Abstract TP74: Risk Assessment Of The Acute Stroke Diagnostic Process
Jane Holl,Matthew Maas,Sarah Song,Alejandro Vargas,Elida Romo,Shyam Prabhakaran
DOI: https://doi.org/10.1161/str.53.suppl_1.tp74
IF: 10.17
2022-02-01
Stroke
Abstract:Introduction: Acute stroke (AS) is a high-harm, high-cost condition that affects nearly 800,000 people/year in the US. Proven, time-sensitive treatments can reduce disability. However, to deliver an AS treatment, a timely and accurate diagnosis is first needed. Yet, diagnostic error is the most common type of error in AS, occurring in ~10% of AS patients, and higher in patients with mild or atypical presentations. Hypothesis: The identification, characterization, and ranking of failures of the AS diagnostic process, by clinicians who provide AS care in the ED, is an essential step prior to designing feasible, robust, and effective solutions to reduce AS diagnostic error. Methods: A Learning Collaborative (LC) of clinicians involved in the AS diagnostic process at 3 health systems in Chicago, IL participated in a Failure Modes, Effects, and Criticality Analysis (FMECA) to identify the steps in the AS diagnostic process, failures of each step and their underlying causes, and to characterize each failure’s frequency (F), impact (I) on making a timely and accurate AS diagnosis, and any existing safeguards (S), using standardized scores. A risk priority index (RPN=FxIxS) and a criticality number (CN=FxI) was calculated for each failure and rank ordered. Results: In a series (N=7) of 60-90 minute virtual sessions, the LC, comprised of Emergency Medicine (N=9), Neurology (N=10), and Radiology (N=1) clinicians of all professions (MD, RN, Technician) and levels (resident, attending, coordinator), created an AS diagnostic process map and risk table. The process map included 27 specific steps. The highest risk steps were failure to use a severe stroke/large vessel occlusion scale (RPN=432; CN=72); inability to establish patients’ last known well (RPN=384; CN=48); failure to use an AS screening scale (RPN=384; CN=54); lack of a “witness” of the event to confirm information (RPN=378; CN=42); and failure to recognize potential stroke and activate a stroke code at triage (RPN=288; CN=48). Conclusion: This study, for the 1 st time, reveals specific targets, particularly in the early phase of the AS diagnostic process, for which solutions should be designed (e.g., standardized process to use a severe stroke/LVO tool) to reduce AS diagnostic error.
peripheral vascular disease,clinical neurology