Machine-Learning

In the healthcare industry, misdiagnosis endangers patient safety and impacts more than 12 million American adults every year. According to the Society to Improve Diagnosis in Medicine (SIDM), a person dies every 9 minutes due to misdiagnosis, and it is the #1 reason for malpractice suits. Although the United States has a well-developed healthcare system with accomplished physicians, its very plausible that physicians are capable of making mistakes. The top three cognitive errors that lead to misdiagnosis, also referred to Dx error, include: mistakes gathering information, mistakes in analyzing information, and premature closure. However, these issues are beyond the scope of being solved at the indivdual level. In healthcare environments, physicians are stretched beyond their capacity; there are too many patients, too many symptoms to diagnose, and too little time for them to be 100% diagnostically reliable with each and every patient. Diagnosis is complex and challenging, and healthcare workers are human-beings, and thus fallible. Aside from cognitive errors, systemic racism and social disparities influence the diagnosis of patients. The University of Leeds in the UK found that women are 50% more likely to receive the wrong diagnosis after having a heart attack, compared to men. Additionally, a study published in the journal, Psychiatric Services, found that African-Americans with severe depression are more likely to be diagnosed as having schizophrenia. These results reveal serious changes need to be made to the decision-making process, in respect to diagnosing illness.