As self-created electronic physician documentation replaces transcribed clinical documentation, what are the implications of that shift for physician workflow, efficiency, and clinical effectiveness, and ultimately, for care quality? Those questions were explored by a panel of clinical informaticists, all of whom have engaged in recent research and study of some of these issues. That discussion took place during a Nov. 6 session at the AMIA 2012 Symposium, currently being held at the Chicago Hilton & Towers in downtown Chicago, and sponsored by the American Medical Informatics Association. Not surprisingly, one of the big topics of discussion was the oft-discussed tension between supporting a rich narrative within the patient record, and facilitating EHR-focused efficiency.

Peter J. Embi, M.D., MS, associate professor in the Division of Rheumatology & Immunology and in Biomedical Informatics at The Ohio State University, led off a discussion with three other experts: Charlene Weir, Ph.D., R.N., associate research professor in the Hartford Center of Geriatric Nursing Excellence at the University of Utah; Kenric Hammond, M.D., clinical associate professor at the VA Puget Sound Health Center in Tacoma, Wash.; and S. Trent Rosenbloom, M.D., M.P.H., associate professor of biomedical informatics, associate professor of internal medicine and pediatrics, and associate professor of nursing, at Vanderbilt University in Nashville.

Making use of the term “computerized provider documentation,” or CPD, Hammond underscored the fundamental tension between the administrative, analytical, and performance improvement uses of physician documentation, and its direct patient care uses. “The strength of CPD is that it improves communication,” Hammond said. “The vulnerability is that CPD takes over. I’ve heard estimates of up to 25 percent of physicians’ time going to documentation. But a balance needs to be struck between the administrative uses, which are essential—otherwise, the organization dies—and the clinical uses—otherwise the patient dies. So my plea,” he said, “is, ecosystem members unite! Clinicians, administrators, and information scientists need to work together to characterize and understand the ecosystem and understand everyone’s needs.”

Meanwhile, Rosenbloom briefly described several studies at Vanderbilt in the past few years that have examined documentation issues. In one small study, he said, Vanderbilt researchers uncovered five core factors that were linked with physician satisfaction with electronic documentation systems. These were efficiency; accessibility to notes within the physician workflow; a balance between the structure needed to ensure accuracy and completeness while at the same time supporting descriptive expressivity; document quality; and physicians’ desire for the note to support improved patient care quality.

In that study, “We found that some of these factors were in tension with each other, particularly accuracy versus efficiency, and completeness versus efficiency,” Rosenbloom noted.

In a different study of physicians at Vanderbilt, Rosenbloom reported, he and his colleagues found a considerable divergence between physician adoption of narrative text-based documentation systems, which were far more widely adopted, and structured documentation systems, which were “very poorly adopted.” Significantly, he noted, “We learned that champions really influenced which modes were used” in different settings. “And different tools found different niches. A simple web form might be widely used for intake in the ED’s intake triage, for example.

The discussion that ensued among the four presenters, and in a subsequent question-and-answer session that became both substantive and highly granular, served to underscore the complexities involved in moving forward to optimize physician documentation processes. And though no issues were “resolved,” the session allowed for the airing of many nuanced questions and arguments in an area of intense interest and concern to all healthcare informaticists.