Comparing Efficacity of Analgesia Between Ultrasound-guided Nerve Block and Local Infiltration After Wrist Fracture Surgery

Patients were allocated to treatment group according to a computer-generated randomisation sequence by blocs. Randomization was obtained and all data recorded through an electronic report form (URC Robert Debre).

Demographic and medical information were obtained in a pre-operative anaesthesia consultation kept with the rest of the patient's written file.

All interventions were carried out under regional anaesthesia consisting of an ultrasound-guided axillary block with 25-35 mL of lidocaine 1,5% with epinephrine, targeting the median, ulnar, radial, musculocutaneous and lateral cutaneous nerves. The analgesic intervention was realised according to group allocation, at the end of surgery: after skin closure and before making the cast. To ensure the blinding of the patient, the surgical sterile drapes were kept until the end of the analgesic intervention

• In the peripheral nerve block group (BN), the anaesthesiologist performed an ultrasound-guided block of the median and radial nerves at the lower third of the arm. Each nerve was encircled with 10 mL of ropivacaine 0,475%.
• In the local infiltration group (INF), the surgeon injected 10mL of ropivacaine 0,75% directly all peri-fracture tissues: fracture site, tendon synovial sheaths hypodermis and dermis regarding surgical incision.

An operating room or aesthetic nurse timed the duration of realisation of the analgesic intervention.

The intensity of proximal motor nerve block was recorded before patient discharge from the post anaesthesia care unit, using a modified Bromage Scale.

In each group the pain management was standardised: in the post anaesthesia care unit (PACU) the patients received 1g of acetaminophen and 550mg of naproxen (NSAID) as well as 20 mg omeprazole given orally. This treatment was continued in the orthopaedic ward, every 6 hours for acetaminophen, 12 hours for naproxen and every day for omeprazole. If the pain score reached 3/10, rescue medication was given: 10 mg of oral morphine sulphate, available every 6 hours. Upon discharge, the patient was given a prescription with these same medications for two days and clear instructions on the use of rescue opioids.

The patient was asked to record pain scores at 3, 6, 12, 24, 36 and 48 hours, opioid consumption, opioid related adverse events, quality of sleep and satisfaction with pain management on a document intended for this purpose.

The pain scores were recorded using a Numeric Rating scale for pain ranging from 0 (no pain) to 10 (worst possible pain). Total opioid consumption was assessed in milligrams. Quality of sleep as well as patient satisfaction with pain management were recorded using a numerical rating scale ranging from 0 to 10.

Interns or medical practitioners unaware of patient allocation then collected the data. If the patient was discharged within 2 days, data was recovered by telephone. The anonymized data was then recorded in the electronic report form. All electronic data was declared to the CNIL (French National Committee for Data protection and Liberties).

According to our preliminary data, the standard deviation of the mean duration of analgesia for the first 48 postoperative hours was 108 minutes. We chose to set the equivalence margin at 90 minutes, as it seemed clinically relevant to describe a difference in duration of analgesia for the patient. With a risk alpha = 5% and beta = 10% and 10% of patients expected not assessable, the estimated number of patients to include was 72 (36 in each group).

As this study aims to show equivalence, a per-protocol analysis will be carried out in first intention. An intention-to-treat analysis will follow to assess the results' reliability and treatment's efficiency.

Patient characteristics will be described for the two groups.

The primary study endpoint was the delay between analgesic intervention and occurrence of a pain score greater than 3/10 during the first 48 postoperative hours. The equivalence margin of 90 minutes was considered clinically relevant, defining an equivalence interval between -90 and + 90 minutes. If the confidence interval of the difference in analgesia duration between the two techniques is included in this interval, the two techniques will be considered equivalent. On the contrary if the confidence interval of the difference is only partially included of not included in this interval, we will conclude to non-equivalence.

The secondary end-points will be included in a superiority analysis.

• The mean difference for the duration of realisation of analgesic intervention will be compared using Student's t-test or a non-parametric test.
• The intensity of proximal motor block assessed by modified Bromage Scale will be compared using a chi-squared test or Fischer's exact test.
• The number of patients with a numerical pain rating ≤ 3, the number of patients requiring opioids, the frequency of opioid-related adverse events during the first 48 postoperative hours and the quality of sleep on the first night on a numeric scale from 0 to 10 will be compared using a chi-squared test or Fischer's exact test.
• Total opioid consumption (in milligrams) as well as patient satisfaction with pain management during the first 48 postoperative hours will be compared using Student's t-test or a non-parametric test.
• The pain scores recorded at 3, 6, 12, 24, 36 and 48 hours postoperatively will be described with a graphic analysis and compared using a mixed model reflecting the repeated measures.