An Interview with Peter Weir, MD, MPH

Peter WeirARUP Laboratories' workplace clinic serves roughly 4,500 employees, spouses, and dependents, and takes full responsibility for an individual's healthcare, from mental and physical health to disease prevention and chronic disease management. Patients of the clinic include newborns, as well as patients over the age of 70. Seven health care practitioners—two family medicine physicians, four mid-level providers, three physician assistants, one nurse practitioner, and a half-time clinical pharmacist—attend to the needs of the entire patient population. Peter Weir, MD, MPH, is the medical director of the clinic and assistant clinical professor in the University of Utah Department of Family and Preventive Medicine in Salt Lake City. 

Here Dr. Weir gives his perspective on diagnostic errors.

Brian Jackson, MD, MS, of the Patient Safety Focus Editorial Advisory Board, conducted this interview.

Please describe your clinical responsibilities and practice setting.
My role in the clinic is to see my own patient panel, as well as to supervise the medical care of all patients seen in our clinic. I'm fortunate to work with partners who are very competent and knowledgeable within their own area of expertise: pediatrics, women's health, and sports medicine.

So would you describe your clinic as a medical home, in other words a team-based healthcare delivery model that provides comprehensive and continuous medical care to patients with the goal of obtaining maximized health outcomes?
Yes, that's what we're trying to accomplish.

What are some of the challenges that you encounter in diagnostic testing?
Every time I order a test of any kind, I'm very aware of the potential for ordering the wrong test for the patient's condition, as well as the fact that even correctly ordered tests have weaknesses, for example false negatives and false positives. Ordering the incorrect test happens far more often than healthcare professionals probably realize. For example, in our clinic, many providers including me were ordering the serum H. pylori antibody test that gives very limited information. We later learned how much more appropriate the H. pylori breath test and stool antigen test are for determining if a patient has an active infection.

At ARUP, we have an unusual patient population in that they tend to like being tested and often ask us to test for things that may not be clinically indicated. An example would be a patient who wants a test for a biomarker to screen for cancer even though the individual has no clinical indications. Ordering a cancer biomarker in this situation can lead to panic and a wild goose chase if the result comes back elevated.

Another common problem is dealing with unexpected results. I've learned to repeat a lab or study if it doesn't fit the clinical picture. Not infrequently, the repeat comes back with a normal result and avoids wasting a lot of time and resources. For example, I remember monitoring serial hemoglobins on a hospitalized patient with suspected upper gastrointestinal bleeding to determine if the patient would need an immediate transfusion and/or an urgent EGD to look for the source of bleeding. One hemoglobin result came back 50% lower than the previous one.  I rushed to the patient's bed expecting to find him very unstable and pale, yet he looked no different from when I last saw him. I ordered a CBC to confirm the result and then watched the phlebotomist as he drew the blood from a vein that the patient's IV fluids were connected to. I realized in an instant that the last blood sample was diluted 1:1 with IV fluids, which led to the erroneous result.

Do you encounter diagnostic testing issues that are specific to mid-level providers?
I have found that inexperienced clinicians, not necessarily mid-level providers, tend to over-order tests when they are uncomfortable with a clinical situation. The problem they run into, however, is the more tests that are ordered, the more interpretation of results that is needed. For example, consider a patient with unusual joint pain. A less experienced provider might begin the workup with a panel of rheumatologic lab tests rather than carefully taking a history, doing a complete physical exam, and then targeting the appropriate laboratory work-up. Ordering panels of lab tests that are not well thought-out can generate misleading, and sometimes conflicting, results, and leads to confusion, unnecessary referrals, and patient anxiety.

Can you think of other examples of patient harm resulting from diagnostic testing? 
The one that first comes to mind is the PSA test. I have followed the prostate cancer screening controversy for the past 12 years. In the late 1990s, many of us were concerned that there was no data to prove that mortality rates were decreasing despite the huge increase in prostate cancer detection. The worry was that we were catching a whole bunch of very slow, indolent tumors that probably would never have caused any clinical manifestation. In the last few years, large randomized controlled trials have confirmed that fear: PSA screening in asymptomatic men has not significantly reduced mortality. I am now reluctant to order PSA screening in men for fear that I will be doing the patient more harm than good.  

If you had a magic wand to wave over the clinical laboratory, what would you change?
I would somehow bring the expertise from the clinical laboratory into our clinic. I am surrounded by physicians and scientists who have an exceptional knowledge base that I wish I could tap into at the point-of-care. No doubt, every provider has limits to his/her own knowledge, and collaboration with colleagues often leads to better care for patients.