Statistical analysis plan of the head position in acute ischemic stroke trial pilot (HEADPOST pilot)

Olavarría, Verónica V, Arima, Hisatomi, Anderson, Craig S, Brunser, Alejandro, Muñoz-Venturelli, Paula, Billot, Laurent and Lavados, Pablo M (2017) Statistical analysis plan of the head position in acute ischemic stroke trial pilot (HEADPOST pilot). International Journal of Stroke, 12 (2). pp. 211-215. ISSN 1747-4930

[thumbnail of HeadPoST statistical analysis plan] PDF (HeadPoST statistical analysis plan) - Supplemental Material
Restricted to Registered users only


Official URL:


Background: The HEADPOST Pilot is a proof-of-concept, open, prospective, multicenter, international, cluster randomized, phase IIb controlled trial, with masked outcome assessment. The trial will test if lying flat head position initiated in patients within 12 h of onset of acute ischemic stroke involving the anterior circulation increases cerebral blood flow in the middle cerebral arteries, as measured by transcranial Doppler. The study will also assess the safety and feasibility of patients lying flat for ≥24 h. The trial was conducted in centers in three countries, with ability to perform early transcranial Doppler. A feature of this trial was that patients were randomized to a certain position according to the month of admission to hospital. Objective: To outline in detail the predetermined statistical analysis plan for HEADPOST Pilot study. Methods: All data collected by participating researchers will be reviewed and formally assessed. Information pertaining to the baseline characteristics of patients, their process of care, and the delivery of treatments will be classified, and for each item, appropriate descriptive statistical analyses are planned with comparisons made between randomized groups. For the outcomes, statistical comparisons to be made between groups are planned and described. Results: This statistical analysis plan was developed for the analysis of the results of the HEADPOST Pilot study to be transparent, available, verifiable, and predetermined before data lock. Conclusions: We have developed a statistical analysis plan for the HEADPOST Pilot study which is to be followed to avoid analysis bias arising from prior knowledge of the study findings.

Repository Staff Only: item control page