|
Abstract:
|
Background: Poor care transitions have been identified as major contributors to poor quality and waste in health care system in US. Patients leaving from one setting to another, their safety may be affected and contributed to adverse events when important clinical information is missing, incomplete, or inaccurate. Objectives: To visualize the network of patient inter-facility transactions within major US health insurance plan; to measure facilities characteristics and connectivity within health care system; and to test the association between facilities' network properties and patients' health outcomes and utilization. Methods: We will select facilities and providers in major US health insurance plan to create share patient's networks. Providers' patient panel characteristics (patient's average age, patient volume and patient's risk index) will be aggregated from medical claim records. Providers and facilities' network position properties will be calculated within the sharing patient's health care network, using python social network analysis packages. Machine learning model will be used to measure the association between providers' characteristics and patients' health
|
