A new JAMA Internal Medicine commentary article highlighted by PopTort via HuffPost makes for interesting but depressing reading in highlighting the volume of medical errors that may be out there. Set out below are excerpts from the article – note especially the words in bold; the numbers are footnote references, which are excluded here but available in the free full text version.
Why are there so many diagnostic errors? In some sense, the answer is simple: medical diagnosis is extremely difficult. It may not be as difficult as accurately predicting a specific weather forecast 6 months in advance, but it is one of the toughest tasks human minds routinely face. Diagnosing can be messy, and scientific understanding is imperfect. Decisions must be made with limited time and information under conditions of uncertainty, often with inadequate experience or expertise in diagnosing a given symptom or disease. Consistent patient follow-up with feedback on diagnostic performance is usually lacking or biased away from detecting diagnostic errors, creating a serious barrier to ongoing skills improvement. These facts, however, should not dissuade us from trying to reduce misdiagnosis-related harm.
Through scientific research, important lessons have been learned about diagnostic errors that are reinforced by results from the study by Singh et al.5 Most diagnostic errors are linked to clear defects in bedside history taking, physical examination, and test ordering7; the same was found in these authors’ research.5 These process errors are associated with cognitive errors dominated by failure to consider the correct diagnosis,8 compatible with the finding that providers failed to document a differential diagnosis at the initial visit in more than 80% of misdiagnosed cases.5 Diagnostic errors most commonly occur in the context of atypical presentations or nonspecific symptoms.8 Singh et al5 found a high frequency of symptoms apparently “unrelated” to the final diagnosis. Regardless of whether these symptoms were unusual disease manifestations or irrelevant symptoms in patients with discoverable disorders, this finding comports with prior ones that difficult presentations are more likely to result in diagnostic errors.8 (emphasis added)