60% of America’s annual deaths, 1.7 million deaths are due to medical error! Do I have your attention? Those numbers are based on a study [1] of course and are not so much blown out of proportion as just wrong. Extrapolating autopsy findings to hospital deaths peg the deaths from medical error at 40 to 80,000, not the numbers we want to see, but perhaps more realistic and certainly demonstrating the wide range and fuzziness of the numbers. A new study in Diagnosis tries to determine “what diseases account for the majority of serious harms” based on malpractice claims data. A word before continuing, not every iatrogenic death, one due to physicians or healthcare systems, results in lawsuits and while the data used covers about 25% of all cases brought in the US, we still are looking at a biased sample.
The researchers made use of a closed claims database covering nine years of suits focusing on the top three categories associated with malpractice actions; vascular events, i.e., stroke, heart attack, infections like meningitis or sepsis, and cancers. They further refined their focus to diagnostic errors, not treatment errors [2]; and stratified harms following a standard severity of injury scale – "high severity injuries" included significant permanent injuries like loss of a limb, permanent major injury like brain damage, permanent grave injury like the need for lifelong care, and death.
What’s Up Doc?
- Diagnostic errors accounted for 34% (#1) in high severity injuries and 28% (#1 again) of payouts
- The top three disease groupings accounted for 61% of diagnostic errors and 67% of payouts for diagnostic errors
- Within each disease group, you can further identify specific diseases, like lung or breast cancer in the cancer group, or sepsis and pneumonia in the infection group, or stroke or heart attacks in the vascular group. In each group, five specific diseases accounted for 50% of the problems.
- 51% of errors were due to primary care providers, including hospitalists and Emergency Department physicians consistent with the location of the events. Medical, surgical, and radiology specialists accounted for the remaining half in about equal numbers.
There were notable variations in the data
- Diagnostic errors about cancer were more frequent in office settings, while vascular and infectious misdiagnosis was more frequent in Emergency Departments and hospitals. The difference is most likely because of the diseases underlying tempo, infections, and vascular problems are urgent or emergent, with timelines measured in hours; cancer, on the other hand, is diagnosed over a week or so and is more often an outpatient activity.
- Deaths figured more prominently in vascular errors, more than 2 fold likelier than disability with the exception of stroke where the outcomes and ratio were reversed. Again, a reflection of the underlying top 5 diseases' clinical course. Deaths with 1.5 more common in the case of infections, with exceptions for spinal abscess and meningitis where disability was higher. For cancers, disability was the usual outcome, 1.4 fold as likely, with an exception for lung cancer, which has a weak therapeutic response in most instances.
- Misdiagnosis of infection was greater at either end of the age spectrum consistent with the increased incidence of infections in these age groups.
The Why of Diagnostic Error
“Serious harms are disproportionately due to failures in clinical judgment, rather than problem with communication or closing the loop on test results…”
In more than 85% of cases, the underlying cause was “clinical judgment.” More specifically, failure or delay in ordering a test or consultation, too narrow a consideration of possible diagnosis, failure to appreciate or reconcile relevant symptoms or tests, or test misinterpretation. Communication and system issues were distant 2ndand 3rdcauses but were more critical in dropping the diagnostic ball for cancer than the other two categories.
What we have here is a failure to think both inside and outside the box. It is about early closure, where you reach a clinical conclusion too quickly. The diagnostic net is not broad enough, or some symptoms or tests are thought to be anomalies, not red flags of error.
Speed is required for scaling services and generating profitability. You need to diagnose, treat, and move on. As more and more physicians become employees or are otherwise enslaved on the clock of productivity measured by “patient throughput,” I suspect that these types of diagnostic errors might increase. Famed physician, Paul Simon may have said it best. “Slow down, you move too fast.”
[1] Types and origins of diagnostic errors in primary care settings JAMA Int Med DOI: 10.1001/jamainternmed.2013.2777
[2] The overall ranking was surgical errors 28%, medical errors 23%, diagnostic errors 21%
Source: Serious misdiagnosis-related harms in malpractice claims: The ‘Big Three’ – vascular events, infections, and cancers Diagnosis DOI: 10.1515/dx-2019-0019
