Continuous perineural infusion of a long-acting local anesthetic for the relief of pain after an amputation of the lower extremity is becoming more popular3,20,21. This method could be useful for the relief of perioperative pain in geriatric patients, who have a limited tolerance for narcotic analgesics. Although continuous epidural infusion of an anesthetic is effective for the relief of postoperative pain, it necessitates the use of an indwelling urinary catheter because of the associated depression of the baroreceptors of the urinary bladder; therefore, it exposes the patient to the potential morbidity of urinary retention and infection.
The purpose of the present study was to determine whether continuous infusion of a regional anesthetic was effective for the relief of postoperative pain and the prevention of late residual or phantom-limb pain after an amputation of the lower extremity because of peripheral vascular disease.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Read in part at the Specialty Day Meeting of the Orthopaedic Rehabilitation Association, Orlando, Florida, February 19, 1995.
‡Departments of Orthopaedic Surgery (M. S. P., T. P., and L. V.) and Anesthesiology (P. G. N. G.), Loyola University Medical Center, 2160 South First Avenue, Maywood, Illinois 60153.
The present study was approved by the Human Studies Committee at Loyola University Medical Center. Twenty-one consecutive patients who were to have an amputation of the lower extremity because of ischemic necrosis secondary to peripheral vascular disease were divided into two groups with use of a table of random numbers. The treatment group (Group A) included nine patients who were to have a transtibial amputation, one patient who was to have a disarticulation at the knee, and one patient who was to have a transfemoral amputation. The control group (Group B) included seven patients, two patients, and one patient, respectively. The patients included ten men and eleven women who were forty-one to ninety-one years old (mean and standard deviation, 68.3 ± 12.96 years old). Five men and four women had diabetes. This cohort of twenty-one patients had been selected from a group of seventy-five patients who had an amputation of the lower extremity at our institution between January 1993 and October 1994. Patients were excluded from the study if they did not appear to have the cognitive capacity necessary to use patient-controlled analgesia effectively or to complete the McGill Pain Questionnaire.
The level of amputation was chosen on the basis of the clinical findings, the results of transcutaneous oximetry (as a measure of vascular inflow), the level of serum albumin (as a measure of nutritional status), and the total lymphocyte count (as a measure of immunocompetence). Use of these clinical parameters15-17 has increased the rate of wound-healing after amputations for ischemic necrosis in the foot and ankle from 43 per cent (ten of twenty-three4) to 92 per cent (twenty-four of twenty-six17). The operation was performed with use of spinal anesthesia (seventeen patients) or general anesthesia (four patients). No patient was managed with administration of analgesics by intrathecal injection or by use of an indwelling spinal or epidural catheter for the relief of postoperative pain.
Technique of Continuous Infusion of the Anesthetic
The sciatic nerve (in patients who had had a transfemoral amputation) or the posterior tibial nerve (in patients who had had a transtibial amputation or a disarticulation at the knee) was dissected free proximally. Light traction was applied to the nerve end, the nerve was transected as far proximally as possible, and the remaining portion of the nerve was allowed to retract proximally into the muscle mass of the residual limb. Care was taken to avoid crushing even the excised section of nerve, and ligatures and cautery were not used. A 19-gauge Teflon epidural catheter was placed adjacent to the transected nerve and was brought to the surface of the skin through a 16-gauge needle, which was then removed. The catheter was sutured to the skin and then was attached to a numbered bag that contained 250 milliliters of either 0.5 per cent bupivacaine hydrochloride or normal saline solution. The bag was connected to a portable infusion pump. The patients and the staff were blinded to the contents of the bag, which were known only to the research pharmacist.
Postoperatively, a bolus of ten milliliters of solution was given to ensure patency of the catheter. The patient then received a continuous infusion of 0.5 per cent bupivacaine hydrochloride (Group A) or normal saline solution (Group B) at a rate of one milliliter per hour. Patient-controlled analgesia with morphine sulfate, administered intravenously with use of a pump at a concentration of five milligrams per milliliter, also was available during this period1. A maximum of thirty milligrams of morphine could be used every four hours; the lock-out period, during which morphine could not be obtained, was set at ten minutes. No additional analgesia was used. The amount of morphine that each patient used was recorded meticulously. The catheter was removed and the patient-controlled analgesia was discontinued seventy-two hours after the operation.
McGill Pain Questionnaire
The McGill Pain Questionnaire9,11-13 (Fig. 1), designed to provide a quantitative assessment of pain and to be used as a diagnostic tool, was administered to evaluate late residual and phantom-limb pain at three and six months after the operation. It was not administered before the operation or in the early postoperative period because we were not attempting to compare late pain with ischemic preoperative pain or postoperative pain, and it was not used after six months because pain had stabilized by that time18.
The questionnaire required the patient to select subjective descriptors of pain from a list that was divided into four categories: sensory descriptors, such as throbbing, stabbing, sharp, cramping, burning, aching, pounding, pulling, tender, shooting, gnawing, heavy, and boring; affective descriptors, such as tiring, exhausting, and cruel; evaluative descriptors, such as annoying, intense, and unbearable; and temporal descriptors, such as constant, periodic, and brief. The chosen descriptors then were ranked in terms of intensity according to a scale in which 0 indicated no pain; 1, mild pain; 2, discomforting pain; 3, distressing pain; 4, horrible pain; and 5, excruciating pain.
Statistical Analysis
The primary purpose of the present study was to determine whether treatment with bupivacaine hydrochloride provided greater comfort than did treatment with a placebo. The null hypothesis suggested that more morphine would be used in conjunction with bupivacaine hydrochloride than in conjunction with the placebo. The data were analyzed descriptively to note measures of central tendency and dispersion; missing values and any known outliers were identified. The study met the assumptions of homogeneity. Repeated-measures analysis of variance was performed on the test data with a prespecified alpha coefficient of 0.05. The sample available for this pilot study was small; the experimental effect size was calculated with use of the method of Cohen, in which f = (k - 1)F/N. Therefore, f = (1)17.82/21 = 0.9211 (large effect size). By interpolation, power was estimated at 0.89 for this particular sample and was considered adequate.
Administration of the McGill Pain Questionnaire at the three and six-month follow-up evaluations necessitated an examination of test-retest reliability. Melzack10 reported a mean consistency of response of 70.3 per cent for this test on the basis of the results of test-retest studies. We found that after two administrations of the test, the reliability estimates were moderately positive at 60 per cent.
The patients in Group A used significantly less morphine during the first and second days after the operation than did those in Group B (F = 17.82, p < 0.001). There was no difference between the groups with regard to the amount of morphine used on the third postoperative day (F = 4.16, p = 0.01) (Table I). There was no significant difference between the groups with respect to the four-hour periods of measurement (F = 1.75, p = 0.14). There were no known complications related to the use of the catheters or the patient-controlled analgesia, and there were no wound infections.
Fourteen patients (five men and nine women) completed the McGill Pain Questionnaire three and six months after the operation, five patients chose not to complete the questionnaire, and two patients died before the questionnaire was administered. In the section that included sensory descriptors, twelve patients indicated a decrease of pain and two noted an increase of pain; in the section that included affective descriptors, seven patients indicated a decrease in pain, five did not choose any descriptors, one noted that the pain had not changed, and one noted an increase in pain; in the section that included evaluative descriptors, three patients indicated a slight decrease in pain, two did not choose any descriptors, five noted that the pain had not changed, and four noted an increase in pain; and in the section that included temporal descriptors, ten patients indicated a decrease in pain, one noted that the pain had not changed, one did not choose any descriptors, and two indicated an increase in pain.
Total scores from the rankings of the subjective descriptors showed that eleven patients had an over-all decrease in pain, two had a marginal increase in pain, and one had a substantial increase in pain between the three and six-month evaluations. Evaluation of the over-all intensity of pain demonstrated that nine patients had no change in intensity, four had a decrease in intensity, and one had an increase in intensity between the three and six-month evaluations. Over-all, there was no significant difference between Groups A and B with regard to residual or phantom-limb pain (p = 0.34) and there was no significant change in subjective pain between the three and six-month evaluations (p = 0.11).