Evidencia de validez en Chile del Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) / Validity evidence of the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) in Chile

RESUMEN

El objetivo del estudio fue analizar la validez de la versión chilena de un instrumento para la identificación de diferentes niveles de riesgo asociado al consumo de alcohol, tabaco, marihuana y cocaína(Alcohol, Smoking and Substance Involvement Screening Test, ASSIST).Se evaluó la fiabilidad y consistencia del instrumento además dela validez concurrente y discriminante. La muestra fue de 400usuarios de servicios de tratamiento drogas y alcohol ambulatorios y residenciales de la atención primaria de salud, comisarías y empresas. La consistencia interna obtenida fue alta (Alcohol α= .86. marihuana α= .84 y cocaína α= .90). El coeficiente de correlación intraclase (CCI) con la comparación test-retest fue buena para Alcohol (CCI = .66), marihuana (CCI = .74) y cocaína (CCI =.80). Se observó una buena correlación entre los puntajes del ASSIST el puntaje del AUDIT (r de Pearson = .85), ASI-Lite (r entre .66 y .83 para tabaco, alcohol, marihuana y cocaína) y SDS (r = .65). El punto de corte original para la detección del riesgo alto es de 27 puntos, sin embargo, considerando un mejor balance entre sensibilidad y especificidad se modificó ese corte a 21. Los resultados obtenidos en éste estudio demostraron las buenas propiedades psicométricas del ASSIST para la detección de distintos niveles de riesgo asociados al consumo de sustancias en población general de Chile

ABSTRACT

This study aims to psychometrically validate the Chilean version of the Alcohol, Smoking and Substance Involvement Screening Test ASSIST. Specifically, this study is interested in evaluating the reliability, consistency and concurrent and discriminant validity of this instrument. The sample was composed for a total of 400 people from four different settings treatment centers (residential and ambulatories), primary health care, police stations and companies. The reliability of the ASSIST was high (α = .86 for Alcohol, α = .84 for marijuana and α = .90 for cocaine). The intra class correlation coefficient (ICC) with test-retest comparison was statistically significant for Alcohol (ICC = .66), marijuana (ICC = .74) and cocaine (ICC = .80). There were statistically significant correlations between the ASSIST and the AUDIT score (Pearson’s r = .85), the ASSIST and the ASI-Lite score (r between .66 and .83 for tobacco, alcohol, marijuana and cocaine), and the ASSIST and the SDS score (r = .65). The original cutoff point for high risk detection was 27 points, however, in order to have a better balance between sensitivity and specificity the cut was changed to 21 points. The ASSIST presents good psychometric properties and therefore is a reliable and valid instrument to be used as a mechanism to detect risk levels of substance use in the Chilean population