Initial Management of Patients with Suspected Stroke in the SARS-CoV-2 Era: Effects on the Door-to-Picture Time.

Objective: To examine the effectiveness of a newly developed emergency room (ER) protocol to treat patients with stroke and control the spread of SARS-CoV-2 by evaluating the door-to-picture time. Methods: We retrospectively enrolled 126 patients who were transported to our ER by ambulance with suspected stroke between April 15 and October 31, 2020 (study group). A risk judgment system named the COVID level was introduced to classify the risk of infection as follows: level 0, no infection; I, infection unlikely; II, possible; III, probable; and IV, definite. Patients with COVID levels 0, I, or II and a Glasgow Coma Scale (GCS) score >10 were placed in a normal ER (nER) without atmospheric pressure control; the medical staff wore standard personal protective equipment (PPE) in such cases. Patients with COVID level II, III, or IV, and a GCS score of ≤10 were assigned to the negative pressure ER (NPER); the medical staff wore enhanced PPE for these cases. The validity of the protocol was assessed. The door-to-picture time of the study group was compared with that of 114 control patients who were transported with suspected stroke during the same period in 2019 (control group). The difference in the time for CT and MRI between the two groups was also compared. In the study group, the time spent in the nER and NPER was evaluated. Results: In all, 118 patients (93.7%) were classified as level I, 6 (4.8%) as level II, and 2 (1.6%) as level III. Only five patients (4.0%) were treated with NPER. Polymerase chain reaction tests were performed on 118 out of 126 patients (93.7%) and were negative. No significant differences were observed in age, sex, neurological severity, modalities of diagnostic imaging, and diagnosis compared with the control group. The median door-to-picture time was 18 (11-27.8) min in the study group and 15 (10-25) min in the control group (p = 0.08). No delay was found on CT (15 [10-21] vs. 14 [9-21] min, p = 0.24). In contrast, there was an 8-min delay for MRI (30 [21.8-50] vs. 22 [14-30] min, p = 0.01). The median door-to-picture time was 29 min longer in patients treated with NPER than in those treated with nER, although the difference was not significant due to the small number of patients (47 [27-57] vs. 18 [11-26] min, p = 0.07). Conclusion: Our protocol could optimize the use of medical resources with only a 3-min delay in the door-to-picture time in an area without explosive outbreak. Unfortunately, the effectiveness of the protocol in preventing infection could not be verified because of the low incidence of COVID-19. When developing and modifying an institutional protocol, recognizing the outbreak status surrounding each institution is important.