Effect of Patient-controlled Perineural Analgesia on Rehabilitation and Pain after Ambulatory Orthopedic Surgery: A Multicenter Randomized Trial

After obtaining approval from the Institutional Review Board of Lapeyronie University Hospital Center (Montpellier, France) and written informed consent, we prospectively enrolled 85 adult patients scheduled to undergo ambulatory, unilateral, acromioplasty, or hallux valgus surgery, who desired peripheral nerve block for the perioperative period, in two university hospitals and one private institution. Inclusion criteria were age 18 yr or older; American Society of Anesthesiologists physical status I, II, or III; understanding of the possible complications related to local anesthetic, the study protocol, and care of the catheter and elastomeric pump; and a nurse from a home care service managing the patient twice a day and capable of removing the catheter in the evening of postoperative day 3. Patients who did not cooperate and those who had psychological disorders or linguistic difficulties that might interfere with pain assessments were excluded. Medical exclusion criteria were severe bronchopulmonary disease, blood clotting impairment, hepatic or renal insufficiency, history of recent local or systemic infection, known allergy to the trial drugs, any previous damage to the nerve, plexus neuropathy or neuraxis disease, and cardiac conduction problems (second- or third-degree atrioventricular block). In addition, patients who refused, patients who had participated less than 1 month previously in a therapeutic trial, and those who were already participating in another study were not included in the study.

Patients were premedicated with 0.1 mg/kg oral midazolam 1 h before surgery. They had a peripheral intravenous catheter inserted and were placed in the prone position for popliteal nerve block and supine for interscalene block. Preoperatively, patients were monitored during peripheral nerve blocks according to the standard guidelines published by the French Society of Anesthesiology and Critical Care Medicine. Oxygen (6 l/min) was administered via  a facemask. Experienced anesthetists performed all of the blocks. The following procedures were standardized: a conventional aseptic procedure was used to insert interscalene block for those undergoing acromioplasty and popliteal block catheters for those undergoing hallux valgus surgery; the anesthesiologist wore a mask, cap, sterile gown, and gloves. The puncture site was prepared with 10% povidone iodine solution, and surrounding areas were covered with sterile drapes. The continuous interscalene and popliteal blocks were performed in all patients using nerve stimulators (Stimuplex® HNS 11; B. Braun Melsungen AG, Melsungen, Germany) according to the modified lateral technique for interscalene block19and the Singelyn technique for popliteal block20before perioperative sedation or induction of general anesthesia. Insulated short-beveled needles (Plexolong®; Pajunk, Geisingen, Germany, or Contiplex®; B. Braun Melsungen AG) were connected to the nerve stimulator. The placement of the needle was considered successful when a specific muscle contraction (triceps muscle or plantar flexion) was obtained with a current output of less than 0.5 mA (frequency 1 Hz and impulse duration 100 μs). The nonstimulating 20-gauge catheter was then inserted through the needle for 3 cm (interscalene block) to 5 cm (popliteal block). The catheter was subcutaneously tunneled for 4 cm through the insulated needle and secured with adhesive strips and a transparent adhesive dressing (Opsite®; Smith and Nephew Medical Ltd., Hull, United Kingdom). If the catheter could not be placed after three attempts, the patient was withdrawn from the study. After tests for endovascular placement (aspiration; injection of 2 ml lidocaine, 2%, with 1:200,000 epinephrine), anesthesia was induced via  the catheter with 30 ml ropivacaine, 0.5%, with gentle aspiration between divided doses. A block was considered successful when sensory (inability to recognize cold temperatures with an ether-soaked cotton swab on the tip of the first and third finger [interscalene block] or dorsal and plantar parts of the foot [popliteal block]) and motor block (inability to extend the arm [interscalene block] involving the radial and median nerve or plantar and dorsal flexion abolished [popliteal block]) were present 30 min after injection of ropivacaine. Specific nerve distribution of sensory blockade was evaluated. The block was considered to have failed when sensory and motor blocks were not noted or when patients needed general anesthesia because of pain during surgery. Intraoperatively, patients received a 10- to 30-μg · kg^–1· min^–1propofol infusion titrated to the patient's desired level of sedation. If preferred by the patient, higher doses of propofol were used to administer general anesthesia, and mechanical ventilation was applied via  a Proseal® laryngeal mask (Laryngeal Mask Company, Henley-on-Thames, United Kingdom) with a 1:1 mixture of nitrous oxide and oxygen. Durations of surgery and tourniquet were noted. After surgery, patients were admitted to the postanesthesia care unit after a wound dressing and an arm or leg splint had been applied. All patients received 4 mg intravenous ondansetron as standard antiemetic prophylaxis.

After successful placement of the block and catheter, patients were randomly assigned to one of three groups using a computer-generated table (fig. 1). Ten minutes after arriving at the postanesthesia care unit, patients received either patient-controlled intravenous morphine (PCA morphine group, 0.5 mg/ml, bolus 2 ml, lockout period 12 min, no basal infusion) via  250-ml-reservoir disposable pumps with PCA chambers (PCA infusor®; Baxter, Maurepas, France) or 0.2% ropivacaine in one of the two dosing regimens: 7-ml/h continuous infusion without bolus (continuous infusion group) or 5-ml/h basal infusion and a patient-controlled bolus dose of 2 ml available every 12 min (basal-bolus group) via  the same type of disposable pump (Multirate infusor®; Baxter). In the PCA morphine group, the perineural catheter was removed in the postanesthesia care unit.

The patients were transferred to the surgical ward for the first postoperative night. A nurse or an anesthesiologist blinded to the trial made a follow-up visit on the morning of the first postoperative day, and the patients were discharged home with the disposable pump if they met discharge-to-home criteria (satisfactory pain control and the ability to mobilize safely with or without assistance devices as assessed by a physical therapist).

The patient and nurse from the home care service were given standard postoperative outpatient instructions and written instructions on the use of the elastomeric pump. The nurse managed the patient twice a day at home and was instructed to fill and change the elastomeric pump with a new one if necessary. The patient was given a form to record the worst visual analog scale (VAS) pain scores (ranging from 0 mm for no pain to 100 mm for the worst imaginable pain) recorded 10 min after arrival in the postanesthesia care unit (H0); 1, 4, and 12 hours after H0; and every morning during physiotherapy or walking over 72 h. Throughout the study period, patients were directed to take 1 g paracetamol orally four times a day. The use of rescue analgesia was standardized. If the VAS score remained more than 30 mm, 100 mg ketoprofen was taken orally, and then a new VAS value was reported on the form 1 h after. The maximal dose was 300 mg daily. Consumption of morphine, ropivacaine, and rescue analgesics, as well as the number of boluses used, were noted at the end of the study period. Patients and nurses could contact a physician at any time during the study period by telephone. Follow-up telephone evaluations were performed by a blinded observer (S.B.) at 24, 48, and 72 h after surgery to determine the number of doses of oral analgesics consumed, the occurrence of any side effects (e.g. , local anesthetic toxicity signs [nurse], dizziness, weakness, paresthesia, nausea and vomiting, pruritus, urinary retention, mechanical problems with the catheter, ileus, sleep disturbance, sedation, acute respiratory failure, local vein inflammation, fever). According to the type or the patient's impact of the side effect, the observer informed the anesthetist responsible for the patient and possibly asked for a hospital admission. If complete anesthesia of the patient's surgical extremity appeared at any time, the nurse was instructed to clamp the disposable pump until the patient regained feeling and then restart the infusion. The catheter was removed by the nurse in the evening of the third day. The patients and physicians graded their satisfaction with the technique at the end of the study period (very satisfied, satisfied, mild satisfied, not satisfied).

Every morning, a physical therapist visited the patients. The physical therapist initiated early rehabilitation in accordance with the objectives of the surgical team on the day after surgery and every morning at home during the study period. In hallux valgus surgery patients, he applied passive plantar and dorsal flexion movements and encouraged the patients to walk as long as possible at home. The patients wore a Barouk shoe. This heel weight-bearing shoe with total relief of forefoot pressure enhanced early mobilization. In patients scheduled to undergo acromioplasty, a sling was applied in the operating room. Every morning, the physical therapist applied passive pendular exercises: flexion below 60°, abduction below 30°, and circumduction. He supervised the patient in doing slight isometric exercises: external rotation, internal rotation, and extension.

For all patients, the time when a patient was able to walk for 10 min, free of any symptoms or adverse events and without any aid devices or human help, was noted by the physical therapist in the hospital or at home. The patients reported their daily activities in their diary on the evening of days 1, 2, and 3: walk in the house, professional occupation, domestic work in the kitchen, took a shower, mental activity, occupied with the children; and graded it: no activity, activity with assistance, free activity. They also reported the reasons for the absence of activity or the need for assistance: pain, fatigue, nausea/vomiting, numbness, paresthesia, dizziness, fear of falling, other cause of activity limitation. The diary was returned to the surgical center in an envelope via  the French Postal Service.

Statistical analysis was performed using SAS software version 8.2 (SAS Institute, Cary, NC) in the Medical Computer Programming Department of the University Hospital of Montpellier, France.

Sample size calculations were centered around our two hypotheses that a basal infusion of ropivacaine via  a perineural catheter combined with patient-controlled bolus dose on one hand optimizes postoperative functional recovery and on the other hand decreases postoperative pain compared with patient-controlled intravenous morphine. We considered a 50% reduction in the time when a patient was able to walk for 10 min (mean time decreases from 40 h to 20 h) and in pain scores (mean pain score decreases from 30 mm to 15 mm on a scale of 0–100) to be clinically relevant. Based on an SD of 15 for the 10-min walking test and of 20 for the pain score values and assuming a two-sided type I error protection of 0.05 and a power of 0.80, 21.4 patients in each group were required to reveal a clinically significant difference for the 10-min walking test, and 30 patients were required for pain scores values. We chose the highest value to demonstrate statistically significant data on functional recovery and pain management. Postoperative rehabilitation data on day 1 were chosen as the primary criteria for analysis. The normality of distribution was determined using the Shapiro-Wilk test.

Parametric data are reported as mean ± SD. Nonparametric data are graphically reported as median with 25th-75th and 10th-90th percentile whiskers or textually noted using the median (25th-75th centiles). For normally distributed data, multiple comparisons were made using one-way analysis of variance. For nonparametric data, the Kruskal-Wallis test was used when appropriate.

Categorical data were analyzed using the chi-square test or Fisher exact test when appropriate. When a significant difference appeared, a two-by-two comparison was performed and a Bonferroni correction was applied. P < 0.05 was considered significant.

Eighty-five patients were approached for inclusion in the study. Of the 85 patients, 2 were eliminated from the data analysis because of obstruction of a popliteal catheter in a continuous infusion group patient and dislodgement of an interscalene catheter in a basal-bolus group patient, both in the operating room. Fifteen interscalene and 15 popliteal nerve blocks were included in each of the two perineural analgesia groups of patients. Ten acromioplasties and 13 hallux valgus surgeries were included in the PCA morphine group (fig. 1). There were no statistically significant differences between the groups in demographics or duration of surgery or tourniquet (table 1). All patients studied had complete motor and sensory blockades before surgery. Eight patients had supplemental general anesthesia during surgery in each group (patients' choice). The 22nd patient of the PCA morphine group was readmitted to hospital on day 2 for an acute respiratory depression because he received half of the entire content of the elastomeric pump during the second night after surgery. A mechanical problem with the PCA chamber was responsible for this adverse event. This patient received 50 mg intravenous morphine in 12 h. Although the study was originally designed to include 30 patients per group for VAS pain scores and 21.4 for functional recovery exercises, an interim analysis indicated a highly significant increase in functional exercise capacity and reduction in VAS pain scores and opioid-related side effects of the patients in both continuous perineural analgesia groups. Consequently, in compliance with the recommendations of our institutional review board, the study was aborted after the 23rd patient was tested in the PCA morphine group. The β risk remained at 20% after the reduction in our patient population. Six patients in the PCA morphine group, 2 patients in the continuous infusion group, and 1 patient in the basal-bolus group had episodes of breakthrough pain and were discharged from the hospital on day 2 (P < 0.05).

The time when a patient was able to walk for 10 min was significantly higher in the PCA morphine group: 40.5 (16–44) h, 20.5 (17–42) h, and 12.5 (4.5–20) h, respectively, for the PCA morphine, continuous infusion, and basal-bolus groups in median (25th-75th centile) values. The basal-bolus infusion of ropivacaine significantly optimized the patient's daily activities at home (fig. 2). One hundred percent of the basal-bolus group patients had complete free activity on day 3. The reasons for no activity or need for assistance during home activity are reported in table 2. The reasons were quite different in the PCA morphine group compared with both perineural analgesia groups, in particular regarding pain and fatigue.

During the 72-h infusion period, both groups of patients receiving ropivacaine experienced significantly less postoperative pain during movement compared with patients receiving PCA intravenous morphine (fig. 3). The total amount of ketoprofen (500 ± 100, 200 ± 100, and 100 ± 100 mg in the PCA morphine, continuous infusion, and basal-bolus groups, respectively) was significantly increased in the PCA morphine group. The basal-bolus group had significantly decreased consumption of ropivacaine (377 ± 22 ml or 754 ± 44 mg vs.  488 ± 14 ml or 976 ± 28 mg). Nausea/vomiting, local vein inflammation, sleep disturbance, mechanical problems with the catheter, and dizziness significantly increased in the PCA morphine group (table 3). Patients experienced significantly more paresthesia in the continuous infusion group (table 3) in comparison with the basal-bolus group. No infection, local anesthetic toxicity, dyspnea (interscalene block), or disposable elastomeric pump problems were noted in the perineural infusion groups.

There were significant differences between the PCA morphine group and both perineural analgesia groups with regard to satisfaction with analgesia techniques for patients and physicians. The basal-bolus group had a higher number of satisfied patients and physicians (table 4).

This multicenter randomized clinical trial demonstrates that, in comparison with PCA intravenous morphine, a perineural infusion of ropivacaine at home via  a disposable elastomeric pump provides statistically significant improvements in patients' functional exercise capacity and free home daily activity after ambulatory orthopedic surgery. VAS pain scores during physiotherapy and patient satisfaction were also improved. The basal-bolus infusion was the ideal technique because it decreased the consumption of ropivacaine and reduced the number of adverse events.

Each group was divided equally into patients scheduled to undergo acromioplasties and hallux valgus surgeries. These procedures are eligible for ambulatory practice and are similar regarding the level of postoperative pain.21Other published studies have focused on shoulder or foot surgery regrouped hallux valgus correction and ankle open reduction or open rotator cuff repair and arthroscopic capsulotomy.10–12,15The postoperative pain and surgical management of these procedures are completely different.

The pain evaluations during the postoperative days were not performed under blinded conditions because of the clinical setting of the study. Rigorous scientific methods would have required placing a perineural catheter and attaching a PCA morphine pump in all patients (morphine or saline depending on the group). Because only the analgesic technique tested in each group would be used, obvious ethical reasons restrained our application of this method.

The novelty of this study lies in the 72-h postoperative analysis of the postoperative analgesic technique impact on the quality of patients' daily activities and functional exercise capacity. We demonstrated that the basal-bolus infusion of ropivacaine significantly optimized the patients' daily activities at home on days 2 and 3 and that PCA morphine increased the time necessary for a patient to walk freely during 10 min. To our knowledge, it is the first time that a randomized multicenter study has demonstrated that patients receiving a perineural infusion for 72 h had higher quality functional outcome during the course of the first 3 days after ambulatory orthopedic surgery. The difference reported between both perineural infusion groups may be due to (although not significant) an increase in some adverse events in the continuous infusion regimen (i.e. , sleep disturbance, slight paresthesia, dizziness). Interestingly, the main reasons for no activity or need for assistance during home activities were essentially pain and fatigue in the PCA morphine group. This result is in agreement with previous studies. Beauregard et al.  6reported that postoperative pain after ambulatory surgery was severe enough to interfere with daily activities in a substantial number of patients. Postoperative pain limited activity level in 73% of patients at 24 h and 61% after 48 h. Walking activity was limited in 69% and 49% of the patients, respectively. Wu et al.  22reported recently that poorly controlled postoperative pain limited physical activity and function of patients after total hip and knee surgery. In our study, 70% and 100% of the patients in the basal-bolus group had no limitation in their activity level or walking activity at 48 and 72 h, respectively. One might infer from the current study that excellent pain relief coupled with a reduction in side effects would contribute to facilitate postoperative mobilization at home. In inpatient shoulder and foot surgery, the effects of continuous perineural analgesia on the success of physical rehabilitation have not been fully studied,23,24but studies have reported that the improved analgesia provided by continuous femoral blocks resulted in faster short-term functional recovery of knee flexion.16,17For outpatients, such benefits were unknown until now. For clinically relevant reasons, we assessed our patients' functional exercise capacity with the measurement of the time achieved for a patient to walk freely during 10 min. We chose this time and not the distance covered during 10 min because it would be difficult to obtain a credible evaluation of the distance at home. On the other hand, in the postoperative period, we measured the capability of the patients to realize their daily activities at home with or without assistance. Controlled studies have covered hospitalized patients' perception of their recovery in the postoperative period.22,25–27However, in the validated survey instruments (Short Form-36 items, Short Form-12 items, Quality of Recovery score-40), some of the items used may not be appropriate for assessment of the quality of postoperative functional recovery; some questions related to limitations of either physical or mental health in patients at work, and redundancies appear between the pain domains of the questionnaires and VAS pain scores. Finally, Wu et al.  22reported that problems are encountered with any health-related quality of life instrument in evaluating the effect of postoperative pain on quality of recovery on a daily basis, because these instruments were not designed to assess recovery in short time frames.

Our results regarding painful foot and shoulder surgery confirm the significant superiority of perineural analgesia over opiate analgesia and are in agreement with controlled studies in hospitalized patients20,28,29and those involving patients discharged home with a perineural catheter.10–12,15,30In-hospital investigations highlighted the difference between perineural blocks and PCA intravenous morphine in providing more effective pain control during physiotherapy and mobilization,16,17,20,28,29but this comparison lacks a placebo-controlled trial regarding postoperative pain management at home.9–11,15,30Home PCA morphine is not considered standard after ambulatory surgery, but studies provide evidence that in the postoperative period, PCA morphine improves pain control and decreases opioid-related adverse effects compared with conventional opioid treatment.31,32As such, we thought that authors may have overstated the difference in postoperative pain measurements at home between patients receiving perineural infusion of local anesthetic and those receiving only oral analgesics in the event of breakthrough pain.9–13Our data suggest that the use of perineural analgesia to decrease levels of postoperative pain can facilitate earlier discharge from the hospital. White et al.  30reported that 40% of the patients in a continuous popliteal nerve block group (vs.  none in the placebo group) were able to be discharged home on the day of their foot surgery. Similarly, 26% of our PCA morphine group patients were not discharged home on the morning of day 1 because of episodes of breakthrough pain versus  6% in the continuous infusion group and 3% in the basal-bolus group.

We reported a lower incidence of nausea/vomiting, dizziness, and sleep disturbance in perineural analgesia group patients. These results are similar to previous studies10–15after outpatient orthopedic surgery. Patients rated vomiting as the most undesirable side effect after surgery,33and dizziness could limit free walking and the patient's daily activities. Consequently, patient satisfaction, as well as physician satisfaction, was significantly reduced in the PCA morphine group. In the current investigation, there were no medical complications attributable to the regional anesthesia technique (local anesthetic toxicity and complications secondary to sensory or motor blockades) in either perineural analgesia group. Continuous infusion group patients reported a higher incidence of slight paresthesia than patients in the other two groups. Borgeat et al.  23and White et al.  30also reported slight paresthesia or “tingling” in 30–80% of patients during continuous interscalene or popliteal nerve block. This adverse event did not affect the patients' ability to ambulate. No patient experienced dyspnea (interscalene block patients). Borgeat et al.  34reported limited diaphragm muscle strength impairment when using 0.2% ropivacaine in continuous interscalene blocks.

As reported by other authors after outpatient orthopedic surgery,12,13the use of a 0.2% ropivacaine basal-bolus infusion technique decreased overall ropivacaine consumption compared with a continuous infusion without limiting perioperative benefits. This limitation is of clinical importance because Zink et al.  35,36demonstrated short- and long-term myotoxic effects of local anesthetics in a clinically relevant porcine model of continuous perineural analgesia.

In summary, our data demonstrate that for outpatient shoulder and foot surgery, a perineural basal-bolus infusion of 0.2% ropivacaine via  a disposable elastomeric pump optimizes postoperative functional recovery, postoperative analgesia, and patient satisfaction while decreasing opioid requirements and their associated side effects, and consumption of ropivacaine. In the past decade, advances have been made to reduce postoperative pain at home. Investigations should now be focused on improvements in quality of functional outcome. Continuous peripheral nerve blocks are probably one of the key elements.