• Anthony P Delaney, Sandra L Peake, Rinaldo Bellomo, Peter Cameron, Anna Holdgate, Belinda Howe, Alisa Higgins, Jeffrey Presneill, Steve Webb, and ARISE investigators.
• Malcolm Fisher Intensive Care Unit, Royal North Shore Hospital, Sydney, NSW, Australia. adelaney@med.usyd.edu.au
• Crit Care Resusc. 2013 Sep 1; 15 (3): 162-71.

BackgroundThe Australasian Resuscitation in Sepsis Evaluation (ARISE) study is an international, multicentre, randomised, controlled trial designed to evaluate the effectiveness of early goal-directed therapy compared with standard care for patients presenting to the emergency department with severe sepsis.ObjectiveIn keeping with current practice, and considering aspects of trial design and reporting specific to non-pharmacological interventions, our plan outlines the principles and methods for analysing and reporting the trial results. The document is prepared before completion of recruitment into the ARISE study, without knowledge of the results of the interim analysis conducted by the data safety and monitoring committee and before completion of the two related international studies.MethodsOur statistical analysis plan was designed by the ARISE chief investigators, and reviewed and approved by the ARISE steering committee. We reviewed the data collected by the research team as specified in the study protocol and detailed in the study case report form. We describe information related to baseline characteristics, characteristics of delivery of the trial interventions, details of resuscitation, other related therapies and other relevant data with appropriate comparisons between groups. We define the primary, secondary and tertiary outcomes for the study, with description of the planned statistical analyses.ResultsWe have developed a statistical analysis plan with a trial profile, mock-up tables and figures. We describe a plan for presenting baseline characteristics, microbiological and antibiotic therapy, details of the interventions, processes of care and concomitant therapies and adverse events. We describe the primary, secondary and tertiary outcomes with identification of subgroups to be analysed.ConclusionWe have developed a statistical analysis plan for the ARISE study, available in the public domain, before the completion of recruitment into the study. This will minimise analytical bias and conforms to current best practice in conducting clinical trials.

11 keywords...

hide…