Abstract
The Kessler 6 (K6) scale has been widely used to screen for serious mental illness (SMI) in general populations. Given that populations with high burden of HIV may be more likely to experience poor mental health outcomes, it is important to validate measures used to assess SMI among these groups. Using data from CDC's National HIV Behavioral Surveillance (NHBS) system in the United States, we examined psychometric characteristics of the K6 scale, including exploratory factor analysis, scale reliability, construct validity, and review of interviewer feedback to explore scale performance in the field. Analyses confirmed a single factor structure and demonstrated that the K6 scale had high internal consistency and construct validity. Although interviewers noted that the effort item was difficult for some participants to understand, psychometric properties were not greatly improved with the deletion of the item. The results of this analysis suggest that the K6 scale is appropriate for use among NHBS populations with a high burden of HIV, including persons who inject drugs, heterosexually active adults at increased risk for HIV infection, women who exchange sex for money or drugs, and gay, bisexual, and other men who have sex with men.