Implementation of Physician Order Entry: User Satisfaction and Self-reported

Ob’ t’ Iec Ives: To evaluate user satisfaction, correlates of satisfaction, and selfreported usage patterns regarding physican order entry (POE) in one hospital. Design: Surveys were sent to physician and nurse POE users from medical and surgical services. Results: The users were generally satisfied with POE (mean = 5.07 on a 1 to 7 scale). The physicians were more satisfied than the nurses, and the medical staff were more satisfied than the surgical staff; satisfaction levels were acceptable (more than 3.50) even in the less satisfied groups. Satisfaction was highly correlated with perceptions about POE’s effects on productivity, ease of use, and speed. POE features directed at improving the quality of care were less strongly correlated with satisfaction. The physicians valued POE’s off-floor accessibility most, and the nurses valued legibility and accuracy of POE orders most. Some features, such as off-floor ordering, were perceived to be highly useful and reported to be frequently used by the physicians; while other features, such as “quick mode” ordering and personal order sets, received little self-reported use. Conclusions: Survey of POE users showed that satisfaction with POE was good. Satisfaction was more correlated with perceptions about POE’s effect on productivity than with POE’s effect on quality of care. Physicians and nurses constitute two very different types of users, underscoring the importance of involving both physicians and nonphysicians in POE development. The results suggest that development efforts should focus on improving system speed, adding on-line help, and emphasizing quality benefits of POE. n JAMIA. 1996;3:42-55. LEE ET AL., Physician Order Entry Affiliations of the authors: Department of Psychology, Harvard University, Cambridge, MA (FL); and The Center for Applied Medical Information Systems Research, Brigham and Women’s Hospital, Boston, MA (JMT, CDS, DWB). At the time this study was done, Dr. Lee was at Harvard University, Cambridge, MA. She is currently at the University of Michigan, Ann Arbor, Ml. Correspondence: Fiona Lee, PhD, The University of Michigan, School of Business Administration, 701 Tappan Street, Ann Arbor, MI 48109-1234. e-mail: [email protected] Reprints: David Bates, MD, Division of General Medicine and Primary Care, Brigham and Women’s Hospital, 75 Francis Street, Supported in part by the Knox Bequest, Department of Psychology, Harvard University, and by Research Grant ROlHS07107-01 from the Agency for Health Care Policy and Research, NIH. Boston, MA 02115. Received for publication: 5/22/95; accepted for publication: 8/28/95. Journal of the American Medical Informatics Association Volume 3 Number 1 Jan / Feb 1996 43 One important application of informational technologies in medicine is physician order entry (POE).‘,’ In POE, physicians enter orders directly into the computer. POE has a number of benefits such as ensuring legibility and allowing faster transmission of orders, but perhaps its greatest area of potential is the opportunity in offering physicians online information support.3 For example, when a physician orders digoxin, the system can check recent results of relevant tests such as serum potassium level, creatinine, and digoxin level, and displays the information on the screen. The system can also perform multiple checks on orders, such as alerting the physician if a patient has a known allergy to the drug being ordered. Providing online information with POE also has the potential to reduce costs: for example, the computer can notify physicians about redundant or low-yield test ordering.3-5 Despite the potential benefits of POE, many attempts at implementing it have failed or met with high levels of user resistance.‘z3 For example, implementation of POE at University of Virginia (UVA) Medical Center resulted in widespread discontent among physicians, and led to a boycott of POE by residents until changes were made.6,7 POE implementation efforts have stumbled for a variety of reasons, including lack of sensitivity to users’ suggestions, and the large workflow change that POE brings.6,7 To be successful, implementers should frequently monitor areas of satisfaction and dissatisfaction among users, and should understand as well as possible how POE is actually used.* This study uses a survey to: 1) evaluate users’ overall satisfaction with POE, 2) identify factors associated with satisfaction and dissatisfaction with POE, 3) measure users’ perceptions about the frequency they used specific POE features, and 4) measure users’ perceptions about the usefulness of specific POE features. This information may have implications for systems development, training, and how POE can be most effectively framed and “marketed” within a hospital.