ICD-10: The Death of Urosepsis and the Opportunity to Decrease Medical Errors

Posted on Tue, Oct 20, 2015
ICD-10: The Death of Urosepsis and the Opportunity to Decrease Medical Errors

By Michael Lozano, MD, FACEP

The other day while I was completing my online ICD-10 training, I came to a great realization. October 1, 2015, marks the date of the eradication of urosepsis.

We’ve all seen patients with this condition. Invariably, EMS brings them in from the nursing home. Common features include incontinence manifest by urine-soaked bed sheets, borderline low blood pressure, delirium and maybe fever – or rumors of a fever. Other associated factors of the classic presentation are underlying dementia and an incomplete history because the skilled nursing facility nurse “just received the patient” and was not fully familiar with them.

Over the years, the term urosepsis became sort of shorthand – ER jargon – for the type of patient described above. It was easy to blame the urine. After all, it was right there up in your nose. Moreover, it gave us an opportunity to admit the patient to the hospital without having to dig too deep in the differential diagnosis. It would be clear that the patient was sick and needed to be in the hospital. Whatever you wanted to call it was OK as long as the patient was not going back to the nursing home and did not linger too long in the ED. Our diagnostic heuristic was satisfied, and we could move on to the next case.

Urosepsis is gone now. It has not gone because of some breakthrough therapy; it’s relegated to the history of medicine because of ICD-10 and its requirement for diagnostic specificity. There is not a code for urosepsis in ICD-10 like there was in ICD-9. As we know, in the world of coding taxonomy, if there is no code for something it doesn’t exist.

An Opportunity to Decrease Medical Errors – and Clinician Biases

While your first reaction is to recoil against any additional documentation burdens, you would do better to embrace this as an opportunity to decrease the chance of committing a medical error in your practice. Really, I’m serious. The case of the sick geriatric patient in the emergency department is a veritable minefield of potential medical errors, and we should take this as a learning opportunity.

Cognitive errors in medicine are more about the decision-making process rather than a simple lack of knowledge. Berner and Graber’s 2008 study reported the rate of diagnostic error in the emergency department to be between 0.6% and 12%.[1] Thus, it’s reasonable to explore areas of error mitigation in emergency medicine. By understanding our biases and their contribution to cognitive errors, we can avoid these cognitive errors.

There are two types of decision-making approaches – intuitive and analytic. The intuitive approach (also known as reflexive) is an automatic one based on pattern recognition. This would be along the lines of a gunshot to chest plus large pneumothorax equals chest tube. The second type is the analytic, or problem-solving, approach. It takes a lot more brainwork to process type-2 decisions, so your brain naturally will want you to default to type-1 decision-making as much as possible.

As we mature in our profession, we become more capable in certain areas of practice. According to conscious competence theory, we evolve to the state of unconscious competence. At that point, we are applying a lot of type-1 decision making. True wisdom in emergency medicine is demonstrated when one knows when to switch from type-1 to type-2 problem solving. That’s the wisdom to pick out the zebra in the herd of horses.
Certain biases will have an effect on our type-2 thinking and may potentially derail our rational thought process. Anchoring bias occurs when you latch on to a particular diagnosis and will not diverge from it – often despite information that doesn’t support your initial impression. Think about your thought process when the nurse puts “drug seeker” on the chart, or when you get a sign-out that states, “Nothing to worry about on this one. I saw her two days ago for the same thing.”

Confirmation bias occurs when you filter information and disproportionally add weight to data that supports your initial diagnosis. Yes, we all unconsciously delude ourselves to a greater or lesser extent. The danger of both anchoring and confirmation bias is that we risk premature closure – shutting down internal debate, and flipping into the reflexive process of admitting or discharging. In the case of the “uroseptic” patient, our biases steer us in the direction of the “easy” diagnosis, and we run the risk of missing any one of the many differential diagnoses that would explain the constellation of signs and symptoms described in the first paragraph.

I would argue that ICD-10’s impetus to document with more descriptive specificity becomes an insurance policy against confirmation bias and premature closure. By taking away the reflex “easy button” of urosepsis, the intellectual challenge shifts to the discernment of where the patient sits on the continuum of SIRS to sepsis to severe sepsis to septic shock – if at all.

For so many years, our brains sought to pigeonhole what was effectively geriatric delirium into the small box of urosepsis. It was easy. It solved the nagging need to shift from type-2 to type-1 thinking. Once you went down that path, all you needed was a little fluid bolus (but not too much) and an antibiotic. You could then move on.

I hope that spending a little more time using the analytical side of our brain results in less misdiagnosis and better care for our patients.

Michael Lozano, MD, FACEP, has been an Executive Vice President for EmCare’s South Division since 2009. Prior to his current role, he worked for EmCare as the medical director and chairperson of the Department of Emergency Medicine at Northside Hospital in St. Petersburg, Fla. He also is the medical director for Hillsborough County Fire Rescue in Tampa, Fla. In his role as medical director for Florida’s Urban Search and Rescue System (U.S.A.R.) Task Force 3, Dr. Lozano has responded to several disaster scenes, including the aftermath of hurricanes Charley, Ivan and Katrina. He is a highly skilled physician with over 20 years of medical and leadership experience.

[1]Berner ES, Graber ML. Overconfidence as a cause of diagnostic error in medicine. Am J Med 2008 May;121(5 Suppl):S2-S23. 


Blog post currently doesn't have any comments.