Misdiagnosing bipolar disorder--do clinicians show heuristic biases?

J Affect Disord. 2011 May;130(3):405-12. doi: 10.1016/j.jad.2010.10.036. Epub 2010 Nov 18.


Background: Bipolar disorders (BD) are often misdiagnosed. Clinicians seem to use heuristics instead of following the recommendations of diagnostic manuals. Bruchmüller and Meyer (2009) suggest that 'reduced sleep' is a prototypic criterion that increases the likelihood of a bipolar diagnosis. This study examines if this criterion specifically elevates the likelihood of a bipolar diagnosis or if the finding of the study mentioned above is rather due to the total number of criteria. Furthermore, we want to replicate the finding that patients offering a causal explanation for their manic symptoms are misdiagnosed more often. Additionally, we examine therapeutic attributes that might influence diagnostic decisions as well as treatment consequences following a (mis-)diagnosis.

Methods: 204 Psychotherapists were presented with a case vignette describing someone with a BD and were asked to make a diagnosis. Symptoms and the total number of criteria varied systematically within the vignettes but each still fulfilled enough diagnostic criteria to be diagnosed as bipolar.

Results: Almost 60% of the clinicians made misdiagnoses. A correct diagnosis did not depend on the specific criterion of 'reduced sleep' but on the total number of criteria. The causal explanation as well as therapeutic attributes did not significantly influence diagnostic decisions. However, the study showed that a misdiagnosis can lead to severe consequences concerning the treatment recommended by clinicians.

Limitations: The validity of case vignettes is discussible.

Conclusions: It seems as if specific symptoms might not be of so much relevance as assumed. Instead, clinicians seem to follow the additive model when making diagnoses.

MeSH terms

  • Bipolar Disorder / diagnosis*
  • Bipolar Disorder / psychology
  • Bipolar Disorder / therapy
  • Diagnosis, Differential
  • Diagnostic Errors* / statistics & numerical data
  • Humans
  • Physicians*
  • Psychiatric Status Rating Scales / standards
  • Sleep