HeadPoST: Rightly positioned, or flat out wrong?. (8th May 2018)

• Record Type: Journal Article
• Authors: Alexandrov, Anne W.; Tsivgoulis, Georgios; Hill, Michael D.; Liebeskind, David S.; Schellinger, Peter; Ovbiagele, Bruce; Arthur, Adam S.; Caso, Valeria; Nogueira, Raul G.; Hemphill, J. Claude; Grotta, James C.; Hacke, Werner; Alexandrov, Andrei V.
• Is Part Of: Neurology. Volume 90:Number 19(2018)
• Publication Date: 2018-05-08
• Subjects: Neurology -- Periodicals; Neurology -- Periodicals; Neurologie -- Périodiques; 616.8
• DOI: 10.1212/WNL.0000000000005481
• Languages: English
• ISSNs: 0028-3878

Abstract

Abstract : Objective: To critique the Head Positioning in Stroke Trial (HeadPoST) study methods in relation to preceding research findings in an aim to clarify the potential efficacy of positioning interventions and direction for future research. Methods: Head positioning research prior to the conduct of HeadPoST was reviewed by a team of international stroke experts, and methods and findings were compared to HeadPoST. Results: Methods used to select HeadPoST patients differ substantially from those used in original head positioning studies, in particular enrollment of all types of stroke. HeadPoST enrolled primarily minor strokes (median NIH Stroke Scale 4, interquartile range [IQR] 2–8) without vascular imaging confirmation of subtype; elapsed time from stroke symptom onset to the initiation of intervention was late (median 14 hours, IQR 5–35), and time from hospital admission to enrollment was delayed (median 7 hours, IQR 2–26). Intervention integrity was not reported, including ability to achieve/maintain 30° head elevation in beds lacking head elevation capabilities. Deterioration or improvement associated with the intervention is unknown as serial assessments were not completed, and the trial's 3-month outcome was powered using unrelated study data. Conclusions: The design of HeadPoST was suboptimal to measure differences produced by the intervention. Future head positioning trials in discrete patient cohorts (in particular, large vessel occlusion) with endpoints supported by pilot work are required to understand the efficacy of this simple yet potentially important intervention.