Tackling The Prevalence Of False Diagnoses In American Medicine
By Darren Schulte, CEO, Apixio
Last month, the Institute of Medicine (IOM) released a report which declared half of all Americans will experience a false diagnosis in their lifetime. Think about it: one of every two people in the U.S. will receive an incorrect diagnosis. The report also noted that diagnosis errors are implicated in one out of every ten patient deaths, and that they account for up to 17 percent of hospital adverse events.
Diagnosis errors have historically received little attention because the data simply doesn’t exist to quantify the full scale of the problem. Clinicians often don’t know that a mistake has occurred until there is either a second opinion or a tragic event that occurs from not getting the diagnosis right the first time.
The IOM states some keys for better diagnosis imclude: more effective teamwork among clinicians and patients; enhanced training and education for healthcare providers; a payment system that rewards time spent for getting to the bottom of a patient problem; and a dedicated focus on new research. These recommendations are well-put and address the structural issues underlying false diagnosis. I would like to expand upon them by offering, from my clinical experience, three steps that can be taken to improve diagnosis accuracy.
The first step in providing a correct diagnosis is to listen carefully to the patient and document symptoms. Unfortunately, though this seems simple, there is less and less of this happening today. As a recent study on diagnostic errors published in the Journal of the American Medical Association found, “Most process breakdowns were related to the clinical encounter, wherein practitioners are almost always pressed for time to make decisions.”
With pressures building for primary care providers to see more patients each day coupled with the burden to follow regulatory and billing requirements, the voice of the patient is getting lost. Without time to methodically consider patients’ conditions, physicians fall back on their individual experiences and biases. “Common things are common.” For example, all coughs and fevers are thought to be infectious respiratory diseases. Some could be lung cancer.
Diseases tend to present in similar ways. Fever is a common response to infection – it doesn’t point to any one disease by itself. Asking questions to get to specific aspects of illness is critical. Doctors are taught to think of many diseases which could explain the collection of patient symptoms, exam findings, and test results. This is called the “differential diagnosis.” It is important to tease apart the presenting symptoms and signs to arrive at the diagnosis. Inferring a diagnosis from limited information will lead to an incorrect conclusion. Through careful deductive reasoning, a physician can arrive at a correct diagnosis.
The second step in providing accurate diagnosis is ensuring flawless follow-up. Twenty-five percent of lab tests or radiology exams ordered are delayed, incorrect, or never get to the ordering physician. Presumably a good number of these tests were intended to help diagnose a patient’s condition.
Lastly, technology may be able to play a helpful role with diagnosis. If the patient story is well recorded and the entire medical chart is available, a cognitive computing system can provide a list of likely diagnoses for the physician to consider. This technology can also suggest tests to help narrow the differential diagnoses if an answer is not already apparent. With enough information, the correct diagnosis can be reached with the assistance of this type of technology. The physician is not being replaced; the intelligent computer is an assistant.
Aside from technology, the real key to the getting the diagnosis correct is giving the providers the time and space to listen to and process the patient story; there are no shortcuts here. We should eliminate all non-clinical activities from the daily routine of the physician, such as pharmacy refills, filling out forms and orders, and coding. And we should provide incentives in the payment system for primary care physicians to take time with their patients. An extra ten minutes during a visit, and access to the entire patient record, can save countless expensive tests and help prevent an incorrect diagnosis.
About The Author
Darren Schulte, MD, has over eleven years of industry experience in healthcare analytics and technology. Darren served as the Chief Medical Officer and President of Apixio prior to being appointed CEO in 2014. Before joining Apixio, Darren served in executive leadership roles at Alere, Anvita Health, and Resolution Health. Darren co-developed 25 clinical measures endorsed by the National Quality Forum to assess ambulatory care quality measurement using electronic data. Darren is a nationally recognized speaker on the topics of healthcare analytics and quality improvement. Darren received his BS degree from Berkeley, MPP degree from Harvard, and MD from Stanford. He trained in Internal Medicine at UCSF, and is co-author of one U.S. patent.