When a potassium of 8.8 comes back on a patient who looks well and the one before was 4.1, no experienced physician treats that number first. The first question is whether the value is correct, and the second is how the specimen was drawn and transported. A tourniquet left on too long, a small-bore needle, a slow draw, and red cells shear and spill potassium into the tube. The result is accurate for the tube and wrong for the patient.
Every clinician knows to do this. We are trained to distrust a value that does not fit the patient, and we are trained to ask what the chain of custody did to it before we act.
Health care leaders read hemolyzed specimens every week, and they arrive as operational dashboards. I call the path those numbers travel the data proposition. A clinician observes something at the bedside. It is captured in the record, or captured partially, or not captured at all. What is captured is mapped to a discrete element, or it lands in unstructured text where no query will find it. The element is pulled by an analytics query, a vendor’s quality engine, or a federal reporting interface, each of which may read it differently than the clinician meant it. Then it is aggregated, calculated, and rendered as a figure on a slide, a payment adjustment, or a recommendation. By the time an executive sees the number, it has been through six or eight transformations, and it is reliable only if every step was performed correctly.
In most organizations, that path has never been followed end to end and audited. Each step is owned by someone who knows their own step well and the steps on either side of it poorly. The clinician does not know how what they entered was mapped. The analyst does not know what the clinician was thinking. The dashboard designer does not know what the analyst’s math is doing. The executive does not know any of it. So the number gets read as if it were a calibrated laboratory value, when it is closer to a many-times-contaminated specimen.
Real physician voices, twice a week
Free, and one click to unsubscribe.
You can recognize the artifacts in your own institution without any new analysis:
- A length of stay built on a discharge time the unit enters at the end of a shift
- A bundle compliance rate that measures whether a box was checked rather than whether the bundle was delivered
- An internal readmission count that cannot see the hospital across the street
- A case-mix index that reports your patients were healthier than they were, because the documentation that would have shown otherwise sits in text no query reads
The note itself is the first link. A record that is half copied forward has a currency problem, because no one can tell which findings are today’s. It has a correctness problem, because a resolved condition carried forward is now wrong. And it has a concordance problem, because the copied text and the current results are drifting apart. Documentation templates were built to capture billing rather than clinical thinking, and what they produce is a long note that tells the next clinician almost nothing. When the source degrades, everything drawing from it inherits the corruption.
A survey of how organizations assess the quality of their own record data found little consistency and little generalizability, and recommended validated systematic methods over improvisation. Most organizations are still improvising.
What a physician does with a suspect value is the whole method. You do not re-treat the patient on the new number, you do not split the difference between the three draws, and you do not launch an investigation. You note that the specimen was hemolyzed, identify the value as an artifact, and continue the original plan. Then, in a well-run service, you note the recurring problem with that phlebotomy team and ask whether a process needs fixing. Organizations almost never take that last step with their own data, and it is the step that would stop the next bad decision rather than this one.
Two questions have answers inside systems your organization already owns:
- Pick one number on your board packet. Who owns each step between the bedside observation and that number, and has anyone ever audited the transformations against the chart?
- When a figure on a dashboard disagrees with what the clinicians on that unit say is happening, which one does your organization investigate?
An institution that investigates the clinicians has decided the dashboard is a calibrated instrument. It is not. It is a specimen, and someone has to be qualified to look at the tube.
Thomas Powell is a physician executive, a career chief medical information officer (CMIO), and the founder of CMIO, LLC, which provides interim physician informatics leadership and advisory work to health systems. He has held the CMIO seat in academic, community, pediatric, and integrated delivery systems, and in most of them he was the first.
At the University of Arkansas for Medical Sciences (UAMS), he was CMIO and associate vice chancellor and led the move to Epic across the hospital and 43 clinical groups. At Aspirus Health, he was vice president and CMIO of an 18-hospital system with 1,300 employed physicians and advanced practice clinicians, and at Miami Children’s Hospital, now Nicklaus Children’s, he took 38 pediatric subspecialties from paper to Cerner. He served the State of Arkansas as medical and executive officer of the Office of Health Information Technology and the state health information exchange.
He trained at Duke University as a National Library of Medicine fellow in clinical informatics under Ed Hammond, and at UAMS he created Arkansas’s first clinical informatics master’s and doctoral program and helped build the state’s only clinical informatics board fellowship. He writes and speaks on applied clinical informatics, the governance of clinical AI, electronic health record usability, and the physician executive role.
He is the author of The Diagnostician: American Healthcare Is Sick, Misdiagnosed, and Getting Worse; It’s Time to Call the Doctor, published November 3, 2026, which applies the clinical diagnostic method to the health care organization itself. He shares updates on X.
