Factor structure and reliability of the Italian adaptation of the Hypomania Check List-32, second revision (HCL-32-R2)

Objective To assess the psychometric properties of the Italian adaptation of the Hypomania-Check-List 32-item, second revision (HCL-32-R2) for the detection of bipolarity in major depressive disorder (MDD) treatment-seeking outpatients. Methods A back-to-back Italian adaption of the "Bipolar Disorders: Improving Diagnosis, Guidance, and Education" English module of the HCL-32-R2 was administered between March 2013 and October 2014 across twelve collaborating sites in Italy. Diagnostic and Statistical Manual Fourth edition (DSM-IV) diagnoses were made adopting the mini-international neuropsychiatric interview, using bipolar disorder (BD) patients as controls. Results In our sample (n=441, of whom, BD-I=68; BD-II=117; MDD=256), using a cut-off of 14 allowed the HCL-32-R2 to discriminate DSM-IV-defined MDD patients between "true unipolar" (HCL-32-R2-) and "sub-threshold bipolar depression" (HCL-32-R2+) with sensitivity=89% and specificity=79%. Area under the curve was.888; positive and negative predictive values were 75.34% and 90.99% respectively. Owing to clinical interpretability considerations and consistency with previous adaptations of the HCL-32, a two-factor solution (F1="hyperactive/elated" vs. F2="irritable/distractible/impulsive") was preferred using exploratory and confirmatory factor analyses, whereas items n.33 ("I gamble more") and n.34 ("I eat more") introduced in the R2 version of the scale slightly loaded onto F2 and F1 respectively. Cronbachs α=.88 for F1 and.71 for F2. Limitations No cross-validation with any additional validated screening tool; treatment-seeking outpatient sample; recall bias; no systematic evaluation of eventual medical/psychiatric comorbidities, current/lifetime pharmacological history, neither record of severity of current MDE. Conclusions Our results seem to indicate fair accuracy of HCL-32 as a screening instrument for BD, though replication studies are warranted.