To determine if providers feel stroke alert treatment decisions are influenced by patient crying and if providers treat patients who cry during stroke alerts differently. Secondary aims include determining if crying during stroke alerts predicts presence of stroke.

Stroke alerts are frequent events, with over 1,200 each year at University of Colorado Hospital (UCH).  Treatment decisions are guided by thrombolytic criteria, risk factors, and possibly emotional factors. There is no literature that addresses the impact of patient crying on decision making in stroke alerts.

Surveys were administered to neurology trainees assessing experience with thrombolysis, guideline adherence, and whether providers felt treatment decisions were influenced by patient crying. ED stroke alerts at UCH from May to July 2017 were prospectively collected, including documentation of whether patients were crying, angry or laughing during the alert, treatment offered, and whether patient was determined to have stroke. All statistical analyses were performed in R version 3.5.3.

Data were collected for 179 of 257 stroke alerts during the study period. 9% of patients were crying, 2% were angry, and 3% were laughing during stroke alerts. Patients who cried were significantly younger (p=0.003) and female (p=0.031). There was no significant difference in treatment (p=0.295) or further diagnostic (p=0.457) decisions between crying and not crying patients. 6% of crying patients were found to have a stroke compared to 23% of not crying patients (p=0.11).

While trainees expect patient crying to impact treatment decisions in stroke alerts, no significant differences were found in treatment or further diagnostic decisions for crying patients during stroke alerts. Crying patients are significantly more likely to be younger and female, and may be less likely to have a stroke, though this did not reach statistical significance.