Effect of Butorphanol Tartrate on Shock-Related Discomfort During Internal Atrial Defibrillation
Background—In patients with atrial fibrillation, intracardiac atrial defibrillation causes discomfort. An easily applicable, short-acting analgesic and anxiolytic drug would increase acceptability of this new treatment mode.
Methods and Results—In a double-blind, placebo-controlled manner, the effect of intranasal butorphanol, an opioid, was evaluated in 47 patients with the use of a step-up internal atrial defibrillation protocol (stage I). On request, additional butorphanol was administered and the step-up protocol continued (stage II). Thereafter, if necessary, patients were intravenously sedated (stage III). After each shock, the McGill Pain Questionnaire was used to obtain a sensory (S), affective (A), evaluative (E), and total (T) pain rating index (PRI) and a visual analogue scale analyzing pain (VAS-P) and fear (VAS-F). For every patient, the slope of each pain or fear parameter against the shock number was calculated and individual slopes were averaged for the placebo and butorphanol group. All patients were cardioverted at a mean threshold of 4.4±3.3 J. Comparing both patient groups for stage II, the mean slopes for PRI-T (P=0.0099), PRI-S (P=0.019), and PRI-E (P=0.015) became significantly lower in the butorphanol group than in the placebo group. Comparing patients who received the same shock intensity ending stage I and going to stage II, in those patients randomized to placebo the mean VAS-P (P=0.023), PRI-T (P=0.029), PRI-S (P=0.030), and PRI-E (P=0.023) became significantly lower after butorphanol administration.
Conclusions—During a step-up internal atrial defibrillation protocol, intranasal butorphanol decreased or stabilized the value of several pain variables and did not affect fear. Of the 3 qualitative components of pain, only the affective component was not influenced by butorphanol. The PRI evaluated pain more accurately than the VAS.
Although the amount of energy needed for internal cardioversion of atrial fibrillation is substantially reduced when using the right atrium–to–coronary sinus electrode configuration in combination with a biphasic shock waveform, intravenous administration of anesthetics is usually necessary to reduce the discomfort of patient shock.1 2 3 4 5 The anesthetic properties of some of these drugs and their intravenous mode of administration may diminish some of the advantages of internal cardioversion over external cardioversion. The shock-related discomfort has components of pain and fear.2 Many factors could contribute to the mechanism producing discomfort, including skeletal muscle contraction, nerve stimulation, and psychological factors.1 The purpose of this study was to evaluate, in a double-blind placebo-controlled manner, the effect of a transnasally administered opioid analgesic (butorphanol tartrate) on patient-reported pain and fear during internal cardioversion of atrial fibrillation.
Forty-seven consecutive patients with atrial fibrillation (29 men and 18 women; mean age 63±12 years, range 28 to 85) in whom conversion to sinus rhythm was clinically indicated underwent internal atrial defibrillation. Clinical evaluation of the patients included a history, physical examination, routine laboratory and thyroid function tests, a 12-lead ECG, and transthoracic and transesophageal echocardiography. The exclusion criteria of this study have been described previously.6 The administration of antiarrhythmic drugs was not interrupted for the procedure. Anticoagulation was interrupted 2 days before the procedure and restarted after the procedure. The study protocol was approved by the Medical Ethics Committee of the Academic Hospital Maastricht, and written informed consent was obtained from all patients.
Butorphanol tartrate (Bristol-Myers Squibb) is a synthetic opioid analgesic of the phenanthrene series. After intranasal administration, onset of analgesia occurs within 15 minutes, with a peak effect at 30 to 60 minutes after a 1-mg dose. The elimination half-life is 4.7 hours in individuals, ages 20 to 40 years, and 6.6 hours in individuals older than 65 years. The recommended starting dose for nasal administration is 1 mg (1 spray in 1 nostril), and a second 1-mg dose may be given 60 to 90 minutes after the first dose.7
The analgesic was supplied in a market package containing 2.5 mL of a 10 mg/mL solution of butorphanol tartrate. A solution of 0.8% sodium chloride was used as placebo. Drug and placebo were transferred to empty 2.5-mL bottles by a pharmacist of the Academic Hospital Maastricht, who also performed the randomization. Before each procedure, 2 bottles with metered pumps (0.09 mL/stroke) were delivered: 1 labeled with and containing placebo or butorphanol, according to the randomization, for the first intranasal administration and 1 labeled with and containing butorphanol for a potential second administration.
McGill Pain Questionnaire, Dutch Language Version
The McGill Pain Questionnaire (MPQ) is widely recognized as a valid and reliable instrument to measure pain, with a Dutch language version (MPQ-DLV) available.8 The MPQ-DLV contains 20 subclasses of 3 to 4 pain-descriptive words arranged in progressively increasing intensity. Twelve subclasses of 3 words describe sensory modalities of pain experience, 5 subclasses of 3 words represent affective aspects, and 3 subclasses of 4 words represent the evaluative dimension of pain. The list of words is scored according to the method described by Melzack.9 Two major indexes are obtained; the pain rating index (PRI) and the number of words chosen. As shown by Vlaeyen,10 the PRI and the number of words chosen appear to intercorrelate highly. Therefore we included only the PRI for further analysis. The PRI is based on the rank values of the words. The values of the words chosen by a patient are then added up to obtain a score separately for the sensory (PRI-S), affective (PRI-A), and evaluative (PRI-E) words, in addition to providing a total score (PRI-T).
All patients were instructed on their participation in answering the questions of the MPQ-DLV before the start of the procedure. After each shock, except when the patient was intravenously sedated, the 63 pain-descriptive words of the 20 subclasses were read to the patient. The patient could only select from each subclass the word describing the sensation at that moment. If in a subclass none of the words represented his pain, nothing was recorded and the examiner proceeded to the next group of words. Besides the list, 2 visual analogue scales (VAS), 1 for quantification of pain (VAS-P) and 1 for quantification of fear (VAS-F), were completed after each defibrillation shock. The VAS were numbered from 0 to 100; 0 represented no pain or fear (VAS-F) and 100 represented the maximum score for pain or fear during shock delivery. The PRI of the 3 individual pain dimensions (sensory, affective, and evaluative) and their total were calculated from the patient answers for each shock delivery.
Internal Atrial Defibrillation Protocol
Two 6F defibrillation catheters (Elecath, Electro-Catheter Corp) were inserted into the right femoral vein and positioned with 1 in the anterolateral right atrium and 1 in the coronary sinus. With the right atrial catheter used as the cathode and the coronary sinus catheter as the anode, a 6/6 ms biphasic defibrillation shock, synchronized to the R wave, was delivered from an external defibrillator (Ventritex HVS-02). To prevent inadvertent induction of ventricular fibrillation, shocks were delivered only after R-R intervals >500 ms.11 A quadripolar catheter (USCI) was positioned at the right ventricular apex to allow temporary ventricular postshock pacing.
Once all catheters were in position, 1 dose of nasal spray of either placebo or butorphanol was administered in 1 nostril (stage I; see Figure⇓). Randomization between placebo and butorphanol was only performed at stage I. After 30 minutes, a 90-V (≈0.5 J) shock was delivered from the Ventritex HVS-02. If the shock was unsuccessful at converting atrial fibrillation and if the patient agreed to have a stronger intensity shock delivered, a 0.5-J step-up protocol until 3 J and thereafter a 1-J step-up protocol was followed until successful cardioversion was obtained or until the patient requested discomfort relief. Five minutes was allowed for the next shock to be delivered. After each shock was delivered, the patient was asked the level of discomfort by use of the MPQ-DLV and the VAS for pain and fear. If the patient requested additional medication for discomfort, he or she received 1 dose of butorphanol (1 mg) nasal spray in the other nostril (stage II). After 30 minutes, the previous shock intensity was repeated to obtain paired shock intensity; discomfort perception data in the same patient and the step-up protocol continued until successful cardioversion was obtained or until the patient again requested discomfort relief. On the basis of the clinical situation, intravenous etomidate (0.2 mg/kg) or midazolam (0.1 mg/kg) was then given and shocks delivered until successful cardioversion was obtained (stage III). The defibrillation threshold was defined as the lowest shock energy (voltage) that converted atrial fibrillation into sinus rhythm. After each defibrillation, the voltage and the impedance of each phase of the biphasic waveform were recorded from the external defibrillator and the total delivered energy was computed.
On the basis of the data of a pilot study,12 it was determined that a sample size of 41 patients would be required to detect a 25% reduction in the VAS-P (α=0.05, β=0.20) at the same shock intensity comparing the on drug/off drug states (assuming the worst case of no correlation between the placebo and drug response). Furthermore, if the drug permitted the delivery of a shock at an intensity of 150% without significantly increasing the VAS-P, this difference would, as well, be detected with a sample size of 46 (α=0.05, β=0.20). Therefore a minimum of 46 patients were to be enrolled.
Results are expressed as mean±SD or SEM. Comparison of mean values was performed with the nonparametric Wilcoxon rank sum test for unpaired samples and with the Wilcoxon signed-rank test for paired observations. Proportions were compared with the use of the χ2 test for contingency tables. Linear regression was used to calculate the slope of each variable (VAS, PRI) on the corresponding shock number for each patient. A positive slope indicated an increase in pain or fear when more shocks with increasing intensity were administered to the patient. Individual slopes were averaged for the placebo and butorphanol groups during stages I and II of the study. All results were considered to be significant at P<0.05. Statistical calculations were performed with SAS (SAS software version 6.11) and S-PLUS (S-PLUS software version 3.1, StatSci Europe) statistical packages.
Of the 47 enrolled patients, 24 were assigned to butorphanol and 23 to placebo on the basis of initial randomization. The clinical and echocardiographic characteristics of both patient groups are reported in Table 1⇓. No significant differences were found between the groups according to sex, age, duration of atrial fibrillation as disease, duration of treated atrial fibrillation episodes, presence of structural heart disease, use of antiarrhythmic drugs, left atrial diameter, and left ventricular ejection fraction.
Outcome of Internal Atrial Defibrillation
All patients were successfully cardioverted at a mean threshold of 4.4±3.3 J (0.5 to 15) or 258±94 V (90 to 510), without complications. No significant difference was found when patients were initially randomized to butorphanol or to placebo with regard to their mean defibrillation threshold (4.3±3.4 vs 4.6±3.2 J; P=0.76). The mean number of shocks delivered to obtain sinus rhythm did also not differ between the butorphanol and placebo groups (7.6±4.0 vs 8.3±4.5 shocks; P=0.58). In the butorphanol group 183 shocks were given and in the placebo group 192 shocks. The number of patients through the various stages of the study is shown in the Figure⇑. In 24 (51%) of 47 patients, sinus rhythm was obtained without the need for intravenous sedation, with a mean energy of 2.3±1.5 J (range 0.5 to 6.9) or a mean voltage of 190±59 V (range 90 to 340) and a mean number of shocks of 4.8±2.5 (range 1 to 11). In contrast, in 23 (49%) of 47 patients requiring intravenous sedation, successful cardioversion of atrial fibrillation occurred at a mean energy of 6.6±2.9 J (range 2.6 to 15.3) or a mean voltage of 329±68 V (range 210 to 510) and a mean number of shocks of 11.1±2.4 (range 7 to 17).
There was no significant difference between the patient groups nor between successful and failed cardioversions for the average number of shocks delivered. However, for those patients undergoing both study stages, there was a significant decrease of the average number of shocks delivered going from stage I to stage II. In the placebo group, the average number of shocks decreased from 4.0 to 2.6, whereas in the butorphanol group it decreased from 3.3 to 2.5 (Table 2⇓). There was also no significant difference between the patient groups nor between successful and failed cardioversions for the mean amount of energy delivered during the first 2 study stages (Table 3⇓).
Analysis of Pain and Fear Variables During Stages I and II
Comparison of Placebo Group With Butorphanol Group for Stages I and II
All patients received a first shock of 90 V at the start of stage I. There was no evidence of a VAS-P difference between the placebo and butorphanol groups at 90 V (P=0.75), nor was there any evidence of a VAS-F difference (P=0.51), a PRI-T difference (P=0.84), a PRI-S difference (P=0.77), a PRI-A difference (P=0.34), or a PRI-E difference (P=0.99).
For patients in stage I, the slope of the regression line for each pain (except for PRI-A in patients randomized to butorphanol) and fear variable and the corresponding shock number was significantly positive, indicating that pain and fear increased with the number and the intensity of shocks received (Table 4⇓). No significant difference was found between the placebo and butorphanol groups according to the mean slope of the different pain and fear variables during stage I. Nevertheless, there was a definite trend for PRI-T, PRI-S, and PRI-A. During stage II, there was also no significant difference between both patient groups for the mean slope for VAS-F and PRI-A, whereas the mean slope for VAS-P showed a trend. The mean slopes for PRI-T, PRI-S, and PRI-E, however, became significantly lower in the butorphanol group than in the placebo group. Furthermore, during stage II, the slope for VAS-P and the 4 PRI slopes were no longer significantly different from zero in the butorphanol group. This may indicate that the second intranasal administration of butorphanol in patients randomized to butorphanol prevented a further rise in pain despite the fact that shocks with increasing intensity were delivered.
Comparison of Placebo Group With Butorphanol Group for Patients Who Received the Same Shock Intensity Ending Stage I and Going to Stage II
Eighteen of the 23 patients randomized to placebo and 17 of the 24 patients randomized to butorphanol requested (additional) butorphanol and were again defibrillated by use of the previous shock intensity before continuation of the step-up protocol. In the patients randomized to placebo, the mean VAS score for pain, PRI-T, PRI-S, and PRI-E became significantly lower after butorphanol administration, whereas no significant difference was found for the mean VAS score for fear and PRI-A. In the patients randomized to butorphanol, no significant difference was found for the pain and fear variables after a second butorphanol administration (Table 5⇓).
Intravenous Sedation (Stage III)
No side effects from butorphanol administration were observed. There was no significant difference in the number of patients of the placebo and butorphanol groups who required intravenous sedation for discomfort relief to obtain a successful cardioversion (13 patients of the placebo group and 10 of the butorphanol group; P=0.47). Intravenous etomidate was given to 12 patients of the placebo group and to 7 of the butorphanol group; the mean dose was not significantly different for both groups (15.0±5.8 vs 15.4±5.9 mg; P=0.89). One patient in the placebo group received 12 mg midazolam intravenously and 3 patients in the butorphanol group received a mean dose of 7.5±1.8 mg. The mean amount of energy at which patients requested intravenous sedation also did not differ between the placebo and the butorphanol groups (3.1±1.4 vs 2.7±1.6 J; P=0.58). The corresponding values in voltage for both patient groups were 221±54 and 205±62 V (P=0.52).
Shock-related discomfort may be due to pain and/or fear and may result from several factors such as skeletal muscle contraction, nerve stimulation, and psychological components.1 2 To date, most studies have attempted to cause less pain by lowering the atrial defibrillation threshold by use of technological improvements of the defibrillation system,13 14 shock characteristics,3 14 15 and optimal electrode localization4 16 or by administration of antiarrhythmic drugs.17 To our knowledge, the current study is the first to evaluate in a double-blind placebo-controlled manner the effect of an analgesic on pain and fear during internal atrial defibrillation. All patients of this study perceived pain and fear during internal cardioversion, and the value of both discomfort parameters increased after serial delivery of shocks with increasing intensity. As expected, fear was not affected by the opioid, but butorphanol diminished pain perception during internal defibrillation. The dose of butorphanol needed to obtain this effect depends on the time of analysis. By analysis of the effect of butorphanol on pain over time, the study reveals a trend toward an effect of 1 mg butorphanol. A total of 2 mg of the pain-relieving drug prevented a further rise in pain despite the delivery of shocks with increasing intensity. By analysis of the effect of butorphanol on pain at well-defined moments of the study, 2 observations were made. First, after the first shock there was no difference in pain perception between the placebo and butorphanol groups; however, when patients received the same shock intensity ending stage I and going to stage II, administration of butorphanol clearly decreased pain perception in the former placebo group. This confirms1 2 3 5 13 that the delivered voltage and/or energy play a role in pain perception. At the beginning of the step-up protocol, a low-energy shock did not result in pain perception. In contrast, at the moment when the patients requested additional medication for discomfort and a higher amount of energy was delivered in comparison to the beginning of the study, pain perception clearly decreased in the placebo group when butorphanol was given. Second, at this partial crossover point of the study, a redose of the analgesic did not further diminish pain in the butorphanol group. One can hypothesize that butorphanol reduces pain perception to a certain level that cannot be lowered further despite redose of the drug.
Of the 3 qualitative components of pain, only the affective component was not influenced by butorphanol at the partial crossover point nor over time. This finding is in agreement with the lack of an effect of an analgesic on fear because the affective dimension of pain is related to emotional qualities of pain perception such as stress, fear, and autonomic reactions. A further increase of the sensorial component, indicating physical pain characteristics such as pressure and temperature, and the evaluative component of pain, reflecting a cognitive measurement of pain severity, was prevented by the second dose of the analgesic despite the fact that shocks with increasing intensity were delivered.
Although previous work used the VAS4 or a similar verbal scale1 2 3 5 13 to evaluate pain during the course of a step-up atrial defibrillation protocol, this study showed that the total PRI of the 3 analyzed pain dimensions appear to be more accurate than the VAS. In a comparison of both patient groups during stage I of the study, the PRI-T revealed a trend toward an effect of 1 mg of butorphanol. Furthermore, a significantly lower PRI-T during stage II of the study protocol undoubtedly demonstrates that the second dose of the pain-relieving drug prevents a further rise in pain despite the delivery of higher-intensity shocks. In contrast, the VAS did not show a difference between placebo and butorphanol on pain relief during stage I and showed only a trend toward an effect on pain perception after 2 mg of the opioid. It is known that there exists only a moderate correlation between the VAS and MPQ despite the very different patient groups studied.10 18 In addition, VAS measures are known to be more difficult to use and understand in some patients.19
Recently, the safety and efficacy of an implantable atrial defibrillator (Metrix Atrioverter) was evaluated in 51 patients with recurrent atrial fibrillation.20 The mean atrial defibrillation threshold of the currently available model 3020 was 2.9±1.1 J at implantation and was slightly lower 3 months thereafter. In our study, approximately the same mean amount of energy (3.1±1.5 J for the placebo group and 2.7±1.6 J for the butorphanol group) was tolerated by 49% of the patients before they required intravenous sedation. The other 51% of the patients were successfully cardioverted without the need for intravenous sedation, at a mean energy of 2.3±1.5 J. The favorable effect of butorphanol, ease in intranasal administration, absence of side effects, and ability to alleviate pain at the required amount of energy needed for successful Atrioverter therapy makes the drug suitable for ambulatory use with the Atrioverter. Nevertheless, our data show that psychological components causing anxiety substantially contribute to discomfort during the delivery of sequential shocks. Until now, shock-related discomfort is only evaluated in the catheterization laboratory. Because an Atrioverter is intended for ambulatory use, it may also be important to consider the effect of surrounding environmental stimuli and patient posture on shock tolerability. The somewhat distracting effects of the surrounding stimuli may alter the patient’s perception of the shocks. Because posture may affect the relative location of the electrical field with respect to nerves and muscles, shocks delivered in a supine position may be perceived differently than shocks delivered in an upright position. Additionally, previous studies showed that tolerability of internal atrial cardioversion by use of a step-up protocol was more related to the number of shocks delivered than to the shock intensity.4 5 This suggests that patients with an Atrioverter may better tolerate the delivery of 1 or a few higher-intensity shocks rather than a greater number of lower-intensity shocks.
This study, using a step-up internal atrial defibrillation protocol, demonstrates an effect of an opioid analgesic, butorphanol, on pain perception during internal atrial defibrillation. Of the 3 qualitative components of pain, only the affective component was not influenced by butorphanol. The PRI-T evaluated pain more accurately than the VAS score. If, in patients treated with an Atrioverter, shock-related discomfort is mainly due to pain, transnasal butorphanol would provide a method to relief pain in a fast and convenient manner in the outpatient setting. Nevertheless, this study also shows that psychological components causing anxiety substantially contribute to discomfort during defibrillation. Control of the anxiety component of discomfort will require further clinical evaluation.
- Received August 26, 1998.
- Revision received December 8, 1998.
- Accepted December 29, 1998.
- Copyright © 1999 by American Heart Association
Lévy S, Ricard P, Gueunoun M, Yapo F, Trigano J, Mansouri C, Paganelli F. Low-energy cardioversion of spontaneous atrial fibrillation: immediate and long-term results. Circulation. 1997;96:253–259.
Verkes RJ, Vanderiet K, Vertommen H, van der Kloot WA, van der Meij J. De MPQ-DLV, een standaard Nederlandstalige versie van de McGill Pain Questionnaire voor Belgie en Nederland. In: van der Kloot WA, Vertommen H, eds. De MPQ-DLV, een standaard Nederlandstalige versie van de McGill Pain Questionnaire: achtergronden en handleiding. Lisse: Swets & Zeitlinger; 1989:57–69.
Vlaeyen JWS. Chronic Low Back Pain: Assessment and Treatment From a Behavioral Rehabilitation Perspective. Thesis. University of Maastricht. Lisse/Rockland: Swets & Zeitlinger; 1991.
Ayers GM, Alferness CA, Ilina M, Wagner DO, Sirokman WA, Adams JM, Griffin JC. Ventricular proarrhythmic effects of ventricular cycle length and shock strength in a sheep model of transvenous atrial defibrillation. Circulation. 1994;89:413–422.
Timmermans C, Rodriguez LM, Smeets JLRM, Ripley K, Ayers GM, Wellens HJJ. Effect of intranasal butorphanol on discomfort during internal atrial defibrillation. J Am Coll Cardiol. 1997;29:2(suppl A):474A. Abstract.
Cooper RAS, Plumb VJ, Epstein AE, Kay GN, Ideker RE. Marked reduction in internal atrial defibrillation thresholds with dual-current pathways and sequential shocks in humans. Circulation. 1998;97:2527–2535.
Cooper RAS, Johnson EE, Wharton JM. Internal atrial defibrillation in humans: improved efficacy of biphasic waveforms and the importance of phase duration. Circulation. 1997;95:1487–1496.
Melzack R, Katz J. The McGill Pain Questionnaire: appraisal and current status. In: Turk DC, Melzack R, eds. Handbook of Pain Assessment. New York, NY: Guilford Press; 1992:152–168.
Wellens HJJ, Lau CP, Lüderitz B, Akhtar M, Waldo AL, Camm AJ, Timmermans C, Tse HF, Jung W, Jordaens L, Ayers G, for the Metrix Investigators. Atrioverter: an implantable device for the treatment of atrial fibrillation. Circulation. 1998;98:1651–1656.