DOP089. Measurement of IBD disease activity using home-based e-health technologies
W. van Deen, J. Choi, A. Zand, C. Ha, E. Inserra, L. Eimers, A. Centeno, B. Roth, D. Cole, T. Getzug, E. Kane, L. Connoly, M. Ovsiowitz, A. Ho, M. van Oijen, E. Esrailian, D.W. Hommes, UCLA Center for Inflammatory Bowel Diseases, Division of Digestive Diseases, University of California, Los Angeles, Los Angeles, United States
Development of e-health technologies is accelerating due to a shift from ‘symptom-oriented’ to ‘prevention oriented’ care. The potential of monitoring patients at home offers great potential for preventive care. However, accurate e-health monitoring tools have yet to be developed. We (1) evaluated how patient reported outcomes predict disease activity and relate to quality of life; (2) tested the feasibility of collecting patient reported outcomes through e-health; and (3) developed an ‘IBD app’ for iOS and Android that allows easy and user-friendly collection of health outcomes and easy communication between healthcare provider and patient.
Disease activity indices were assessed in consecutive IBD patients. The predictive value of each patient-reported component was assessed using logistic regression analyses. As gold standards SCCAI and partial Mayo were used for UC, and CDAI and HBI for CD. The added value of the biomarkers CRP and fecal calprotectin was assessed as well. The feasibility of using e-health for the reporting of patient reported outcomes was evaluated using a web based application, which was subsequently developed into a mobile app for iOS and Android.
107 UC patients and 78 CD patients were included, of which 31 (29%) and 17 (22%) had active disease respectively. For UC, strong predictors of disease activity were urgency (73% with urgency had active disease, 5% without urgency had active disease) and blood in stool (68% with blood in stool had active disease, 3% that did not). Addition of calprotectin increased specificity to 100%, though sensitivity decreased to 67%. Abdominal pain predicted disease activity in CD; 71% of patients with abdominal pain had active disease, versus 4% without pain. 93% of CD patients with more than 2 stools per day had active disease, versus 6% that had 2 or less. Combining the outcomes abdominal pain and >2 stools per day predicted active disease with 70% sensitivity and 100% specificity. We did not detect additional value of a biomarker in CD. In September 2012 we launched a web-based application that is effectively utilized now by 335 patients. A mobile IBD app, co-created with patients, was tested and approved as of November 24, 2013 (search ‘UCLA eIBD’ in iTunes and Google Play stores).
E-health development offers great potential for continuous monitoring of patients at home, allowing early detection of disease activity and improving care delivery. Patients can participate in their care by signaling meaningful health outcomes during year-round monitoring. We showed that disease activity can be predicted using patient reported outcomes, and can be readily collected through e-health applications.