ACL Surgery Necessity in Non-Acute Patients (ACL SNNAP): a statistical analysis plan for a randomised controlled trial

Abstract

Background: Rupture of the anterior cruciate ligament (ACL) is a common injury, primarily affecting young, active individuals. Despite surgical intervention being the more common treatment for patients suffering ACL ruptures, current management is based on limited and generally low-quality evidence. We describe a statistical analysis plan (SAP) for the ACL SNNAP randomised controlled trial, which aims to investigate the necessity of surgical management in patients with ACL injuries. Methods/design: ACL SNNAP is a pragmatic, multi-centre, superiority, parallel-group randomised controlled trial in participants with a symptomatic non-acute ACL deficient knee. Participants are allocated in a 1:1 ratio to either non-surgical management (rehabilitation) or surgical management (reconstruction) with the aim of assessing the efficacy and cost-effectiveness. The primary outcome of the study is the Knee Injury and Osteoarthritis Outcome Score (KOOS4) at 18 months post-randomisation. The KOOS4 score at 18 months will be evaluated using a linear regression model adjusting for recruitment centre and baseline KOOS4 scores, allowing for intra-centre correlation. A secondary analysis of the primary outcome will be carried out using an area under the curve (AUC) approach using treatment estimates obtained from a mixed model using baseline, 6 months, 12 months, and 18 months post-randomisation outcome data. Secondary outcomes will be measured at 18 months and will include return to activity/level of sport participation, intervention-related complications, the EQ-5D-5L questionnaire, all 5 individual subscales of the KOOS questionnaire, the ACL-QOL score, expectations of return to activity and cost-effectiveness of the interventions. Missing primary outcome data will be investigated through a sensitivity analysis. Full details of the planned methods for the statistical analysis of clinical outcomes are presented in this paper. The study protocol for the ACL SNNAP trial has been published previously. Discussion: The methods of analysis for the ACL SNNAP trial have been described here to minimise the risk of data-driven results and reporting bias. Any deviations from the analysis methods described in this paper will be described in full and justified in the publications of the trial results. Trial registration: ISRCTN ISRCTN10110685. Registered on 16 November 2016