Bupivacaine Versus Ropivacaine on Diaphragmatic Motility and Ventilatory Function

Overview

Background: Outpatient continuous interscalene brachial plexus blocks are used to control pain after shoulder surgery, with infusions of 0.125% bupivacaine or 0.2% ropivacaine. There have been no studies comparing the effects of these two formulations. The major concern is hemidiaphragmatic paresis, and since ropivacaine preferentially blocks sensory fibers, it may cause less blockade of the phrenic nerve. This study was to evaluate the effects of continuous interscalene brachial plexus infusions, with the hypothesis that respiratory function is more affected by 0.125% bupivacaine than 0.2% ropivacaine, with equal effects on pain relief. Methods: All patients underwent baseline spirometry and ultrasound evaluation of diaphragmatic excursion, followed by interscalene catheter placement for their surgery, then randomized to receive a pump containing 0.2% ropivacaine or 0.125% bupivacaine. Patients returned to the hospital the following day for spirometry, ultrasound reevaluation, and evaluation of their pain level since discharge.

INTRODUCTION Shoulder surgery is associated with severe postoperative pain and continuous interscalene brachial plexus blocks can control this pain. Compared to single-shot injections, continuous interscalene blocks can increase patient satisfaction, decrease narcotic requirements, which then decreases postoperative nausea and vomiting , improve quality of sleep and decrease hospital length of stay . The concept of continuous peripheral nerve blocks at home (CPNBH) with disposable pumps began in 1998 in Sweden. The authors chose 0.125% bupivacaine for interscalene catheters and their choice was not based on any evidence in the literature, but rather "because this concentration was already available at the institution for epidural infusions" (personal communication, Dr. Narinder Rawal, January 2013). Two years later, as ropivacaine became more widely available, others chose to evaluate the safety and efficacy of 0.2% ropivacaine for continuous interscalene infusions. Again, the choice of this specific concentration was not evidence-based, but rather "because the company made it in this concentration" (personal communication, Dr. Stephen Klein, January 2013). To date, published reviews show that most studies use either of those formulations, yet there have been no studies comparing the effects of these two drugs and concentrations, particularly on pain relief and on major side effects of CPNBH. One of those major side effects is ipsilateral phrenic nerve paresis. This is because blockade of the brachial plexus at the interscalene groove with a single large-volume injection causes diaphragmatic hemiparesis in 100% of patients. This incidence is decreased to 20% with a continuous infusion of a dilute solution11, where 9% of patients have a subjective feeling of shortness of breath. An ideal infusate in CPNBH would control postoperative pain and have minimal effects on the phrenic nerve. Ropivacaine has been shown to have blockade properties that differ from bupivacaine; specifically, ropivacaine seems to preferentially block sensory nerve fibers, while weakly blocking motor fibers. It is not known whether this translates into a more attenuated block of the phrenic nerve with ropivacaine. Therefore, the purpose of this study was to evaluate the effects of continuous interscalene peripheral nerve blocks with 0.125% bupivacaine versus 0.2% ropivacaine, comparing their effect on pain relief and respiratory function. The primary hypothesis is that respiratory function is more affected by bupivacaine than ropivacaine, with equal effects on pain relief. METHODS Patients who were scheduled for moderately to severely painful unilateral shoulder surgery who agreed to continuous interscalene catheter placement were enrolled in the study and gave written consent. All patients underwent baseline studies, which included ultrasonic evaluation of diaphragmatic excursion and bedside spirometry. All were placed in a 45-degree upright position on a hospital stretcher. Assessment of the range of motion of both hemidiaphragms was performed with a curvilinear low frequency ultrasound probe The probe was placed subcostaly in the mid-axillary line, angled 45 degrees cephalad on both sides, using the liver on the right and the spleen on the left as acoustic windows. If there was difficulty in obtaining a satisfactory image (frequently the case on the left side), the probe was moved laterally toward the anterior axillary line until the diaphragm was visualized. Range of motion of the diaphragm with normal respiration was assessed visually to exclude any pre-existing diaphragmatic motion abnormalities, followed by sigh and sniff tests. During the 'sniff test', movement of the diaphragm was evaluated from the resting expiratory position during quick inspirations of air taken through the nose. During the 'sigh test', the range of diaphragmatic motion from the resting expiratory position to deep inspiration was measured. The normal caudad movement of the diaphragm was recorded with the distance traveled in centimeters. Each measurement was performed 3 times and the best value was recorded. A mark was placed on the patients' skin where measurements were taken, to ascertain that the diaphragm was assessed from the same site during the follow-up visit. Bedside spirometry was performed for each patient, again in a 45-degree upright position in a hospital stretcher. After instructing the patient, forced expiratory volume at 1 second (FEV1), forced vital capacity (FVC) and peak expiratory flow (PEF) were measured three times, and the best measurement was recorded. Following this, with standard noninvasive monitoring and oxygen via nasal cannula, all patients had interscalene catheter placement with midazolam sedation (up to 5 mg). All catheters were placed using and nerve. After draping the area and sheathing the ultrasound probe (SiteRite; Bard Access Systems, Salt Lake City, UT), a local anesthetic skin wheal was placed and a 17-gauge Tuohy insulated needle was inserted in-plane with a posterior approach, through the middle scalene muscle. The prevertebral fascial plane was punctured and the needle tip was placed in the area of connective tissue between the C5 and C6 nerve roots. A 19-gauge stimulating catheter was then inserted through the needle and attached to the nerve stimulator, set at a current of 0.48 mA, 100 microsecond pulse duration and 2 Hz frequency. The nerve stimulator settings were kept constant throughout the procedure. With ultrasound visualization and muscle twitches of the upper arm (biceps, triceps or deltoid), the catheter was threaded 5-6 cm beyond the needle tip. If stimulation was lost during threading, the catheter was gently withdrawn, the needle was moved slightly and the catheter was threaded again with ultrasound visualization and visible muscle twitching. If this was not possible after five minutes (from the first catheter threading), the patient was removed from the study. Once the catheter was threaded successfully, the needle and catheter stylet were removed and the catheter was withdrawn until the skin marking was 2-3 cm beyond the needle depth. The nerve stimulator was reattached to the catheter to verify the presence of a muscle twitch at 0.48mA, and a 5 ml volume of 1.5% lidocaine with epinephrine 1:200,000 was injected under direct ultrasound vision, confirming local anesthetic spread. Prior to transport into the operating room and after negative aspiration, all catheters were injected with 30 ml plain mepivacaine 1.5%, followed by an assessment of the quality of the sensory and motor block. A general anesthetic was administered with propofol induction and maintenance with vecuronium and sevoflurane. No opioids or additional benzodiazepines were given. Three hours after the initial injection, whether patients were in the operating room or in the Post Anesthesia Care Unit (PACU), an additional 10 ml plain mepivacaine 1.5% was given through the catheter. Upon arrival to PACU, pain control was assessed (on a numerical scale from 0 to 10), catheter and dressing integrity were verified and patients were randomized by the pharmacists, using a computer-generated randomization table, to receive a disposable pump (On-Q CB004, I Flow Corporation, Lake Forrest, CA) containing 400 ml of 0.2% ropivacaine or 0.125% bupivacaine and set at a rate of 6 ml/hr. Except for the pharmacist, all other personnel were blinded to the content of the pumps. The following day (20-28 hours after discharge), patients returned to the hospital and were asked about their highest pain level (on a numerical scale from 0 to 10) since surgery and the number of pain pills taken since discharge, difficulty breathing, the amount of numbness in the hand, and any problems with the catheter or the pump. The catheter sites were examined and when necessary, the dressing was reinforced. After this, patients underwent pulmonary testing, again on a stretcher in a 45-degree upright position. Repeat ultrasonographic evaluation of diaphragmatic excursion (sniff and sigh tests) and bedside spirometry were performed. For the ultrasonographic evaluation, the previously marked sites were used for 3 measurements, recording the best effort. If on the sigh and/or sniff test a caudad movement of the hemidiaphragm was seen, the distance was measured, recorded and assigned a positive value (+). If paradoxical cephalad movement of the diaphragm was observed, the distance was given a negative value (-). Following completion of the ultrasonographic evaluation, spirometry was performed. Again FEV1, FVC and PEF were measured and the best of 3 attempts was recorded. Patients were sent home with CPNBH for an additional two days. Per routine, an anesthesiologist called them every day to assess their pain level, shortness of breath, numbness, and integrity of the catheter site. Statistical Analysis Determining the study size was allowing for a 15% difference in forced expiratory volume at 1 second (FEV1) between the two. A non-inferiority power analysis was chosen to obtain an alpha error probability of 0.05 and a 1 minus beta error probability of 0.8. A sample size of 30 patients (15 per group) was chosen to give the study sufficient power. A T-test (unmatched pairs) was used to compare continuous data between the two groups. If data had a non-normal distribution, a non-parametric Kruskal-Wallis ANOVA was used to examine differences between means. Categorical data was expressed as mean + standard error. Statistical significance was set at a p-value ≤ 0.05, two-tailed.