Objectives: The primary objective of this study was to assess the effectiveness of two prehospital stroke screens in correctly classifying patients suspected of having a stroke. Secondarily, differences in the sensitivity and specificity of the two screening tools were assessed.
Methods: We performed a retrospective assessment of the Cincinnati Prehospital Stroke Scale (CPSS) and the Medic Prehospital Assessment for Code Stroke (Med PACS) between March 1, 2011, and September 30, 2011, in a single emergency medical services (EMS) agency with seven local hospitals all classified as stroke-capable. We obtained data for this analysis from the EMS electronic patient care reports (ePCRs) and the Get With The Guidelines(-)Stroke (GWTG-S) registries maintained by the two local health care systems by matching on patient identifiers. The Med PACS was developed specifically for the EMS agency under study by a local team of neurologists, emergency physicians, and paramedics. All of the physical assessment elements of the CPSS were included within the Med PACS. Two additional physical assessment items, gaze and leg motor function, were included in the Med PACS. We classified patients as CPSS-positive or -negative and Med PACS-positive or -negative if any one of the physical assessment findings was present. We determined the presence of a hospital discharge diagnosis of stroke from GWTG-S. We calculated sensitivity and specificity with resultant 95% confidence intervals.
Results: We enrolled 416 patients in this study, of whom 186 (44.7%) were diagnosed with a stroke. The Med PACS scale demonstrated a sensitivity of 0.742 (95% confidence interval [CI] 0.672-0.802), while the sensitivity for the CPSS was 0.790 (95% CI 0.723-0.845). The sensitivity of the CPSS was significantly higher than that of the Med PACS, with a difference of 0.048 (95% CI 0.009-0.088; p = 0.011). The specificities of these two scales were low, Med PACS 0.326 (95% CI 0.267-0.391) vs. CPSS 0.239 (95% CI 0.187-0.300), and the specificity of the Med PACS was significantly higher compared with the CPSS, with a difference in specificity of 0.086 (95% CI 0.042-0.131), p < 0.001.
Conclusion: The two stroke scales under study demonstrated low sensitivity and specificity, with each scale performing marginally better in one of the two metrics assessed.