The Challenge of Screening for Bipolar Disorder

Reliable and valid self-administered screening tests for depression and many of the anxiety disorders have been around for many years.  However, development of a good screening instrument for bipolar affective disorder has been disappointing.  Although several instruments have been proposed as reasonable for clinical care, there is limited psychometric support for widespread utilization.

The Mood Disorder Questionnaire (MDQ) is one of the most widely studied of the questionnaires that screen for bipolar disorder.  Although, it has some evidence of validity when tested in patients with well-established mood disorders, it has received very limited studies in more challenging screening populations, i.e. general psychiatric outpatient clinics or primary care clinics. 

Dr.  Mark Zimmerman and colleagues from Rhode Island Hospital and Brown Medical School recently published an analysis of the reliability and validity of the MDQ in a series of 752 psychiatric outpatients.  Dr. Zimmerman has been conducting tremendous research in the psychometrics and diagnosis of a variety of mental disorder for over 20 years.  I trust his work and given that he did not develop the MDQ, I think he is likely to be unbiased in this type of research.

Sensitivity and specificity are two key components of a good screening testing.  Unfortunately as one increases, the other decreases and so determining a cut point for screening test is a compromise between false negatives and false positives.   The sensitivity for a screening test is commonly set for 90% meaning that 90% of the population with the disorder (i.e. bipolar disorder) would screen positive with the screening test.

In the Brown University study of the MDQ in a general psychiatric outpatient population:

• Several cut points were  analyzed for sensitivity and specificity of bipolar disorder as defined in a comprehensive DSM-IV interview all outpatients received
• A cut point that provided a 90% sensitivity was defined and implemented (this cut point resulted in a 61% specificity)
• Unfortunately the positive predictive value using this cut point was only 22.1%--meaning of all those that had a positive screen, only 22% in fact had bipolar disorder

The authors note that the MPQ (and all other bipolar disorder screening tests) do not appear to be add sufficient value to general psychiatric practice.  They note that competent clinicians asking the diagnostic criteria are unlikely to find benefit in routine clinical practice.  There may be some slight benefit in that those screening negative may require less attention in further assessment of bipolar disorder risk.  But this minimal benefit is probably offset by the time and energy needed to administer and score the questionnaire.

The authors also note that for primary care physicians with little psychiatric training, the MPQ may aid in identifying patients at risk for bipolar disorder who may need referral for further evaluation.  Primary care screening for bipolar disorder should occur in the context of collection of screening test data for the more common conditions of depression, anxiety disorder and substance dependence.

Photo of St. Louis Cardinal Albert Pujols batting in spring training game against the Minnesota Twins in 2011 in Jupiter, Florida.

Zimmerman M, Galione JN, Ruggero CJ, Chelminski I, Dalrymple K, & Young D (2011). Are screening scales for bipolar disorder good enough to be used in clinical practice? Comprehensive psychiatry PMID: 21406301