Validation of the Substance Abuse Screener in American Sign Language (SAS--ASL).

Abstract

The study objectives were to adapt and validate a substance use disorder (SUD) screening instrument in American Sign Language (ASL) to be used to identify those deaf individuals who have a high probability of having an SUD. The goal was to develop an accurate screening instrument that balanced sensitivity and specificity while imposing minimal response burden on respondents. A sample of 198 deaf participants in behavioral health, family social service, and educational programs that provide specialized services for deaf individuals was interviewed to obtain clinical diagnoses for current (past 12 months) SUD according to criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, and completed a 42-item version of the Substance Abuse Screener in American Sign Language (SAS-ASL). We used Rasch and discriminant function analyses to reduce the instrument to 28 items, then divided the sample into a development subsample, used to formulate a scoring routine, and a validation subsample to assess correspondence with clinical diagnoses. To provide validation data on the shortened SAS-ASL, an independent sample of 62 respondents was diagnosed and completed the screener. The SAS-ASL instrument demonstrated good person reliability (.85), sensitivity (.90), and specificity (.84) in the primary validation sample, and 100% screening accuracy with 62 respondents in the second validation sample. The SAS-ASL provides a standardized SUD screening for the deaf population. The adaptability of the instrument to electronic administration lends itself to a continuum of technologically supported services for a high-risk population that is disenfranchised for most community-based behavioral health services.