In Reply Dr Ross proposes the adoption of software that interprets physician-selected data and provides guidance derived from high-quality information sources. Currently web-based differential diagnosis generators such as Isabel, VisualDx, and DXplain do that to some degree.1
At issue is the uptake of such assistance. As the recent JAMA Internal Medicine studies elegantly demonstrated, lack of time and overconfidence are major barriers to physicians’ seeking assistance from knowledge sources and decision support.2,3 It takes so much time to enter information once into the medical record that physicians are understandably disinclined to enter the same information again into another program (eg, PubMed, diagnostic support, Google). The key to harnessing any software program’s potential is to have it detect, extract, and interpret the relevant variables from the electronic medical record (and other inputs) without interrupting physician workflow.
Dhaliwal G. “Expert Opinion” Software for Medical Diagnosis and Treatment—Reply. JAMA Intern Med. 2014;174(4):639-640. doi:10.1001/jamainternmed.2013.13789