| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Thoracic Surgery, Roswell Park Cancer Institute and State University of New York, Buffalo, New York
b Department of Biostatistics, Roswell Park Cancer Institute and State University of New York, Buffalo, New York
c Department of Anesthesiology and Pain Medicine, Roswell Park Cancer Institute and State University of New York, Buffalo, New York
Accepted for publication December 30, 2008.
* Address correspondence to Dr Demmy, Department of Thoracic Surgery, Roswell Park Cancer Institute, Elm and Carlton Sts, Buffalo, NY 14263 (Email: todd.demmy{at}roswellpark.org).
Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5–8, 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Methods: Thirty patients who had non-rib-spreading thoracoscopic operations under general anesthesia were prospectively randomized to no local anesthetic infusion (control), intermittent bolus (30 mL every 6 hours), or continuous infusion (5 mL/h). Bupivacaine (0.25%) was delivered through the pleural infusion channel of a specially designed single silicone 28F chest tube. Total intravenous fentanyl patient-controlled analgesia (boluses with basal rate) infused in the first 24 hours postoperatively was the designated primary study end point. Escalations of analgesic therapy, including ketorolac administration, were standardized across all groups. Nurses assessed pain control at onset and every 6 hours by visual analog pain scales (VAPS, 100 mm). VAPS were repeated 10 minutes later to assess any opioid or bupivacaine bolus effects.
Results: No study-related adverse events occurred. Compared with controls, pooled VAPS scores and 24-hour fentanyl consumption were significantly lower for the intermittent and continuous administration groups (1753 vs 1180 vs 1177 µg/24 h, respective median; p = 0.04) Early (6-hour) VAPS analgesic responses were more certain for intermittent (10 of 10) and continuous (10 of 10) patients than controls (7 of 10, p = .04). Five continuous patients successfully maintained VAPS scores below 20 mm throughout the study vs 3 intermittent and 2 controls (p = .045).
Conclusions: Intermittent or continuous intrapleural bupivacaine infused through the chest tube reliably reduces postoperative pain and 24-hour opioid usage in thoracoscopy patients.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Many thoracic surgeons have since increased their use of thoracoscopic techniques and have observed marked pain improvement after chest tube removal. Intuitively, the chest tube should account for large proportion of pain stimulation in video-assisted thoracic surgery (VATS) because smaller wounds made without rib spreading are less traumatic. Also supporting this hypothesis is the finding that the use of fewer chest tubes reduces pain even in thoracotomy patients [4]. Therefore, this study was proposed to reevaluate the use of intrapleural bupivacaine as well as its focused delivery around a known pain stimulant, the intrapleural catheter.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
|
Fentanyl-Only (Group F)
After the operation, the effects of the anesthesia gradually decreased until patients became alert and started to detect pain. At the first complaint of pain (time A) and request for additional analgesia, the initial VAPS measurement was taken to provide a baseline score. Immediately after, group F patients received IV fentanyl while the PCA device was connected; thereafter, it served as their primary source of narcotic analgesia. The initial PCA settings were 40-µg/h basal rate, with 20 µg every 10 minutes as needed by the patient (160 µg/hour maximum). If the pain was not controlled within 4 hours of starting the fentanyl, a 50-µg/h basal rate with 30 µg every 10 minutes as needed (230 µg/h maximum) was used.
VAPS measurements were taken 10 times. Five occurred at times A, 6, 12, 18, and 24 hours after completion of the operation. The others occurred 10 minutes after A (to show the effect of fentanyl infusion) and 10 minutes after the other 6-, 12-, 18-, and 24-hour times. The latter VAPS measurements were made to be directly comparable with the before and after bupivacaine bolus measurements in the IB arm (below). At each of these time points, the total amount of opioids used by each patient was recorded directly from the PCA device.
Intermittent Intrapleural Bolus Delivery of Bupivacaine (Group IB)
At the first complaint of pain and request for additional analgesia (time A), the initial VAPS measurement was taken to provide a baseline score. Immediately after, group IB patients received an initial bolus of 0.25% bupivacaine intrapleurally. The initial bolus was reduced proportionally so that a full 30-mL dose could be given at 6 hours and then proceed normally thereafter. For instance, if the request occurred 3 hours before the 6-hour mark, 15 mL of bupivacaine was infused. These patients were also started on IV PCA fentanyl, exactly the same as patients in group F with the same escalation protocol.
Fentanyl Plus Continuous Intrapleural Delivery of Bupivacaine (Group C)
The time point A, VAPS measurement occurred just like the other arms. Immediately after, the intrapleural chest tube was primed by a 5-mL loading dose of bupivacaine, and the clamp was released on an On-Q (I-Flow Corp, Lake Forest, CA) 5 mL/h continuous intrapleural infusion of bupivacaine already connected to the chest tube in the operating room. The 5-mL/h rate of bupivacaine administration continued through the entire 24-hour postoperative period. These patients were also started on IV PCA fentanyl and managed like group F.
Escalation Protocol
If the fentanyl escalation described for group F was insufficient for any patient, additional 50-µg boluses were prescribed as needed. If pain control remained insufficient, ketorolac (15 to 30 mg IV) was administered after approval by the study investigators.
Early Study Termination
The chest tube and fentanyl infusions discontinued early in patients ready for hospital discharge before study completion. The following formula was used to adjust the value for the primary study endpoint to be consistent with the other patients in the group:
|
|
Statistical Analysis
The Kruskal-Wallis test indicated the sample of 30 subjects was predicted to achieve 80% power to correctly detect differences among the means at a significance level of 0.05 to allow detecting the difference incorrectly, if the standard deviation between treatments was 0.62 times of that within treatments under the normal distribution assumption.
The predetermined primary end point analysis was the comparison of the overall amount of fentanyl delivered to patients by each treatment. Comparison of fentanyl total and F adjusted among the three arms was done using the Kruskal-Wallis test. Pairwise comparisons were done in conjunction with a Bonferroni correction.
The second objective was the comparison of the rates of successful pain management among patients who received VATS by each treatment. The Fisher exact test was used to study the association between success rate and treatment groups. Success was defined as the value of each patient's VAPS scores at time of 6, 12, 18 and 24 hours showing improvement (reduction) of at least 20 mm compared with their own VAPS score at time 0. Descriptive statistics such as frequencies and relative frequencies were computed for categoric variables. Failure to control pain was defined as the inability to reduce the VAPS to less than 40 mm. Numeric variables were summarized using simple descriptive statistics such as the mean, standard deviation, and range. A 0.05 nominal significance level was used in all testing. All statistical analyses were performed using SAS 9.1.3 software (SAS Institute, Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
|
Figure 3 plots total fentanyl consumption and discrete VAPS pain scores over time for each group. These were similar with similar median VAPS scores at pain onset. Thereafter, pain scores generally fell below the acceptable pain threshold of 40, with the continuous group having the lowest VAPS score despite less fentanyl use. Successful pain control (< 20 mm) was demonstrated best in the continuous group.
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Thoracic surgeons and anesthesiologists have tried various regional techniques to reduce postoperative discomfort. Several prospective, randomized studies showed benefits to intrapleural anesthetic deliveries [1–3; 7–9]. However, other investigators who attempted to replicate those results found that intrapleural bupivacaine alone could not lower the pain scores to acceptable levels, that a placebo effect was at play, and that the methodologies themselves were prone to considerable variations [10–12]. In our literature review, we found variations in catheter types, placements, and potentials for migration in the pleural space. All could result in variable deliveries to the areas of pain causation as well as different losses of the local anesthetic through the chest tube system. Different dosing frequencies or concentrations of bupivacaine were also tested, with pleural deliveries of every 4 hours and higher doses (0.5%) sometimes yielding more effective results [13]. However, differences in pleural environment, such as infection or inflammation, can increase the systemic absorption of bupivacaine to toxic levels [14]. At RPCI we chose as an institutional standard a slightly lower dose and longer schedule to provide a wider range of safety for those at risk because of a smaller body size or unrecognized propensity for fast absorption.
To achieve better clinical outcomes, more reproducible methods were established elsewhere, such as placement of anesthetic catheters in subpleural, perivertebral spaces and the elimination of wound delivery altogether by using epidural anesthesia [15, 16]. Because of rapid discharges of patients after VATS, however, we found that epidural anesthetics with their associated side effects could potentially lengthen the stay of patients. Accordingly, we sought a simple way to control regional chest tube pain. We decided to reexamine intrapleural delivery, because many of the factors causing variability noted are not an issue with small wounds and a delivery device focused at a suspected site of major pain causation.
It is important to note the advantage of routinely using a chest catheter like the one used in this study. Because it takes no more time to place, the option to use regional local anesthesia comes with little additional investment. Its cost is similar to other soft silicone tubes that may be more comfortable than the common stiff polyvinyl chloride tubes. Furthermore, the side port is a convenient sclerosant portal for the occasional patient with a prolonged air leak who requires pleurodesis.
The cause of postoperative thoracic pain is complex and multifactorial, but still not well understood. Avoidance of trauma to the intercostal nerves is an important component. It is interesting that the degree of early thoracic pain may indicate the scope and duration of chronic pain after thoracotomy [17]. Our observations indicate that about two-thirds of VATS patients have dramatic reductions in chest pain after removal of the chest tube trigger, making it the target for this current research. It is reasonable to consider different designs such as release of the anesthetic at multiple sites along the catheter. Alternatively, the catheter could be made smaller, and in fact, the product used in this study comes in an 18F size that we use selectively.
Cost is an important consideration, although the drugs used in this study are relatively inexpensive. If the median reduction in the amount of fentanyl infused in the study is attributed to the local anesthetic, this results in a 24-hour cost-savings of $15.00. However, the incremental fully loaded cost of the intermittent boluses was approximately $17.60, and the continuous infusion system cost about $10 for the local anesthetic and $200 for the disposable pump that was usually discarded with unused local anesthetic. Alternatively, a standard infusion pump could be used for about $5 but will require more nursing attention. Depending on nursing and disposable costs for continuous or intermittent anesthetic infusions, the return on investment will change but will probably be justified if reduced pain translates into faster discharge and fewer complications. The nurses at RPCI readily taught each other intermittent infusions because of the satisfaction of immediately reducing pain by infusing bupivacaine for their patients who were not responding to narcotic therapy. Intermittent therapy, however, tends to cause swings in drug levels that may be less effective for pain control than a continuous infusion.
One limitation of this study is that there could be more effective applications of intrapleural bupivacaine, such as higher concentrations or shorter intervals between infusions. We chose a schedule unlikely to cause systemic toxicity, but certain patients were probably underdosed. This might have accounted for a tachyphylactic effect seen in the experimental groups, where a few subjects began to ask for additional pain control options midway through the study. Although not addressed in this research, it is reasonable to surmise that if bupivacaine demonstrates to a patient that a lower pain sensation is attainable, then that patient will try to reach that level of analgesia should the effect of bupivacaine wear off. Patients in the continuous group may have asked for more adjustments because they did not have the distraction of bolus therapy.
Another limitation is that the study did not test alternate methodologies for controlling postoperative pain, such as epidural anesthesia, cryoablation analgesia, perivertebral blockade, and preemptive anesthesia [9, 15, 18]. Unfortunately, many of these techniques require additional operative time, specialized equipment, modification of the intraoperative practices, and may have associated complications or prolonged effects. It was presumed (correctly for most subjects) that only a brief course of pain control was needed, so we chose this simplistic approach and brief study duration.
The relatively brief stays in this group might have been because of careful case selection, and therefore, one might wonder whether the proposed method is actually durable beyond 24 hours or applicable to a broader population. Alternatively, we wonder whether the careful attention to pain management may have accounted for the excellent results of these cohorts.
This was a relatively small study, powered to detect the large difference akin to great relief of pain that we observed after delivering bupivacaine boluses. The intermittent group did not have as significant change from control; however, it had more patients with poor lung function, which is a risk factor for chronic pain after thoracotomy [19]. Despite this potential disadvantage, the group performed well. A larger study would lead to more evenly balanced groups and better power to detect more subtle differences in pain control.
We also do not know whether this method of bupivacaine delivery was optimal for all participants. Further study of delivery is warranted, in particular, whether bupivacaine adequately bathes the intercostal nerve below the tube insertion point or the subcutaneous tissues. Also, nonsteroidal drugs like ketorolac may have been used better as a routine medication for all patients rather than as part of an escalation protocol. We moved away from its routine use because of its occasional adverse effects on renal or platelet function.
In addition, this study could be constructed differently in an attempt to determine the response to intrapleural bupivacaine early in order to titrate the dosage to an effective level or schedule. That way, patients would only undergo treatments that seem to have a measurable response. This might be useful before attaching a longer duration infusion system such as the On-Q product, which has a modest cost and potential wastage of drug in the reservoir if discarded early or ineffective. A crossover design could have been used to see if bupivacaine improved the pain scores for the fentanyl-only group, or the use of saline controls rather than bupivacaine could have determined whether there was a placebo effect in this trial. An extra open thoracotomy control arm, while interesting, was not included because of the disappointing earlier research cited, the bias of greater pain stimuli expected for these patients, and institutional selection bias (because VATS is attempted in more than 80% of pulmonary cases).
Intermittent or continuous intrapleural bupivacaine infused by chest tubes was reliable enough to demonstrate a reduction in postoperative pain and 24-hour opioid usage in VATS patients in this modest clinical trial. Continuous therapy will become our preferred approach as a consequence of this study. Intrapleural anesthetic therapy should be considered as an adjunct to pain control after thoracoscopy but may increase patient requests for additional pain interventions because of tachyphylaxis, wider swings in pain stimulation, demonstrating that less pain is attainable, or unknown mechanisms.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
And secondly, you mentioned in your manuscript that you had previously used an intermittent bolus, but it appears in your outcomes that a continuous infusion provided the best results. How have the results from this study changed your practice and do you have any plans for future studies to answer some of your questions and contradictions in your study? I, again, appreciate you providing me the manuscript in advance, I look forward to your responses, and I appreciate the opportunity to discuss your paper.
DR DEMMY: Thank you, John. I think opioids in the older population is a problem, and that is why I was interested in using this type of therapy. My interest in thoracoscopy started in the high-risk elderly population, so this seemed to help a lot by reducing the narcotic needs for these older patients. You can add epinephrine to reduce systemic absorption and increase the dose from 0.25% to 0.5% to improve efficacy based on data from the 1980s. If you increase dose, you do increase the risk of systemic bupivacaine toxicity.
As far as changing, we are now going to try using continuous for all our patients. The tricky thing is if you use an On-Q pump [I-Flow Corp, Lake Forest, CA] and patients go home the next day, you are throwing away a $200 device full of bupivacaine. What we are wrestling with is how to decide who is going to need a longer stay, like a patient with an air leak, and perhaps start with a standard infusion pump, which is cheaper. We could then convert to an On-Q to improve patient mobility and require less nursing work.
DR BRYAN FITCH MEYERS (St. Louis, MO): Regarding the use of bupivacaine, the main questions are how to deliver it and how long to deliver it, and you are comparing intrapleural versus the extrapleural with the On-Q. Can you describe briefly how you place the On-Q? I have sent a few patients home with an On-Q pump in place if they get discharged on day one or two after either a VATS [video-assisted thorascopic surgery] wedge resection or a VATS lobe, and that is very similar to what other surgeons are doing after peripheral surgery, plastic surgeons and orthopedic surgeons, with the use of the On-Q. The portability of it and the self-regulated dosing of the flow rates makes it quite safe to send home with the patient. So you are actually decreasing narcotics on discharge as well, which you can't really do with a chest tube infusion.
DR DEMMY: I apologize for not making that clear. When you think On-Q, you may assume intraincisional or paravertebral. The pump was just hooked up to the side port of the same chest tube. My experience has been with VATS that most of the pain is from the chest tube, and just pulling the chest tube greatly reduces the patient's pain. We use this type of chest tube routinely for every patient, because it gives you the option of starting intrapleural anesthetic therapy. The other neat thing about it is that you can use it to perform pleurodesis on patients with prolonged air leak. You just hook up the syringe and shoot in your sclerosant. So it is a nice little adjunct in that way as well.
DR MEYERS: You don't use a catheter and don't use it in the way that the On-Q device is designed to be used?
DR DEMMY: No. That chest catheter has a standard Luer lock connector, so it makes it real easy to hook the On-Q pump.
DR MEYERS: Well, in that case I think a natural comparison would be using the On-Q the way it was designed to be used vs your system.
DR STEPHEN C. YANG (Baltimore, MD): Todd, that was a great study. I have been using intrapleural bupivacaine since my training 15 years ago. Actually the easiest thing that I find is to just use an epidural catheter that I insert next to the chest tube and we do a continuous infusion. Anecdotally and from the same studies that you talked about from the '80s, it has shown a great response in reducing pain and the need for parenteral narcotics.
My question is, when do you do epidurals, because I am trying to get away from the anesthesiologist wasting their time in trying to get epidurals in, and then I just go straight through the intrapleural catheter that we use continuously. It costs a lot less than the On-Q system and it is very easy to manage because we just set it at one rate and there is very minimal bupivacaine toxicity.
DR DEMMY: We have gotten away from epidurals because it kept patients in the hospital longer, and we saw occasional hypotension from epidural local anesthetic sympathetic blockade. The problem with placing the catheter at midwound like the trials from 20 years ago, is that sometimes it gets too close to the chest tube hole and the anesthetic gets sucked out immediately. That and other causes of delivery variability led to placing the catheter paravertebrally, which is the current preferred On-Q methodology.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
