Study of Peri-Articular Anaesthetic for Replacement of the Knee (SPAARK): statistical analysis plan for a randomised controlled trial assessing the effectiveness of peri-articular liposomal bupivacaine plus bupivacaine hydrochloride compared with bupivacaine hydrochloride alone

Background Up to three quarters of surgical patients receive inadequate pain relief, with 40% of patients reporting severe pain following knee replacement, which may indicate the current pain relief strategies using opiate-based analgesia cannot achieve patient satisfaction. Liposomal bupivacaine is liposome-encapsulated bupivacaine which has been reported to be effective for up to 72 h. The study of Peri-Articular Anaesthetic for Replacement of the Knee (SPAARK) trial has been designed to assess the effectiveness of peri-articular liposomal bupivacaine and bupivacaine hydrochloride compared with peri-articular bupivacaine hydrochloride alone in the management of post-operative pain following knee replacement. Methods/design The SPAARK trial is a multi-centre, patient-blinded, randomised controlled trial. The co-primary outcomes are post-operative recovery assessed by global QoR-40 scores at 72 h and cumulative pain VAS score from 6 to 72 h following surgery. Longer-term measures of the co-primary outcomes are collected at 6 weeks and 6 and 12 months post randomisation, together with secondary outcomes, i.e. the Oxford Knee Score, and the American Knee Society Score. Cumulative opiate use and fitness for discharge are measured up to 72 h post-surgery. The analysis approaches for the primary and secondary outcomes are described here, as are the descriptive statistics which will be reported. The full SPAARK protocol has already been published. Results The co-primary outcomes will be analysed using multivariate linear regression adjusting for stratification factors and other important prognostic variables, including baseline scores in the case of the QoR-40. The adjusted mean difference between the two groups together with 97.5% confidence intervals will be reported for each of the primary outcomes. Other continuous variables will be assessed using the same method. Binary outcomes will be assessed using chi-squared tests. Discussion The paper provides details of the planned statistical analyses for the SPAARK trial and aims to reduce the risk of outcome reporting bias from prior data knowledge. Any changes or deviations from this statistical analysis plan will be described and justified in the final study report.