Gender effects on client-spousal collateral agreement levels in substance abuse posttreatment reports


This study examines gender effects on client-spousal collateral agreement in substance abuse posttreatment reports and assesses the dependability of spousal collaterals in corroborating these reports. Study participants included 178 male clients with female spousal collaterals (male client dyads) and 58 female clients with male spousal collaterals (female client dyads) who completed follow-up interviews 6 months post admission. Interviews were conducted from July 2002 to June 2004. Levels of agreement were examined by dyad type for responses to 13 questions regarding clients' treatment experiences and treatment effectiveness. A simple kappa coefficient was calculated to measure the degree of nonchance agreement between clients and their spouses using SAS/Base, SAS/Stat, and SAS/IML. Agreement levels between the two dyads were compared and examined whether gender differences in agreement levels were significant. Agreement levels were found to be high, ranging from 80.4 to 97.1% for male client dyads and from 76.5 to 100% for female client dyads. Six of the 13 questions for male client dyads and 8 of the 13 questions for female client dyads revealed excellent or substantial nonchance agreement. Questions related to domestic violence had poor nonchance agreement, with "Victim of domestic violence since treatment" having the lowest score for the male client dyad. Only one question, "Currently on probation or parole," showed a significant difference across genders, as males revealed a significantly higher level of agreement. Moderate to substantial agreement was determined for 11 out of the 13 questions regardless of gender effects. This study provides empirical data affirming the high dependability of spousal collaterals to validate responses about clients' substance abuse following treatment. However, the study revealed a high degree of disagreement among both male client dyads and female client dyads regarding questions related to domestic violence (including whether the client had committed or been a victim of domestic violence). Following rigorous gender analyses, this study reveals little difference overall in agreement levels between gender dyads.

Publication Title

Addiction Research and Theory