Computer diagnostics could offer opportunity for pathologists and clinical laboratory managers to add value to clinicians in diagnosing diseases
Efforts are intensifying to develop computer software that successfully emulates the skills of highly proficient diagnosticians. The motivation is increased pressure to reduce medical errors, including misdiagnosis. This is welcome news to many pathologists, who often see physicians ordering the wrong laboratory tests.
Diagnostic Superstar Helps Doctors Become Better Diagnosticians
“Getting better at diagnosis is as important to patient quality and safety as reducing medication errors, or eliminating wrong-site surgery,” stated diagnostician superstar Gurpreet S. Dhaliwal, M.D., Associate Professor of Clinical Medicine at the University of California San Francisco School of Medicine. Dhaliwal is recognized as one of the most skillful clinical diagnosticians in practice, the Times stated.
The unassuming 39-year-old wows audiences at conferences with his diagnostic talents. Beginning with information that could describe hundreds of diseases, he is fed additional bits of information about the case. Sooner than most, he comes up with a diagnosis. It is usually correct.
Dhaliwal splits his clinical practice between the wards of the San Francisco VA Medical Center and its emergency department. “Things can shift very quickly in the emergency room,” he observed in the Times story. “One challenge of this, whether you use a computer or your brain, is deciding what’s signal and what’s noise.”
Most Diagnostic Failures Are the Result of Limiting Possibilities
A 2005 study, identified “premature closure” as the single most common cause of diagnostic error. Premature closure is the failure to continue to consider reasonable alternatives after an initial diagnosis has been reached. The study findings showed that faulty or inadequate knowledge was rare.
According to Martin S. Kohn, M.D., Chief Medical Scientist for IBM Research, most physicians spend five hours or less per month reading medical literature, the Times reported. Dhaliwal, on the other hand, reads insatiably. Kohn’s research work includes the role of expert systems in the clinical decision process, including the use of the Watson supercomputer in healthcare.
Even at his level of diagnostic expertise, Dhaliwal sometimes uses a diagnostic checklist called Isabel. A former London money manager, Jason Maude, created the web-based diagnosis checklist system after his three-year-old daughter Isabel was misdiagnosed by her local hospital with consequences that were almost fatal.
Physicians failed to look beyond the little girl’s case of chicken pox to discover the presence of toxic shock syndrome and necrotizing fasciitis, according to the Isabel Healthcare website. Rather than suing the hospital, Maude and his wife created the program to help clinicians around the world with differential diagnosis.
“Low-frequency events are hard to put on the brain’s palette, and that’s part of Isabel’s strength,” Maude stated in the Times story. “It’s impossible for any one person to remember how each of those diseases presents, because each presents with a different pattern.”
Barriers to Adoption of Diagnostic Software
Dhaliwal offers some explanations for the slow adoption of diagnostic software in clinical settings. It can be difficult to integrate software into a busy work flow, he suggested in the Times account. Further, he observed that most doctors think they don’t need help, especially in routine cases.
“The root cause of diagnostic error is not thinking about what some of the possibilities might be,” stated Maude in a story published by InformationWeek. “The differential forces you to start thinking about it.”
Pathologists and clinical laboratory managers will want to watch this development because they have a role to play. They are experts in both diseases and how lab test technology is used to diagnose disease and guide treatment decisions. Therefore, laboratory professionals should have a useful role in helping to develop the algorithms used in software systems designed to aid physicians in the diagnosis of disease.
—Pamela Scherer McLeod