This weekend I read an excellent article in Clinical Infectious Diseases on recognizing cognitive biases in clinical reasoning.Studies suggest that diagnostic errors occur in 15% of cases across all medical disciplines. Common diagnostic errors include anchoring (locking onto a diagnosis without considering alternatives), availability (labeling diagnoses based what has been recently seen), heuristics (mental short cuts or "rules of thumb''), overconfidence bias, premature closure and outcome bias.
The solution lies in metacognition- increasing physician awareness of their own thought processes or "thinking about thinking.''
As suggested by Groopman and Hartzband, three simple questions should be employed at the bedside.
- What else could this be? Give consideration to alternative diagnoses to avoid premature closure.
- Is there something that does not fit? Make sure that the clinical data truly fits with the presumed diagnosis.
- Is there more than one diagnosis?
