Professor Kai Zheng

User Reaction to IT Systems in Health Care: A Multi-Site Multi-Approach Study
Professor Kai Zheng, School of Public Health and School of Information, University of Michigan

Description

Semester: 
Winter 2008
Lecture Time: 
Friday, April 11, 2008 (All day)
Lecture Location: 

Room 4212, School of Education

Abstract

Healthcare institutions in the United States are under increasing pressure from patients, payers, and regulatory bodies to create a better delivery system that controls operating costs while maintaining quality of care and services. Health IT in general, electronic health records in particular, provides great promise for helping achieve this goal through effectively managing information and facilitating total quality management and continuous quality improvement programs. However, relative to the vast investment in health IT systems, their impact to date has been very limited. Resistance to change by healthcare workers and the misalignment of technology design with established practice styles and clinical workflow are believed to the cause of the deficiencies. We conducted formative and summative evaluations of an electronic health records system deployed in 3 ambulatory primary care clinics at an urban teaching hospital. The study first uses a developmental trajectory analysis approach to stratify clinician users according to their system usage objectively recorded. Users are found to cluster in 3 major categories; each demonstrated distinct longitudinal technology acceptance behavior unfolded over a 10-month time period. Next, sequential pattern analyses are used to examine the subtle use behavior as clinicians navigated through the system’s user interface to perform different clinical tasks. Results show that the ways in which the system was actually used largely deviated from its originally designed purposes. Whenever misalignment occurred, users always developed workarounds that may diminish the value of using the system. In the third part of the study, pre- and post-implementation time motion data were collected. The comparison shows that the introduction of the system brought with dramatic influence on clinical workflow. The post-implementation clinical workspace became more fragmented, with more frequent task switches and less amounts of time spent on direct patient care. This workflow alteration and the associated sociotechnical integration issues may account for the acceptance behavior observed previously.

Recording & Additional Notes

Introducer: Rosalind Keith, Ph.D. student, School of Information