Article Title

Fidelity Evaluation of a Care Coordination Solution for Integrative Medicine: Usage and User Acceptance

Publication Date



program evaluation, referrals, referral networks, quality of care, evaluation research, research administration, contracting, operations, dissemination and implementation of innovations, patient experience/satisfaction


Background: Care coordination is an important aspect of patient care, especially for integrative health, which includes many types of providers on a care team. We developed and piloted an application (Trellis) designed to improve care coordination for patients at the Institute for Health and Healing, an integrative medicine specialty care clinic in San Francisco, California. Trellis is built to improve transparency, awareness and communication between care-team members. We evaluated the usage and user acceptance of Trellis to determine whether we should continue the dissemination process by creating the next version.

Methods: We implemented Trellis and trained the 32 members of the care-team staff. We measured usage by counting active sessions and number of times each feature was accessed. We stratified by provider modality and role, and because most of the practitioners are part-time employees, we weighted usage estimates by hours worked per week. We held a focus group to measure user acceptance. When low usage was identified, we made changes to the application and workflow to improve utilization.

Results: Through July 2016, Trellis has impacted care for 383 (27%) patients with encounters at the Institute for Health and Healing. Overall, 43% of the usage can be attributed to medical practitioner, 20% to allied health practitioners and 37% to administrative staff. Trellis usage evaluation identified 4 super-users and 7 under-users. Changes to the system, including training, bug fixes and onsite support, resulted in better short-term usage, but not long-term. Trellis helped to track referrals; 72% of practitioners refer to themselves and about 28% to specialists of other modalities. Focus groups identified barriers such as culture, workstation set-up and uncertainty about overall purpose.

Conclusion: Trellis adoption was low, only used for one-fourth of patient encounters, but there was enthusiasm for future versions and future use. Weekly usage combined with focus group data indicated that the enthusiasms to use the application could not be maintained over time due to cultural barriers. With this early evaluation, we have provided a foundation of information on which to design future versions of the application.




June 28th, 2017


August 10th, 2017