International Journal of Bipolar Disorders
Bipolar disorder (BD) is a complex and chronic psychiatric disorder associated with severe functioning difficulties. Typically marked by recurrent episodes of (hypo)mania and depression, and symptom overlap with several other psychiatric disorders (e.g., major depressive disorder, schizoaffective disorder, borderline personality disorder), accurate diagnosis presents a unique clinical challenge. Indeed, individuals with BD report an average of 9 years from initial presentation for treatment to an accurate diagnosis (Hirschfeld et al. 2003). At the same time, mounting evidence suggests that a substantial proportion of individuals previously diagnosed with BD may fail to meet formal diagnostic criteria upon structured interview, leading to false-positive cases in addition to the false-negative cases encountered in routine care (Zimmerman et al. 2008). Yet even when an accurate BD diagnosis is obtained, it remains difficult to correctly identify BD subtypes. Clinically, incorrect diagnosis may lead to delays in the delivery of appropriate, evidence-based care. From a research perspective, misclassification of individuals into diagnostic groups, for purposes of group comparison or for evaluation of novel treatment effects (e.g., Sachs et al. 2003), may bias or otherwise undermine the validity of research findings.
The challenges described above underscore the importance of accurate diagnosis and detection in BD. One approach to enhance diagnostic accuracy in BD research is through the establishment and reporting of interrater reliability (IRR). Surprisingly, there are no published guidelines describing this process. We discuss the importance of IRR, briefly note common features and variations, and suggest steps moving forward including greater transparency to facilitate replicability of practices.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Gruber, June and Weinstock, Lauren M, "Interrater reliability in bipolar disorder research: current practices and suggestions for enhancing the best practices." (2018). Psychology and Neuroscience Faculty Contributions. 30.