Prehospital Stroke Identification: Factors Associated with Diagnostic Accuracy

J Stroke Cerebrovasc Dis. 2015 Sep;24(9):2161-6. doi: 10.1016/j.jstrokecerebrovasdis.2015.06.004. Epub 2015 Jul 7.


Background: Stroke patients misdiagnosed by emergency medical services (EMS) providers have been shown to receive delayed in-hospital care. We aim at determining the diagnostic accuracy of Fire Department of New York (FDNY) EMS providers for stroke and identifying potential reasons for misdiagnosis.

Methods: Prehospital care reports of all patients transported by FDNY EMS to 3 hospitals from January 1, 2010, to December 31, 2011, were compared against the American Heart Association Get With The Guidelines (GWTG) database (reference standard) for the diagnosis of stroke. Age-adjusted logistic regression models were generated to explore prehospital patient characteristics which are associated with stroke misdiagnosis.

Results: Of 72,984 patient transports during the study period, 750 had a GWTG diagnosis of stroke, 468 (62%) of which were identified correctly in the field and 282 (38%) were missed. An additional 268 patients were misdiagnosed as stroke when in fact they had an alternative diagnosis. Overall sensitivity was 62.4% (95% confidence interval [CI], 58.9-65.8) and specificity was 99.6% (95% CI, 99.6-99.7). No patients who presented with unilateral weakness, facial weakness, or speech problems were missed, whereas patients with atypical complaints like general malaise, dizziness, and headache were more likely to be missed. Seizures led the EMS providers to both overcall a stroke and miss the diagnosis.

Conclusions: FDNY EMS care providers missed more than a third of stroke cases. Seizures and other atypical presentations contribute significantly to stroke misdiagnosis in the field. Our findings highlight the need for better prehospital stroke identification methods.

Keywords: Prehospital; acute stroke care; emergency medical services; paramedics; stroke; stroke scale.

Publication types

  • Multicenter Study
  • Research Support, N.I.H., Extramural

MeSH terms

  • Adult
  • Aged
  • Diagnostic Errors
  • Emergency Medical Services / methods*
  • Emergency Medical Services / standards*
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • New York City
  • Reference Values
  • Sensitivity and Specificity
  • Stroke / diagnosis*
  • Stroke / pathology
  • Stroke / physiopathology