PAIN after cardiac surgery is most often related to the median sternotomy. 1 Intensity of sternotomy pain has been found to be significantly higher on the first two postoperative days. Developing strategies for optimizing the analgesic management after cardiac surgery has assumed increased importance with the introduction of fast-track discharge protocols requiring early tracheal extubation. 2–4 In addition, postoperative pain is one of the primary concerns of patients admitted to the intensive care unit (ICU). 2,5–7 

The most often used analgesic methods for alleviating pain after cardiac surgery are intravenous opioid analgesics via  patient- or nurse-controlled delivery systems. 2,4 However, the use of parenteral opioid-based analgesic techniques can delay tracheal extubation as a result of their ability to produce drowsiness and respiratory depression. Opioid analgesics also produce adverse effects on the gastrointestinal tract (e.g.  , nausea, vomiting, and ileus) and bladder function (e.g.  , urinary retention). To reduce the adverse systemic effects of opioid analgesics, the use of central neuraxial techniques involving small doses of opioids has become increasingly popular. 8–11 In addition, adjunctive use of nonopioid analgesics that are devoid of respiratory depression has been suggested to facilitate the fast-tracking process after cardiac surgery. 5 However, a recent study evaluating the use of propacetamol (acetaminophen) to supplement opioid analgesia failed to demonstrate any clinically significant benefits. 12 

Hynninen et al.  13 reported that the use of nonsteroidal antiinflammatory drugs as adjuvants to the opioid analgesics in the early postoperative period can produce opioid-sparing effects after cardiothoracic surgery. However, nonsteroidal antiinflammatory drugs can also produce adverse effects that complicate the recovery process after cardiac surgery (e.g.  , bleeding and renal dysfunction). A recent study by Ilfeld et al.  14 found that a local anesthetic infusion at the popliteal sciatic nerve decreased pain and opioid use and improved patient outcome after podiatric surgery. However, controlled studies evaluating the opioid-sparing effects of local anesthetic infusions at the surgical site have yielded conflicting results. 15,16 

Therefore, we designed this clinical study to examine the hypothesis that the infusion of a local anesthetic solution at the median sternotomy site would produce improved pain control and an opioid-sparing effect after cardiac surgery. A secondary objective of the study was to determine if the opioid-sparing effects of the local anesthetic infusion would facilitate the recovery process and improve patient outcome with respect to satisfaction with their pain management.

After obtaining approval from the Institutional Review Board at the University of Texas Southwestern Medical Center and St. Paul Medical Center in Dallas, 45 consenting patients who were scheduled to have an open-heart operation were enrolled in this randomized, placebo-controlled, double-blind study. The patients were assigned to receive one of three solutions: saline (control), bupivacaine 0.25%, and bupivacaine 0.5% at a fixed infusion rate of 4 ml/h during the first 48 h after cardiac surgery. Patient exclusion criteria included known allergy to local anesthetics, paragenes, or para-aminobenzoic acid, clinically significant kidney or liver disease, neurologic dysfunction, chronic use of systemic lidocaine or other antiarrhythmic drugs, insulin-dependent diabetes mellitus, or active bacterial infection.

All patients received intravenous midazolam (2–5 mg) for sedation in the preoperative holding area during placement of the standard invasive monitors. Monitors included those for five-lead electrocardiography, radial intraarterial blood pressure, central venous pressure, pulse oximetry, and capnography and a three-channel Holter monitor (for 24 h). Anesthesia was induced using intravenous sufentanil (0.3–1.0 μg/kg) and intravenous etomidate (0.2–0.3 mg/kg), and tracheal intubation was facilitated with intravenous rocuronium (1 mg/kg). Maintenance of anesthesia consisted of a sufentanil infusion (0.3–1.0 μg · kg⁻¹· h⁻¹) with desflurane (3–6% end-tidal concentration). At the end of surgery, desflurane and sufentanil were discontinued, and a propofol infusion was initiated at 25–50 μg · kg⁻¹· min⁻¹.

The cardiac surgeon placed catheters at either end of the sternotomy incision site during closure of the chest wound. The first catheter was placed in the subfascial plane above the sternum, and the second catheter was positioned above the fascia in the subcutaneous tissue at the median sternotomy incision site. Both catheters were attached to an elastomeric infusion pump (ON-Q® Pain Relief System; I-Flow Corporation, Lake Forest, CA) at the end of surgery before leaving the operating room. This infusion pump consists of a disposable 270-ml elastomeric reservoir that delivers a continuous infusion at a fixed rate of 2 ml/h through each of the two 5-in-long soaker catheters. A one-way valve prevents any backflow into the drug reservoir, which contained saline, bupivacaine 0.25%, or bupivacaine 0.5% solution. An operating room pharmacist who was not involved in the study filled the reservoir with the study medication.

Postoperative evaluations were performed by a blinded observer 4, 8, 12, 24, 48, and 72 h after tracheal extubation, including assessments of the level of sedation (using the Ramsay scale), 17 severity of pain using an 11-point verbal rating scale, where 0 = no pain and 10 = worst pain imaginable, the amount of opioid analgesic medication administered (i.e.  , patient-controlled analgesia morphine use), and patient satisfaction with their postoperative pain management using a 100-point verbal rating scale, where 1 = highly dissatisfied to 100 = highly satisfied. Patients were also questioned regarding the occurrence of pruritis, nausea, and vomiting at each of the previously described time intervals. Urinary retention was defined as the need for reinsertion of a bladder catheter because of an inability to void within 8 h after its removal. The time intervals from arrival in the ICU to tracheal extubation, first bowel sound, first bowel movement, ambulation with assistance, tolerating normal dietary intake, and discharge from the ICU and the hospital were also assessed by the nursing staff according to the institution's standardized protocol for all patients undergoing cardiac surgery. A sample of the mediastinal drainage fluid was obtained 4 h after surgery, and blood samples were obtained 24 and 48 h after surgery for analysis of the bupivacaine concentrations.

An a priori  power analysis suggested that minimum group sizes of 10 would have 80% power to detect a difference between the groups if the local anesthetic solutions decreased the pain scores or patient-controlled analgesia opioid use by 75% compared with saline (based on a study involving the use of continuous local anesthetic infusions after major orthopedic surgery procedures 14), assuming a significance level of 0.05 and an SD of ± 5. Data are expressed as mean ± SD. Normality was assessed with the Kolmogorov–Smirnov test (by using the Lilliefors modification), and depending on the results, either parametric or nonparametric analysis was performed. Demographic, surgery, and anesthetic data as well as doses of drug administered were analyzed and compared by using the Student t  test. Pain, sedation, and patient satisfaction verbal rating scale scores were analyzed by using the Fisher exact or Mann–Whitney U-test, with P  < 0.05 considered statistically significant. Means were analyzed using a type III sum of squares ANOVA, where model = treatment. Frequencies of events were analyzed using a chi-square test. Correlation coefficients for the relationships between the mean total patient-controlled analgesia morphine use and the median verbal pain scores were calculated for each of the three treatment groups using a multiple regression model.

A total of 45 patients were enrolled in the study. However, nine patients were excluded from the data analysis because of failure to initiate the therapy or protocol violations (e.g.  , premature termination of the therapy). In six cases, the local anesthetic catheters either were not placed by the surgeon at the end of the operation (four cases) or were inadvertently removed within 24 h (two cases). Two patients developed serious bradyarrhythmias during the postbypass period (and were withdrawn from the study before initiating the therapy), and one patient (in the control group) developed a cerebrovascular accident and died on the second postoperative day. There were no differences between the three study groups with respect to age, weight, height, American Society of Anesthesiologists physical status, midazolam premedication, anesthesia and surgery times, aortic cross-clamp time, cardiopulmonary bypass time, and the amount of intraoperative opioid medication administered (table 1). Of the 12 controls, 10 underwent coronary artery bypass grafting, and two underwent valve replacement procedures. In the bupivacaine-0.25% group, six patients underwent coronary artery bypass grafting, three underwent valve replacements, two underwent atrial septal defect repairs, and one underwent an atrial myxoma excision procedure. In the bupivacaine-0.5% group, eight patients underwent coronary artery bypass grafting, three underwent valve replacements, and one underwent atrial septal defect repair.

No statistically significant differences were found in the average times to tracheal extubation (table 2). Patients in the bupivacaine-0.25% group recovered bowel sounds and were able to ambulate significantly earlier than were controls (table 2). The urinary catheter was removed significantly earlier in the bupivacaine-0.5% group than in the other two study groups (table 2). More importantly, all patients in the bupivacaine-0.5% group were able to sit up in a chair and ambulate with assistance on the first postoperative day. Patients in the bupivacaine-0.5% group were also able to meet discharge criteria from the hospital significantly faster than were patients in the other two study groups (table 2).

Patients in both bupivacaine treatment groups reported significantly less pain 4, 8, 12, and 24 h after surgery than did controls (fig. 1). The patient-controlled analgesia morphine use was significantly reduced in the bupivacaine-0.5% group compared with the control and bupivacaine-0.25% groups (figs. 2 and 3). In the two bupivacaine treatment groups, a positive correlation was found between the average patient-controlled analgesia morphine use and the patient's median pain verbal rating scale score, with r values of −0.168, 0.638, and 0.392 in the control, bupivacaine-0.25%, and bupivacaine-0.5% groups, respectively (fig. 3). Some statistically significant differences were also found in patient satisfaction with pain management scores between the two bupivacaine treatment groups and the control group (fig. 4). The overall incidences of postoperative nausea and vomiting during the 72-h observation period were 36%, 25%, and 33% in the control, bupivacaine-0.25% and bupivacaine-0.5% groups, respectively.

The bupivacaine concentrations in the mediastinal drainage fluid were consistently higher than the serum bupivacaine concentrations in both bupivacaine treatment groups (table 3). At 24 h after surgery, there was no significant difference in the serum bupivacaine concentrations between the two bupivacaine treatment groups. However, at 48 h (end of the infusion period), the serum bupivacaine concentrations were significantly higher in the bupivacaine-0.5% group than in the bupivacaine-0.25% and control groups (1.3 ± 0.7 vs.  0.5 ± 0.5 and <0.5 μg/ml, respectively). Importantly, the serum bupivacaine concentrations in all bupivacaine-treated patients were less than 4 μg/ml at the end of the 48-h infusion period. Complications related to the local anesthetic delivery system included a catheter tip that was inadvertently broken off during its removal from the incision site, which required reexploration of the wound under local anesthesia.

Mortality and morbidity after surgery seem to be related, in part, to the pathophysiologic response to the surgical trauma and to postoperative complications. 18 Postoperative pain and the use of large doses of opioid under medications can increase adverse effects, which contribute to delays in postoperative recovery after major surgery. 19,20 Therefore, the use of opioid-sparing analgesic techniques that can improve postoperative pain control with less opioid medication might facilitate the recovery process and rehabilitation (e.g.  , resumption of dietary intake and physical activity). 21 Analgesic techniques that improve pain control while minimizing the respiratory depressant effects of opioids in the early postoperative period are essential for fast-tracking patients through the recovery process after cardiac surgery. 8–11 

This “proof of concept” study demonstrated that the use of a continuous infusion of bupivacaine 0.5% (4 ml/h) at the sternotomy site during the early postoperative period is an efficacious method for improving pain control after open-heart surgery. Analogous to the findings described for patients undergoing major orthopedic surgery procedures, 14,22–26 continuous infusion of local anesthetics at the surgical site seems to produce an opioid-sparing effect, which can lead to greater patient satisfaction with their pain management after cardiac surgery. These findings also confirm the anecdotal reports describing local anesthetic (tissue) infusions after bone graft harvest and laparoscopic surgery procedures. 27,28 As suggested Todd and Brown, 29 the modern concept of postoperative analgesia appears to be moving toward peripheral local anesthetic techniques. 21 

Although we failed to demonstrate an improvement in all key outcome variables (e.g.  , tracheal extubation and length of ICU stay), this study involving a relatively small series of patients was not adequately powered to examine the effect of the local analgesic technique on all of the secondary outcome variables. The differing types of cardiac surgery procedures (e.g.  , coronary artery bypass grafting, valve replacement, and atrial septal defect repair) may have exerted a significant impact on the recovery and discharge times because of the small group sizes. The failure to demonstrate a benefit of the opioid-sparing action of the local anesthetic infusion on the times to tracheal extubation and ICU discharge was probably related to the fact that there was no attempt by the ICU staff to “fast-track” these cardiac patients at our university teaching hospital. Furthermore, opioid-related adverse effects (e.g.  , nausea, vomiting, urinary retention, constipation, and ileus) are not strictly dose related. It is also important to note that this study examined a fixed infusion (4 ml/h) of either bupivacaine 0.25% or bupivacaine 0.5%. Further studies are needed to determine the “optimal” infusion rate for and concentration of the local anesthetic. Another valid criticism of this study relates to the location of the local anesthetic infusion catheters. A better outcome might have been achieved with continuous paravertebral intercostal nerve blockade and/or placement of both catheters at deeper tissue planes (i.e.  , subfascial placement). 30 

Another concern relates to the high incidence of catheter-related problems (e.g.  , inadvertent removal during dressing changes and breakage on removal) in this “proof of concept” study. Although attempts were made to in-service all ICU nurses and surgery house staff, mistakes were made and interfered with the successful completion of the study in 13% of the cases. Our evaluation of the safety of this local anesthetic technique was limited to monitoring the patients’ cardiac rhythm and plasma drug concentrations. Of importance, even in the bupivacaine-0.5% group, plasma concentrations remained below the alleged toxic concentration for this local anesthetic in all patients (table 3). Although no ventricular arrhythmias were observed during the 48-h postoperative period, the use of higher infusion rates for bupivacaine 0.5% (> 4 ml/h) may result in plasma bupivacaine concentrations that could produce cardiotoxicity. Whether the use of ropivacaine or levobupivacaine would provide an additional margin of safety when using continuous local anesthetic infusion systems remains to be established in future comparative studies.

In conclusion, the use of a continuous infusion of bupivacaine 0.5% at the median sternotomy site reduced postoperative pain and the need for opioid analgesics after cardiac surgery. The use of infusion of bupivacaine 0.5% also improved some patient outcome variables (e.g.  , time to ambulation) and patient satisfaction with their pain management. This nonopioid pain management technique has the potential to facilitate the recovery process (i.e.  , fast-tracking) after open-heart surgery.