Following institutional review board approval (Peking University Third Hospital Ethics Committee) and trial registration (ChiCTR-TRC-13004058), patients scheduled for elective arthroscopic shoulder surgery were recruited in this prospective, randomized, controlled, double blind clinical trial. Written informed consents were obtained from all patients.
The inclusion criteria were age between 18 and 80 years and an American Society of Anesthesiologists physical status of I and II. The exclusion criteria were patient refusal, neurologic or neuromuscular disease, severe bronchopulmonary disease, coagulation disorders, infection at the injection site, known allergy to one or more medications used in the study protocol, body mass index over 32 kg/m2 and patients treated with opioid analgesics.
Using a computer generated random number table, patients were randomized into three groups: 6.7 ml of 0.75 % ropivacaine (Group 0.75), 10 ml of 0.5 % ropivacaine (Group 0. 5) and 20 ml of 0.25 % ropivacaine (Group 0.25). Sealed opaque envelopes with the study group allocation were opened before the blocks were performed. An independent anesthesiologist who was not involved in the study prepared the study solution.
Routine monitors (pulse oximeter, noninvasive blood pressure cuff, and electrocardiogram) were used, and intravenous access was established. Patients were premedicated intravenously with midazolam 2 mg and dexamethasone 10 mg.
After skin disinfection and infiltration with lidocaine 1 %, the scalene muscles and interscalene brachial plexus were imaged in the short axis using a 38-mm 13–6 MHz linear array ultrasound probe (Turbo SonoSite HFL, Bothell, WA), the transverse process of vertebra C7 was identified by the absence of the anterior tubercle, subsequently, root C7 of the brachial plexus was visualized in short-axis view, then move the probe to detect root C6 and C5. After clear identification of root C6 and C5, then the area posterior to the plexus was analysed for the presence of the dorsal scapular and long thoracic nerves located within or around the middle scalene muscle as a discrete hyperechoic structure containing hypoechoic fascicles within them. Then a short bevel 100-mm, 19-gauge stimulating needle (StimuLong Sono Tsui set, Pajunk, Geisingen, Germany) connected to a Stimuplex (Stimuplex-HNS II A; B. Braun Melsungen AG, Germany) nerve stimulator was inserted using an in-plane technique from posterior to the probe, following a shallower needle trajectory to avoid entering the bulk of middle scalene muscle. If the dorsal scapular and long thoracic nerves had been visualized within the middle scalene, avoid injuring them during the needle passage. But if they were hard to identify by ultrasound, during inserting the stimulating needle, the contractions of the serratus anterior, rhomboids, and levator scapulae could help identify the dorsal scapular (rhomboids and levator scapulae muscle) or long thoracic (serratus anterior muscle), and once these muscle twitches were observed, stop and change the needle passage to avoid nerve injury. Elicitation of a sustained deltoid motor response at 0.3–0.5 mA with the tip of the needle positioned just lateral to the C5-6 roots confirmed the correct placement of the needle. All injections were administered slowly with repeated aspiration to prevent or detect early intravascular injection. The blocks were performed by an experienced physician who specialized in ultrasound and nerve stimulation–assisted interscalene block. After completion of the block, all the evaluations were performed by another blinded independent observer who was not involved in grouping, study solution preparing and interscalene block performing. All patients were blind to their group throughout the study.
Sensory and motor blockades were assessed every 5 min for up to 30 min. Sensory blockade was assessed by using a pinprick in the C5-6 dermatome (lateral region of the arm, innervated by the superior cutaneous nerve of the arm, branch of the axillary nerve) on a 3-point verbal rating scale: 0, normal sensation; 1, dull sensation (analgesia); and 2, no sensation (anesthesia). The onset time for a sensory block was defined as the time elapsed between the end of the block procedure and the moment when the pinprick test yielded a score of 2. Failure to reach a score of 2 within 30 min of interscalene block was considered to be block failure. Motor blockade was determined by loss of shoulder abduction (deltoid sign) and was objectively assessed on a 3-point rating scale: 0, normal movement; 1, diminished movement; and 2, no movement.
Complications, such as hematoma, Horner’s syndrome, hoarseness, respiratory distress and a decrease in SpO2 of more than 5 %, were also assessed during this period.
Before and 30 min after the block, an independent observer evaluated ipsilateral hemidiaphragmatic movement on deep inspiration by ultrasonography using a 5–2 MHz broadband curved array transducer (SonoSite, Inc., Bothell, Wash). Patients were examined in the upright sitting position and scanned from a low intercostal or subcostal approach using the liver or spleen as an acoustic window [3]. Normal caudad movement with deep inspiration was considered to be no paresis, whereas no movement or paradoxical cephalad movement was designated as paresis.
General anesthesia was induced with propofol (target controlled infusion 3–5 μg/ml), fentanyl (1.5 ~ 2.5 μg/kg) and cisatracurium (0.15 mg/kg) and maintained with propofol (target controlled infusion 3–5 μg/ml) and remifentanil (continuous infusion 0.05–0.3 μg/kg/min). Patients were intubated and ventilated with an oxygen/air admixture. Anesthesia management was left to the discretion of the attending anesthesiologist. No other analgesics were administered intraoperatively.
All patients received parecoxib 40 mg bid for the first two days after surgery. Tramadol 100 mg or pethidine 50 mg was given as rescue analgesics for numerical rating pain scale (NRPS) ≥4 or by the patient’s request.
Patients were instructed to rate their pain using an 11-point NRPS ranging from 0 to 10 (0, no pain; 10, worst imaginable pain). NRPS of rest pain was measured before the block, right before discharge from the PACU and at 4, 8, 24 h after the block. The timing and dosage of analgesics were recorded. Twenty-four hours after the block, patients were questioned for NRPS of the worst pain, time of the first shoulder pain, sleep quality (0 = no sleep disturbance because of pain, 1 = sleep disturbance because of pain) and satisfaction with the interscalane block (0–10, 0 = very unsatisfied; 10 = very satisfied).
The blinded investigator recorded grip strength in the operative limb using the same Jamar dynamometer before the administration of intravenous sedation, right before discharge from the PACU and at 4, 8 and 24 h after the block.
The patient was telephone interviewed if they suffered a late complication such as scapula alata, nerve injury and pain along the medial border of scapula or radiating to the arm and forearm related to interscalene block after more than one year since discharge from hospital.
The primary outcome was onset time. Our research hypothesis was that compared with low concentration/high volume group, high concentration/low volume group would result in a shorter onset time. According to our preliminary test, the onset time for 10 mL of 0.5 % ropivacaine is 10 ± 5 min. We considered a time difference of 50 % as being clinically relevant. To obtain a 2-tailed α error of 0.05 and a statistical power of 0.9, the calculated sample size was 27 patients per group. We recruited 33 patients per group in anticipation of 20 % potential dropouts.
Statistical analysis
The SPSS version 20.0 (SPSS, Inc., Chicago, USA) was used. The normality of the continuous data was first assessed with the Kolmogorov-Smirnov test. Normally distributed variables were expressed as the mean ± SD and compared between groups using one way analysis of variance (ANOVA) with an LSD post hoc test. Non-normally distributed variables were expressed as the median (range) and analysed using the Kruskal-Wallis H test. Categorical variables were expressed as the number (%) and compared between groups using a Pearson test. Kaplan-Meier survival curves were constructed (survival was equivalent to unsuccessful block). The log-rank test was used to detect differences in block onset times among groups. P < 0.05 was considered statistically significant.