Correlating web usage of health information with patient medical data

The number of online websites providing health-related information to the general public increases daily. Yet, it is relatively unknown as to how individuals in the general population access information with respect to their own medical status. In this study, clickstream data from an online health information website is analyzed with respect to the user's health insurance claims. The relationships are assessed through the construction and study of intersecting sets of ICD-9 codes in visited web pages and claims made. Results demonstrate that approximately 15% of patients use health information on the web in exact agreement with their medical status. In addition, almost 40 codes were found to be indicative of temporal aspects in user behavior with respect to physician visits.