ABSTRACT NUMBER: ESOC2026A1505 IDENTIFICATION OF STROKE MIMICS IN THE PRE-HOSPITAL SETTING: EVALUATION OF A STRUCTURED HOSPITAL PRE-ALERT QUESTIONNAIRE

Abstract Background and aims A regional standardised questionnaire was developed to facilitate direct referrals from paramedics in the community to hospital hyperacute stroke units (HASU). This study aimed to evaluate the validity of a pre-alert process for distinguishing between strokes and stroke mimics for directing patients to a HASU or emergency department (ED) accordingly. Methods Retrospective data were collected from pre-alert forms and hospital discharge summaries, for referrals made between October 2024 – April 2025 to the HASU. Individual variables recorded on the pre-alert form (Figure 1), were compared against final diagnosis to determine associations. Decisions to accept or reject patients from the HASU were compared against final diagnosis to determine triage accuracy. For patients referred by paramedics, the sensitivity and specificity of the pre-alert process was then compared with the FAST test alone. Results Data from 364 patients was collected. 4 patients with illegible forms were removed, leaving 360 patients, 230 of which were referred by paramedics and 130 by general practitioners (GP) or EDs. Sensitivity & specificity of the pre-alert triage process was 0.909/0.559 compared to 0.950/0.272 when using FAST test alone. Increased age (P<0.001) and contemporaneous raised systolic blood pressure (P = 0.001) significantly predicted a stroke versus a stroke mimic. Conclusions The pre-alert process improved specificity for stroke compared with FAST alone but at the expense of some sensitivity. Consideration of patient’s age and current systolic blood pressure may provide significant value to the decision-making process. Conflict of interest Sam James: nothing to disclose. Aditya Dass: nothing to disclose. Joyce Balami: nothing to disclose. Jessica Redgrave: nothing to disclose Figure 1 - belongs to Methods