Improved paramedic sensitivity in identifying stroke victims in the prehospital setting

The authors have previously shown that San Francisco paramedics without specific training in stroke recognition identified acute stroke victims with a 61% sensitivity and a 77% positive predictive value (PPV). The authors implemented an educational program on stroke to improve paramedic accuracy in stroke recognition.

Twenty-two paramedics volunteered to attend a four-hour seminar about stroke and then were followed prospectively for six months. All encounters with adult patients who were evaluated by both trained and untrained paramedics and were transported to two university hospitals were reviewed. Subjects were identified by paramedic assessment as stroke/transient ischemic attack (TIA) and/or final hospital discharge diagnosis of stroke/TIA after detailed chart review. Sensitivity and PPV for paramedic identification of stroke were calculated.

During the prospective six-month phase, 84 confirmed stroke patients were transported to the target hospitals. Of the 32 who were transported by trained paramedics, all but three were identified as having stroke/TIA, resulting in a sensitivity of 91%. This is significantly higher than the 61% previously found (p=0.01). Nontrained paramedics also increased their sensitivity to 90%. Thirty-eight false-positive patients were identified, resulting in PPVs of 64% for trained paramedics and 69% for all other paramedics.

Institution of an educational stroke program was associated with a significant increase in sensitivity in stroke identification by the paramedics; however, educational influences outside this training program may have contributed to the increased sensitivity. Better education for paramedics, combined with rapid response to stroke victims once identified, may result in improved care for victims of acute stroke.