Abstract
BACKGROUND: Telemedicine has emerged as a critical tool in the management of acute stroke; yet, its impact on clinical decision-making, particularly in the administration of tissue plasminogen activator (tPA), remains underexplored. Understanding how telemedicine influences tPA use and subsequent patient outcomes, such as hemiplegia, is critical for optimizing stroke care protocols. OBJECTIVE: This study aims to assess whether the adoption of telemedicine influences treatment decisions regarding tPA administration in patients with stroke. In addition, we used a causal mediation framework to examine the causal path between telemedicine adoption and the occurrence of hemiplegia via tPA use. Furthermore, we conducted a moderated mediation analysis to investigate the extent to which regional differences (metropolitan vs nonmetropolitan locations) impact this mediated relationship. METHODS: We analyzed data of patients with stroke from Florida's State Emergency Department Database (SEDD), State Inpatient Database (SID), and the Healthcare Information and Management Systems Society (HIMSS) database, covering the years 2010 to 2017, with a focus on telemedicine adoption. The final sample includes 314,665 visits from patients with stroke. A fixed-effects model was used to examine the relationship between telemedicine adoption and tPA use and between tPA use and hemiplegia occurrence. A causal mediation framework was then applied to estimate the average direct effect and the average causal mediation effect of telemedicine on hemiplegia through tPA use. In addition, a moderated mediation analysis was performed to explore how metropolitan status influences the mediation process. RESULTS: We found that telemedicine adoption is associated with a 1% decrease in tPA use (coefficient=-0.010; 95% CI -0.013 to -0.007; P<.001), and that the use of tPA is associated with a 23.7% increased probability of hemiplegia (coefficient=0.237, 95% CI 0.231-0.243; P<.001). Consequently, telemedicine adoption was found to reduce the likelihood of hemiplegia by decreasing tPA usage. The causal mediation analysis shows a negative average causal mediation effect (average causal mediation effect=-0.002, 95% CI -0.003 to -0.002; P<.001), suggesting that telemedicine adoption reduces hemiplegia occurrence, while the average direct effect is not statistically significant (average direct effect=-0.002, 95% CI -0.007 to 0.004; P>.10). Importantly, the positive effect of telemedicine on reducing hemiplegia is observed only among metropolitan patients. CONCLUSIONS: This study provides evidence that telemedicine adoption can improve stroke care by reducing tPA administration, thereby lowering the risk of hemiplegia. However, the benefits appear more pronounced in metropolitan areas, highlighting potential regional disparities in stroke care. These findings underscore the importance of targeted interventions to ensure equitable access to telemedicine, especially in rural and underserved areas. Policy makers should focus on enhancing telemedicine infrastructure and training in these regions to optimize stroke care and reduce health inequities.