Statistical analysis plan for the Robotic-Assisted Surgery and Kinematic ALignment in total knee arthroplasty (RASKAL) registry-nested randomized trial

Aims There is uncertainty over whether robotic-assisted surgery (RAS) or kinematic alignment (KA) improves outcomes in total knee arthroplasty (TKA). The Robotic-Assisted Surgery and Kinematic ALignment in TKA trial (RASKAL) is a registry-nested, multicentre, randomized, blinded, 2 × 2 factorial trial involving approximately 300 TKA patients. RASKAL will assess the effectiveness of RAS, KA, or both to improve clinical, operative, functional, radiological, and survivorship outcomes when compared to computer-assisted surgery (CAS) and mechanical alignment (MA), respectively. We describe the statistical analysis plan (SAP) for the RASKAL trial to ensure transparency of this study. Our purpose is to minimize analysis bias to bolster the internal validity of our findings. Methods We created a SAP for the RASKAL trial to define the analysis of primary and secondary outcomes. The SAP will also state our approach to protocol deviations, withdrawals, management of missing data, and any loss to follow-up. General statistical analysis principles are outlined for the main outcomes, along with statistical tests required for each outcome. Conclusion The RASKAL SAP pre-specifies the statistical plan for the primary outcome of knee-specific function using the Knee injury and Osteoarthritis Outcome Score-12 comparing RAS to CAS and KA to MA up to one year postoperatively. The primary outcome will be reported as a modified intention-to-treat analysis, with secondary per-protocol analyses. Secondary outcomes will include other patient-reported outcome measures, early recovery, operative outcomes, functional outcomes, radiological precision, prosthetic survivorship, and cost-effectiveness analyses.