Electronic Health Information Exchange (HIE) between healthcare providers is increasing, driven in part by public support. HIE holds the potential to improve patient outcomes and provider efficiency by giving healthcare providers access to all relevant patient information. However, HIE will only provide patient benefits if it meets the information needs of providers and their patients. There are several available models of HIE, and it is not clear whether hospitals are selecting the model that best meets their information needs for care delivery or are choosing HIE models based on competitive and collaborative market factors, which may lead them to models that are not optimal for patient care. As currently developed, available HIE models present trade-offs to hospitals. Participation in health information organizations (HIO), third party groups that intend to connect all providers in the market, can connect many participants but often does not support rich or customizable data sharing between participants. On the other extreme, enterprise HIE, such as customized interfaces between hospital partners, provide rich, customized data sharing, but are costly and unlikely to link all partners. In the proposed dissertation, I will use tools from social network analysis to identify hospitals' information needs, measured as the concentration of shared patients between hospitals, and link these needs to HIE model selection. Hospitals whose patients are concentrated in just a few partners may best meet their needs by investing in customized links with those partners, while hospitals with highly dispersed patients will best meet their information needs through HIO participation that promises to connect all partner hospitals. Two market factors, the collaborative structure of cohesion and clustering in the market, and hospital competition, are also likely to influence hospitals' HIE model selection, but may not align with patient information needs, and I will assess the extent to which these factors are driving HIE model selection. Finally, I will measure the benefit associated with hospital participation in HIE. I predict that hospitals and their patients will benefit most from participaton in HIE that matches the hospitals' information needs. The proposed research meets AHRQ's mission to promote policy that will improve the quality, safety and efficiency of patient care by understanding and meeting current information needs and understanding health IT's impact on outcomes (Research Area of Interest 1 & 4 in Special Emphasis Notice NOT-HS-13-011). I hope to provide guidance to organizational leaders and policy makers as they decide if and how to make further investments in HIE.
The proposed dissertation addresses the need to better understand how providers are making use of electronic health information exchange and if current strategies are benefiting patients. I will use social network analysis to identify hospital networks of shared patients that create needs to share health information for care delivery, and assess the extent to which the hospitals' choice of HIE model meets these information needs or is driven by other market factors unrelated to patient information needs. I will then assess whether hospitals engaged in HIE that appears to fit their information needs achieve greater benefits for their patients than hospitals engaged in mis-fitting HIE models, providing guidance to policymakers and organizational leaders as they continue to design HIE policy.
