Abstract B94: An integrated database in colorectal cancer outcomes, prognosis, and epidemiology (The COPE database)

Introduction: With the move towards a completely electronic medical record system in the United States, there is an opportunity to leverage these electronic data for longitudinal epidemiological studies. Capturing data electronically that would otherwise have been manually abstracted reduces errors and increases the possible number of data elements collected; however these integrated data systems are not readily available. Therefore, for the purposes of a longitudinal study of colorectal cancer patient prognosis, we constructed an Oracle database to integrate multiple data sources. Experimental Procedures: A molecular epidemiology study was initiated to measure outcomes, epidemiology risk factors, obesity and inflammation-related molecular markers. An integrated Oracle database was constructed that combines several disparate data sources, including realtime messaging for clinic schedules (HL7), outcomes from a cancer registry and paper-based questionnaires into a single Oracle database. A web-based frontend was created to manage study activities. Results: The IRB-approved cohort study involves the identification of colorectal cancer patients, study-specific consenting and tracking of follow-up questionnaires and blood draws. A web-based interface was created to assist study coordinators and physicians in identifying eligible patients based on a real-time feed of scheduling and admissions. Once a patient has consented to the study, basic demographic information from clinical systems is populated into the study database. A web-based calendar view allows for scheduling study participants’ follow-up visits at 3, 6 and 12 months for blood draws and questionnaires during normal clinic visits. Patient questionnaire data is currently filled out on paper, scanned into electronic form and loaded into the database. Additional data, including anthropometric measurements and blood draw information are manually entered in the system. Although currently entered manually, research tissue collection details will be automatically populated within the study database. The integration of tissue gene expression data and biomarker results is currently under development. Conclusion: An integrated database system for patient accrual, consent status, patient information, questionnaires, anthropometric measurements and blood draws enables the successful collection of data for colorectal cancer prognosis cohort studies. This database serves as a prototype that can be customized for larger cohort studies that focus on cancer prognosis and survivorship. Citation Information: Cancer Prev Res 2010;3(12 Suppl):B94.