Current screening tools for bipolar disorder are poorly validated in the age group most at risk for developing the condition, say researchers.
The highest risk for bipolar disorder is present at age 15–25 years, and prompt treatment is crucial to optimize outcomes.
However, most screening tools have been validated in middle-aged populations with established bipolar disorder. In their review, Thomas Meyer (Newcastle University, UK) and colleagues found just 11 studies that assessed the performance of mania and/or depression screening tools in people at the peak age of risk.
“This finding alone is significant, as it indicates a serious mismatch between the expressed desire to improve the recognition of [bipolar disorder] cases or of individuals at increased risk of [bipolar disorder], and introduce treatment earlier versus the approach to validating any screening instruments,” writes the team in the Journal of Affective Disorders.
The studies examined seven scales: the Centre for Epidemiological Studies-Depression scale, the Bipolar Spectrum Diagnostic Scale, the General Behaviour Inventory (GBI), the Hypomania Checklist, the Hypomanic Personality Scale, and the original and adolescent versions of the Mood Disorders Questionnaire (MDQ).
Meyer et al say that the findings were heterogeneous and differed from those in middle-aged populations. “As such, it seems inappropriate to apply the published cut-off scores for these scales to studies of younger people,” they comment.
Also, the performance of the tools seemed to differ according to the setting. For example, the MDQ performed well in a clinical sample (sensitivity 74%; specificity 80%) but less so in a community setting (sensitivity 11%; specificity 93%).
This point “becomes especially relevant when the goal is identifying individuals at high risk versus cases with an as yet undiagnosed illness,” say the researchers.
By contrast, the 73-item GBI performed better in a community study (sensitivity 76%; specificity 99%) than in a clinical setting (sensitivity 29%; specificity 79%), although the team notes that any conclusions are “undermined considerably by the small sample sizes, some including as few as 13 participants with features of [bipolar disorder].”
In light of the “limited evidence on which to base any recommendations,” Meyer and team “would encourage those employing established scales to calculate optimal cut-off scores for different age groups in the recruited sample.”
They also suggest that researchers could combine two tools – one with high sensitivity and one with high specificity – to maximize case detection.
This article first appeared on ‘MedWire News’ on 9 January 2014.