Delayed arrival to the hospital in stroke is a major unsolved problem in public health that leads to stark and persistent racial and socioeconomic disparities in stroke outcomes. The delay generates health disparities because racial minority and socioeconomically disadvantaged patients arrive later than White patients leading to less access to treatment and worse outcomes. The most common reason for delay is the time spent by the patient and witnesses who decide together to watch-and-wait or go to the hospital. Therefore, we propose that social connectedness is a major determinant of the delay phenomenon. Our team has demonstrated that social network structure around a specific patient determines the flow of information that leads to decisions to act rapidly or slowly. Patients who arrived early had large and loosely connected networks, while those who arrived late had small and close-knit networks. What remains lacking, however, is knowledge of the extent of the social network effect in a more diverse population of stroke patients, its mechanism, and translation into interventions to improve stroke delay and disparities. This understanding is critical to establishing rigor and premise for future social network interventions aimed at reducing disparities in stroke outcomes. Our long-term goal is to design network-based interventions that reduce delay during stroke and ensure equitable access to therapies. Therefore, in this project, we use a dual empirical and social simulation approach to characterize and model social network effects in a diverse patient population. In Aim 1, we will determine whether social networks affect delay in hospital arrival after stroke differentially by race and socioeconomic status. We will capture social network data and time to arrival in 500 racially and socioeconomically diverse patients during their hospital admission. In Aim 2, we will model the potential of network interventions to improve stroke delay in at-risk populations. Using data from the same 500 patients and persons in their network, we will parameterize an agent-based model to represent the dynamic decision-making within the social network during stroke. Then we will evaluate the potential effects of network interventions to improve delay and disparities within the model. Our central hypothesis is that social network metrics will be associated with hospital arrival time, social networks will moderate race and SES differences in arrival time, and that network interventions such as increasing network size will improve outcomes and disparities in social simulations. We have assembled a multidisciplinary team with expertise in stroke, social networks, agent-based modeling, and health disparities to execute this project. The proposed research will provide much needed empirical data on social network effects and the potential of network interventions to address stroke delay and its disparities. These results will have a positive impact by directly setting t...