleura from the posterior wound margin on the vertebral bodies to create an extrapleural paravertebral pocket into which a percutaneously inserted catheter is placed against the angles of the exposed ribs. This difference in the techniques used to place the catheters could explain the lower complication rate observed in our study. This supports the hypothesis of Davies et al.6 that the surgical procedure could be safer and more logical, despite the lack of hard scientific evidence to prove its superiority. However, the surgical approach for TPVB seems to be associated with longer duration of surgery and an increase in bleeding. Ten percent of patients were unable to undergo TPVB due to pleural tearing during disconnection of the osteoligamentous structures. This failure rate is in line with the 6 to 10% failure rate reported with the percutaneous technique in the literature.10–13 Complications may be prevented or at least decreased by using ultrasound guidance, currently being developed.14,15

Our data seem to show the safety of TPVB in terms of respiratory, infectious and haemodynamic side effects. TPVB does not require any specific monitoring other than the conventional follow-up of patients in the postoperative period and does not require the patient to be hospitalised in a costly ICU.6,13

In the present context of multimodal analgesia management, another loco-regional anaesthetic technique that is widely used is continuous wound infusion. In our study, we did not observe that CWC yielded superior analgesia after thoracic surgery. However, CWC is reportedly effective in complex and painful surgeries. Zohar et al.16 showed that significantly less morphine was required with CWC in the context of total abdominal hysterectomy, a complex surgical procedure with a large visceral component. These findings were confirmed in a systematic review by Liu et al.17 who showed that the use of CWCs was associated with improved analgesia at rest and after mobilisation, reduced morphine use and related side effects, increased patient satisfaction and a tendency towards reduced hospital stay. Notably, in the review by Liu et al., these benefits were observed across a wide range of surgical procedures (cardiac, thoracic, abdominal, gynaecological and orthopaedic). However, within the thoracic surgery procedures alone, there was a discrepancy in the catheter positions (two interpleural, four intrapleural, three intercostal and only two extrapleural catheters). The position of certain catheters (closer to PVB or intercostal block) could explain the positive results observed by Liu et al. and the mismatch with our findings. In the review by Liu et al., inclusion criteria varied between studies, and there were a wide range of local anaesthetics used, with different modalities of administration, and various clusters of surgery (such as lung resection, oesophagectomy, midsternotomy), and these factors all render the interpretation and comparison of results difficult.

In the context of thoracotomy, only one retrospective comparison of 110 patients reported an advantage of CWC compared to epidural infusion.4 In this study, 0.25% bupivacaine was administered at a dose of 4 ml h−1 through two 20-gauge catheters using the ON-Q Pain relief system also used in our study. Wheatley et al.4 showed reduced VAS scores and use of narcotics in the CWC group, without any local or general complications. Pain caused by drains (located at a distance from the main incision, i.e. at the 8th to 11th intercostal space) could be one explanation for the failure of CWC to provide effective analgesia in this indication.

The techniques used for catheter placement in thoracic surgery are based on reports from abdominal surgery in the literature, in which it is acknowledged that efficacy is best when local anaesthetics are administered at a deep level rather than subcutaneously.17–19 However, data from cardiothoracic surgery are sparse. In one study of pain after standard median sternotomy, continuous wound infusion with ropivacaine at a dose of 2 mg ml−1 for 48 h led to a significant decrease in postoperative pain, narcotic analgesia requirements and length of hospital stay.20 These results were confirmed by White et al.,21 who showed that bupivacaine 0.5% administered through a double-catheter decreased morphine use by half and also reduced length of hospital stay and had a positive effect on mobilisation. In the context of thoracotomy, retrospective study by Wheatley et al.4 showed a benefit of CWC when one catheter is placed at the level of the pericostal sutures and the other above the fascia in the subcutaneous space. In our study, at the end of the procedure, the surgeon inserted the CWC on both sides of the incision. The placement of catheters involves introduction through a peel-away trocar 3 to 5 cm anterior and superior to the incision. One catheter is placed at the level of the pericostal sutures adjacent to the intercostal nerve bundle and the other is placed above the fascia in the subcutaneous space. The optimal position for placement of CWC remains to be identified in the context of thoracotomy.

In addition to the most effective catheter position, the optimal volumes, concentrations and types of local anaesthetic also remain to be defined. In cardiac surgery, White et al.21 showed a significant benefit with bupivacaine 0.5% compared with bupivacaine 0.25% at a fixed infusion rate of 4 ml h−1.

In our study, we used 8 mg h−1 ropivacaine by pericostal and subcutaneous approach for the CWC group, compared with 20 mg h−1 for the TPVB group. These lower doses (in terms of both concentration and flow rate) may have been insufficient to achieve significant efficacy, considering the extent of the space into which the medication spreads. To the best of our knowledge, no study has ever specifically examined the spread of local anaesthetics from CWC in thoracic surgery.

Lastly, the CWC technique is possible in anaesthetised patients and is simple and rapid to put in place. At the end of the procedure, the surgeon positions the catheter near or on the wound, under direct visual control. There is no risk of nerve lesions, because there is no need for nerve localisation. CWC requires virtually no handling by nurses, thereby limiting the infectious risk. Thus, the equipment is basic and the complication risks negligible, making it possible for patients to remain in conventional care units during the postoperative phase, without the need for expensive intensive care. The cost of CWC is very low because of its simplicity, rapidity and ease of use.17 To date, no toxicity events have been reported with local anaesthetics.17,22

Limitations of the study include the number of patient dropouts, rendering the results of the primary outcome vulnerable as regards the power of the study. Also, we used fixed doses of local anaesthetic and did not add adjuvants, which may enhance the analgesic effect of both the TPVB and CWC infusions. Large-scale CWC investigations are warranted, in order to identify the optimal catheter positions, doses, volumes and local anaesthetic concentrations. Adjuvant therapy (morphine or clonidine) could increase analgesic efficacy of this technique, as could epidural analgesia associations.23 Finally, further investigation of the extent of diffusion and serum levels of local anaesthetics should also be performed in view of the positioning of CWC, considered to be at high risk of systemic absorption. It would also be interesting to assess the impact of CWC on chronic pain, which was not considered in our study.


Our results show the benefit of TPVB regarding postthoracotomy pain and morphine requirements, at rest, and more particularly after coughing or respiratory physiotherapy. These findings are in agreement with those of the literature and current recommendations, and underline the important position of TPVB in postthoracotomy pain care.

In contrast, CWC failed to decrease pain and morphine consumption. This result combined with current data seems to be insufficient to recommend its use in this indication and requires further investigation and larger scale studies.


Assistance with the article: the authors would like to thank Fiona Ecarnot for translation and editorial assistance.

Financial support and sponsorship: none declared.

Conflicts of interest: none declared.


1. Kavanagh BP, Katz J, Sandler AN. Pain control after thoracic surgery. A review of current techniques. Anesthesiology 1994; 81:737–759.
2. Bernard A, Rivera C, Pages PB, et al. Risk model of in-hospital mortality after pulmonary resection for cancer: a national database of the French Society of Thoracic and Cardiovascular Surgery (Epithor). J Thorac Cardiovasc Surg 2011; 141:449–458.
3. Soto RG, Fu ES. Acute pain management for patients undergoing thoracotomy. Ann Thorac Surg 2003; 75:1349–1357.
4. Wheatley GH 3rd, Rosenbaum DH, Paul MC, et al. Improved pain management outcomes with continuous infusion of a local anesthetic after thoracotomy. J Thorac Cardiovasc Surg 2005; 130:464–468.
5. Sabanathan S, Smith PJ, Pradhan GN, et al. Continuous intercostal nerve block for pain relief after thoracotomy. Ann Thorac Surg 1988; 46:425–426.
6. Davies RG, Myles PS, Graham JM. A comparison of the analgesic efficacy and side-effects of paravertebral vs epidural blockade for thoracotomy: a systematic review and meta-analysis of randomized trials. Br J Anaesth 2006; 96:418–426.
7. Marret E, Bazelly B, Taylor G, et al. Paravertebral block with ropivacaine 0.5% versus systemic analgesia for pain relief after thoracotomy. Ann Thorac Surg 2005; 79:2109–2113.
8. Beaudroit L, Ripart J. Nerve blocks of the trunk: indications, techniques, advantages and complications. Ann Fr Anesth Reanim 2009; 28:e79–e83.
9. Joshi GP, Bonnet F, Shah R, et al. A systematic review of randomized trials evaluating regional techniques for postthoracotomy analgesia. Anesth Analg 2008; 107:1026–1040.
10. Naja Z, Lonnqvist PA. Somatic paravertebral nerve blockade. Incidence of failed block and complications. Anaesthesia 2001; 56:1184–1188.
11. Lonnqvist PA, MacKenzie J, Soni AK, Conacher ID. Paravertebral blockade. Failure rate and complications. Anaesthesia 1995; 50:813–815.
12. Marret E, Ohana M, Bonnet F. Paravertebral block. Ann Fr Anesth Reanim 2006; 25:249–252.
13. Karmakar MK. Thoracic paravertebral block. Anesthesiology 2001; 95:771–780.
14. Hara K, Sakura S, Nomura T, Saito Y. Ultrasound guided thoracic paravertebral block in breast surgery. Anaesthesia 2009; 64:223–225.
15. O Riain SC, Donnell BO, Cuffe T, et al. Thoracic paravertebral block using real-time ultrasound guidance. Anesth Analg 2010; 110:248–251.
16. Zohar E, Fredman B, Phillipov A, et al. The analgesic efficacy of patient-controlled bupivacaine wound instillation after total abdominal hysterectomy with bilateral salpingo-oophorectomy. Anesth Analg 2001; 93:482–487.
17. Liu SS, Richman JM, Thirlby RC, Wu CL. Efficacy of continuous wound catheters delivering local anesthetic for postoperative analgesia: a quantitative and qualitative systematic review of randomized controlled trials. J Am Coll Surg 2006; 203:914–932.
18. Beaussier M, El’Ayoubi H, Schiffer E, et al. Continuous preperitoneal infusion of ropivacaine provides effective analgesia and accelerates recovery after colorectal surgery: a randomized, double-blind, placebo-controlled study. Anesthesiology 2007; 107:461–468.
19. Moiniche S, Mikkelsen S, Wetterslev J, Dahl JB. A qualitative systematic review of incisional local anaesthesia for postoperative pain relief after abdominal operations. Br J Anaesth 1998; 81:377–383.
20. Dowling R, Thielmeier K, Ghaly A, et al. Improved pain control after cardiac surgery: results of a randomized, double-blind, clinical trial. J Thorac Cardiovasc Surg 2003; 126:1271–1278.
21. White PF, Rawal S, Latham P, et al. Use of a continuous local anesthetic infusion for pain management after median sternotomy. Anesthesiology 2003; 99:918–923.
22. Marret E, Vigneau A, Salengro A, et al. Effectiveness of analgesic techniques after breast surgery: a meta-analysis. Ann Fr Anesth Reanim 2006; 25:947–954.
23. Ballantyne JC, Carr DB, deFerranti S, et al. The comparative effects of postoperative analgesic therapies on pulmonary outcome: cumulative meta-analyses of randomized, controlled trials. Anesth Analg 1998; 86:598–612.

analgesia; continuous wound catheter; local anaesthesia; pain; paravertebral block; surgery; thoracotomy

© 2012 European Society of Anaesthesiology