This document summarises the pre-specified Statistical Analysis Plan (SAP) for the PETRA randomised controlled trial. PETRA evaluates the addition of pregabalin to standard care and an antidepressant versus placebo plus standard care and an antidepressant in people with treatment-resistant generalised anxiety disorder. The SAP governs the planned analyses of trial outcomes prior to unblinding and database lock and is intended to reduce analytic bias by specifying methods in advance.
The SAP is presented as a stand-alone, pre-specified plan; this manuscript describes that plan rather than reporting trial results. No trial datasets were generated or analysed for this SAP paper.
The primary objective addressed by the SAP is to compare the effect of adding pregabalin versus placebo on anxiety severity in PETRA participants with treatment-resistant GAD. The primary outcome specified in the source is the GAD-7 score measured at week 12.
Secondary objectives, secondary outcomes, and exploratory analyses are referred to in the SAP itself. Where specific secondary endpoints, timing windows, or hierarchies are not reported in the public abstract/metadata, those details are contained in the full SAP document but are not restated here.
PETRA is described as a randomised controlled trial comparing two treatment strategies: pregabalin plus an antidepressant and standard care, versus placebo plus an antidepressant and standard care. The SAP documents how the trial treatment arms will be compared statistically for the primary and other outcomes.
Details such as randomisation ratio, blinding procedures, treatment titration, concomitant medications, inclusion/exclusion criteria, and planned sample size are part of the trial protocol and SAP. If specific numeric or procedural details were not reported in the publicly available abstract or metadata, those specifics are not introduced in this summary.
Ethics oversight was provided by The East Midlands - Leicester Central Research Ethics Committee of the Health Research Authority (REC reference: 23/EM/0192). The authors confirm that participant consent was obtained and that all relevant ethical guidelines were followed.
The clinical trial registration identifier is ISRCTN16993990. The SAP manuscript lists author contact information and institutional affiliations (University College London) for correspondence and accountability.
The SAP sets out the planned statistical principles and methods to be applied to PETRA data. Central elements of such a plan typically include definition of the primary analysis population, treatment of protocol deviations, model selection for the primary outcome, covariate adjustment, multiplicity control, and the timing of analyses. The source specifies that this document describes the SAP for the GAD-7 primary outcome at week 12.
Where the abstract or metadata do not include granular details of model form (for example, whether an analysis of covariance, mixed-effects model, or other approach is chosen), the full SAP should be consulted for precise model specifications. This summary does not invent specific analytic choices that were not explicitly reported in the source.
The SAP addresses how missing outcome data and other data issues will be managed and describes planned sensitivity analyses to assess robustness of findings. The public abstract/metadata indicate that this is a pre-specified analysis plan; however, if the exact imputation methods, assumptions for missing-at-random versus missing-not-at-random, or the number and type of sensitivity analyses are not reported in the source brief, those procedural details are available in the full SAP document and are not restated here.
The SAP manuscript and associated trial materials set out procedures for recording and reporting adverse events and other safety-related outcomes. In the metadata for this SAP preprint the authors declare no competing interests.
The PETRA trial and its SAP are funded by the National Institute for Health and Care Research, Health Technology Assessment (NIHR HTA), grant number 134074. The SAP paper reports that no datasets were generated or analysed as part of this manuscript because it describes the pre-specified analysis plan. The preprint is posted on medRxiv and is made available under a CC-BY 4.0 International license.
This rewritten SAP summary sticks strictly to information reported in the source medRxiv metadata and abstract for the PETRA SAP. Specific analytic recipes, sample size calculations, randomisation details, model parameterisations, and other granular statistical specifications are contained in the full SAP itself; where those items were not present in the abstract or metadata, this summary notes their existence without restating unstated specifics.
For trialists, statisticians, journal reviewers, and clinicians seeking the precise pre-specified analytic methods and exact model statements, the full SAP document associated with the PETRA trial should be consulted directly.