Diagnostic Errors in the COVID-19 Era

Jaya Mallidi, MD, MHS


September 03, 2020

Editor's note: Some details have been changed to protect the patient's identity.

Jaya Mallidi, MD, MHS

It was an unusually quiet Saturday afternoon on call. I was sipping coffee and leisurely reading that day's echocardiograms. On one, ordered for an indication of "dyspnea," there was a moderate-sized pericardial effusion without tamponade physiology. I called the ordering hospitalist to tell him the results.

He mentioned that the patient was admitted the previous day as a "COVID rule-out." Mr Gonzales was a 55-year-old man with progressively worsening exertional shortness of breath and cough over the previous 2 weeks. A chest x-ray showed cardiomegaly, concerning for pericardial effusion. His vital signs were stable and he was admitted with a diagnosis of viral pericarditis and high suspicion for COVID-19. Rapid SARS-CoV-2 testing was negative, but another test was ordered given the high clinical suspicion.

The Patient in Person vs the EHR

I called the lab for the second COVID-19 test result but was told that it would not be available for several hours. With not much clinical activity going on, I decided to go see Mr Gonzales. I donned the necessary personal protective equipment and entered his room.

Mr Gonzales was propped up in the bed. A yellow mask covered his mouth but stopped short of covering his nostrils. His eyes looked tired, like those of a man who had not slept in days. I introduced myself and asked how he was doing. After a long pause, he shook his head and mumbled, "Not good." His voice was unusually hoarse. He could not lie flat comfortably. He could not walk 10 yards without dyspnea. I wondered whether the effusion needed to be tapped. Unusual to have this degree of symptoms without any tamponade physiology.

As I stood at his bedside, unsure of what to do, he raised his hands to the sides of his neck. "It's swollen, my neck," he said. That's when I noticed that his neck was somewhat thick, more than what one would expect for his sturdy build. The supraclavicular fossa (the indentation above the clavicular bone) was obliterated and full. On palpation, I felt the hard, irregular surface of the supraclavicular lymph nodes. Check for malignancy, I thought. I informed him that we would need to do a CT scan of his neck and chest to find out what is going on.

Diagnostic Errors

Diagnostic errors, either missed, inaccurate, or delayed, were a common source of potential patient harm even before this pandemic. The Institute of Medicine, in its report "Improving Diagnosis in Health Care," defined diagnostic error as the failure to establish an accurate and timely explanation of the patient's health problem or communicate that problem to the patient.

Cognitive shortcuts to aid in decision-making are important sources of diagnostic error in medicine. Traditionally the diagnostic process is iterative. Accuracy and timeliness are subjective and largely depend on the patient's clinical presentation. In the COVID-19 era, an accurate diagnosis is focused on ruling the disease in or out, and the rest of the diagnostic workup begins only after this. While these new processes are important from a public health standpoint, for an individual patient, especially one presenting with a non-COVID-19 illness, the virus has become a hurdle in the diagnostic process, already strained by cognitive biases and system-based errors (Figure).

Adapted from Improving Diagnosis in Health Care .

A recent perspective by two patient quality and safety experts, Tejal Gandhi and Hardeep Singh, describes several types of anticipated diagnostic errors in the COVID-19 era. With universal availability of testing, missing COVID-19 in a patient presenting to the ED is unlikely. The more relevant error at this phase of the pandemic is a delayed or missed non-COVID diagnosis. The following common clinical fallacies can lead to diagnostic errors, especially now.

Anchoring bias. This cognitive bias occurs when the decision-making is predominantly based on the initial information provided. With Mr Gonzales, the initial framing was toward COVID-19. It was such a strong anchor that even rapid negative testing did not result in a reevaluation of the diagnosis. In the past few months, I have personally witnessed and read case reports of common disorders such as acute myocardial infarction being missed because of anchoring to a COVID-19 diagnosis.

Availability bias. With the raging pandemic, we physicians have COVID-19 on our minds. All patients presenting with respiratory complaints are assumed to have COVID-19 unless proven otherwise, because this is the most readily available explanation.

Base rate neglect. The prevalence of SARS-CoV-2 has significant geographical variation. In an area with low prevalence, a patient who presents with respiratory symptoms has a higher probability of a non-COVID condition, such as pulmonary edema from myocardial infarction, bacterial pneumonia, bronchitis, COPD, or heart failure. However, the underlying population-based prevalence is often ignored in everyday clinical medicine.

Lack of Time

The tainted practice of modern medicine, where physicians review the EHR before meeting patients, as well as the note bloat from copy-paste problem lists, is a perfect set-up for diagnostic errors. To avoid errors and overcome our cognitive biases, we need to spend adequate time with patients, carefully integrating history, examination, and test results. In the fast-paced, chaotic world of modern medicine, where compensation rewards volume and not cognitive thinking, this needed time is missing.

Instead, we have burnout associated with the pressure to see more patients and generate more wRVUs. During a pandemic, there are new care processes, specific to each hospital, to rule out/rule in COVID-19, not to mention the rational fear for our own safety and the mental fatigue associated with that.

The risk for diagnostic errors is enhanced in this kind of environment. It can be difficult to measure this kind of error, and to date, we do not have objective data to definitively prove that there are more diagnostic errors in the COVID-19 era. I suspect, however, that the rate of diagnostic errors is higher than in pre-pandemic times. It's important for physicians to be extra vigilant during this period.

A Near Miss

The CT scan of Mr Gonzales' chest and neck showed a mass in his lung that extended to the mediastinal structures causing significant laryngeal edema (and his hoarse voice) and superior vena cava syndrome. An urgent biopsy found stage IV adenocarcinoma of the lung. Given the patient's significant respiratory symptoms, the oncologist initiated in-patient radiation treatment in addition to intravenous steroids. Later, Mr Gonzales asked me why no one did the CT scan earlier despite him complaining about his swollen neck. Cognitive bias; a near miss, I thought. "We had to make sure COVID was ruled out," I answered.

Jaya Mallidi is an interventional cardiologist in Santa Rosa, California. An ardent patient advocate, she writes opinion pieces using patient stories as context to highlight problems in the practice of modern-day medicine. In addition, she enjoys digital sketching and playing tennis.

Follow Jaya Mallidi on Twitter

Follow | Medscape Cardiology on Twitter

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.