What inspired you to develop a system to help clinicians understand diagnostic tests?
As a clinical pathology resident, a bright internal medicine resident asked me a simple lab test selection question; he didn’t know how to further evaluate an abnormally prolonged clotting test known as the PTT. It surprised me. I realized I had knowledge about making a diagnosis that he, and most residents in other specialties, didn’t have. I asked him what he would have done if we hadn’t spoken. He told me he would do what every other doctor did – guess which test to pick and guess what the results meant!
So you realized something had to be done, but then the first barriers came up?
Yes. I tried to set up a program to provide patient-specific, expert-driven narrative interpretations of complex clinical lab evaluations while I was a resident. Unfortunately, I could not interest the faculty in participating. I was told by one assistant professor that becoming an associate professor was required, or he would lose his job. And “you do it not by taking care of patients, but by publishing papers!” As a trainee, that was quite a shocking revelation to hear from a mentor.
Undeterred, when I got my first job at the University of Pennsylvania, I created a page to be inserted into patient charts, with blanks for lab test results and a box at the bottom for our interpretations. It was incredibly well received, and patients started to get their diagnoses quickly, rather than needing multiple rounds of tests. But a senior hematologist told me to stop because “my hematology fellows are not seeing enough patients, so my revenue is down.” I knew that was a game I couldn’t win.
I then moved to Massachusetts General as director of clinical laboratories. We initiated interpretations for complex coagulation cases and the program became a source of national and international attention because of its clinical value in establishing rapid, accurate diagnoses. Oddly enough, the interpretations in the chart led to patient referrals to pathology from other doctors in the area (most hematologist-oncologist doctors!), and we developed a widely respected clinical practice for patients with bleeding and clotting disorders. Some patients were very confused, and asked if they were referred in preparation for an autopsy! I assumed they were kidding.
Even then, someone in the hospital’s hematology–oncology division told me that he didn’t need my interpretation. I told him that the other 2,000 doctors in our healthcare system did. He crumpled up the paper and threw it at me.
But you still persisted?
Absolutely. I realized the higher you are in an organization, the fewer people there are to kill your innovations – and the more people there are who are willing to believe they’re worth a try. When I went to a higher position at Vanderbilt, we initiated the same service, and it developed a name – the diagnostic management team or DMT.
DMTs were created for coagulation, transfusion medicine, microbiology, hematopathology, and a few other clinical areas, and each claimed a triumph in improved patient care and decreased cost. Papers were written, and I was asked to speak about the DMT in at least 50 major institutions and many societies. It became abundantly clear that, with thousands of diagnostic tests available and the cost and complexity associated with them, diagnostic experts had to become involved. If they didn’t, somebody was going to get hurt.
Why has nothing changed?
The biggest reason? Financial incentives in the United States. According to statistics from the College of American Pathologists, an average pathologist works about 48 hours per week, makes around $200,000–400,000, and does work almost exclusively in anatomic pathology. There is substantial payment for anatomic pathology services, but almost no payment for clinical pathology advice on test selection and result interpretation. For decades, this misalignment of incentives has, expectedly, greatly minimized the number of experts in clinical pathology disciplines. But the external environment is changing and, thanks to our work with the Institute of Medicine, diagnostic error is no longer flying under the radar.
I’m now at the University of Texas, where we are hoping to create a state-wide group of experts in all major areas of clinical and anatomic pathology. We plan to have DMTs for dozens of clinical areas to bring diagnostic experts electronically to the bedsides of all 26+ million patients. The endgame is better patient outcomes. We shouldn’t allow our patients to die because there is no expert in their immediate environment. Once this is available all over the world, I will have completed the task that started with a simple PTT evaluation query in 1984.
Adopting the DMT model won’t be easy for pathologists, though, especially those who have been working with a different focus for years. Somebody has to provide the example, and we plan to show others how it’s done. It’s going to take the next generation of residents who are trained to provide diagnostic consultations covering all areas of pathology to effect the major change. And I haven’t lost my enthusiasm. It may take a while until the change comes, but I’m going to do this until I can’t do it anymore.
Grandpa was a sulfur miner, and dad a barber – it’s hard for me to complain.