An unconscious bias in this EHR study?

Like many physicians, I struggle with looking at my patients while they are talking and getting the stories that they tell me transcribed as accurately and completely as possible. After I read the article, “EHR use and patient satisfaction: What we learned” by Farber, et al, (J Fam Pract. 2015;64:687-696), I was struck by something.

Of the 126 patients chosen for the research, the educational level breakdown included 75% with at least some college education and 28% with postgraduate education. A study performed by the National Center for Veterans Analysis and Statistics published in 2015 has different statistics.¹ Although a similar percentage had at least some college education, only 10.5% of the men and 12.4% of the women had postgraduate education.

In my practice, most of my patients who have worked with computers empathize with the amount of time that I spend looking at the screen. Those with less education are less agreeable. Since the patients were picked by their physicians to take part in the study, I wonder if there was an unconscious bias present.

Holly Leeds, MD
Auburn, Calif

1. National Center for Veterans Analysis and Statistics. Profile of Veterans: 2013. US Department of Veterans Affairs Web site. Available at: http://www.va.gov/vetdata/docs/SpecialReports/Profile_of_Veterans_2013.pdf. Accessed March 21, 2016.

Author's response:
Dr. Leeds brings up an interesting issue. It is possible that there is an unconscious bias on the part of physicians who participated in this study. Although the demographics are fairly similar to those that she cites, the veterans in our study were somewhat more educated.

If less well-educated subjects participated, this would make the data more impressive, in terms of less satisfaction with physicians who more readily focus their eyes on computer screens rather than on their patients. The fact that we did find this association is important for physicians who use EHR systems.

Neil J. Farber, MD, FACP
San Diego, Calif