Ethical approval was granted by the ethics committee of the Lausanne University Hospital, Switzerland (Commission cantonale d’éthique de la recherche sur l’être humain, study ID number: 318/14) on October 10, 2014, and the trial was registered on Clinicaltrials.gov (NCT02433548) on October 20, 2014. The CONSORT statement was followed in reporting this trial [24]. The study took place in the Emergency department of the Lausanne University Hospital between November 7, 2014 and June 2, 2016. Study participation was proposed to patients over 70 years old who were admitted for to the emergency department with fractured hip. Exclusion criteria were bleeding disorder or presence of anticoagulation, periprosthetic fracture, a known polyneuropathy, body weight below 40 kg, chronic pain condition, patients undergoing chemotherapy, infection at the site of injection, allergy to local anaesthetics and cognitive disorder.
After written informed consent was obtained, subjects were randomly allocated to either the experimental group (FIB group) or the control group (sham injection group) according to a computer-generated list of random numbers (www.randomization.com, seed 20,388). Assignments were concealed in a sealed opaque envelope. An emergency medicine physician, who was not responsible for the patient’s care, prepared the study drugs in syringes and performed the block procedure. A blinded study nurse collected the data. The physicians and nurses responsible for the patient’s care in the emergency department and on the ward were all blinded to the group allocation.
All procedures were performed in the emergency department, by emergency medicine physicians. Electrocardiogram, pulse oximetry, and blood pressure monitors were routinely applied. Oxygen was provided and peripheral intravenous access was established. Patients in the FIB group received an injection of 30 ml bupivacaine 0.5% with epinephrine 1:200,000, following an anatomical landmark-based technique previously described [10, 25]. Briefly, the site of injection was identified and marked 1 cm below the junction of the lateral and middle thirds of a line between the anterior superior iliac spine and the pubic tubercle. After sterilization with a solution of chlorhexidine 2% in isopropyl alcohol 70%, a short-bevel needle (Plexifix® 50 mm, 24 G, BBraun medical AG, Melsungen, Germany) was inserted at a right angle to the skin until 2 losses of resistance were felt, corresponding to the fascia lata and fascia iliaca piercing, respectively. The entire volume was then injected in 5 ml-increments. Patients assigned to the control group received a 5 ml-subcutaneous injection of normal saline in the same location after sterilization of the skin. All procedures were performed by physicians who had experience with the anatomical landmark-based technique and who were not further involved in the study.
After the procedure, patients received acetaminophen 1000 mg every 6 h. Persistent pain (numeric rating scale [NRS] > 3) was treated with intravenous morphine as needed. After transfer to the orthopaedic ward, patients were prescribed acetaminophen 1000 mg every 6 h and subcutaneous morphine 0.1 mg.kg− 1 every 6 h as needed.
We originally planned to compare the two mean pain scores (fascia iliaca block versus sham injection groups) at 45 min. However, given that, despite randomisation, the two groups had different mean baseline scores, we elected to evaluate the longitudinal pain trajectories and tested for parallelism instead. Therefore, the primary outcome was the comparison between groups of the longitudinal pain score profiles at rest over the first 45 min following the procedure. Pain intensity was assessed with an 11-point numeric rating scale (NRS), ranging from 0 (no pain) to 10 (worst imaginable pain). Secondary pain-related outcomes included comparison of the longitudinal pain score profiles on movement over the first 45 min following the procedure; longitudinal pain score profiles at rest and during movement (standardized gentle 15°-elevation of the leg) 4 h, 8 h, 12 h and 24 h after the injection (NRS, 0–10); and cumulative intravenous morphine consumption at 24 h after the injection (mg). Other outcomes were length of stay, and mortality at 3 months. The between group difference in NRS score over time was analysed as a post hoc addition to the protocol in order to explore the potential impact of baseline pain score differences between groups.
Based on published data, mean NRS pain score at rest when patients are untreated was anticipated to be 8/10 with a variance of 2 [18]. We determined a clinically meaningful decrease in pain score at rest to be 2 points after a fascia iliaca block. As a result, we calculated that a minimum of 10 patients per group was necessary to detect a difference between groups, with a power of 90% and an alpha error of 0.05, using a one-sided test. Allowing for a 50% patient drop-out rate due to protocol violation or consent withdrawal, we planned to enrol a total of 30 patients. Continuous and non-continuous data are summarized as medians with 25th–75th interquartile ranges (IQR), means and standard deviations or absolute numbers, when appropriate. Baseline characteristics of the two groups were compared using the Wilcoxon-Mann-Whitney rank sum test for continuous variables (age, pre-intervention drug dose received) and by the Pearson’s chi-squared test for discrete variables (gender, ASA score, etc.). The NRS over time was analysed using linear mixed models with a random effect to account for the correlation induced by the individual differences [26]. The analyses were adjusted for gender, ASA score, and pre-procedure intravenous morphine consumption. The between group difference in NRS was assessed over time for different follow-up periods using the Wald test. The interaction between gender and ASA category was also tested. The best dose-response functional forms for the pre-intervention cumulative intravenous morphine consumption were assessed by the method of fractional polynomials and the goodness of fit by residual analysis [27]. Significance was considered at p < 0.05. Statistical analyses were performed using the Stata 14.2 statistical package (Stata Corporation, College Station, Texas).