Reliability and acceptability of automated telephone surveys among Spanish- and English-speaking mental health services recipients.

Interactive Voice Response (IVR), an automated system that administers surveys over the phone, is a potentially important technology for mental health services research. Although a number of studies have compared IVR to live interviews, few have looked at IVR in comparison to pencil-and-paper survey administration. Further, few studies have included subjects from those populations most likely to benefit from IVR technology, namely patients with lower education levels and non-English-speaking patients. This randomized clinical study, conducted at a community health center serving low-income English- and Spanish-speaking populations, assessed the reliability of an IVR-administered Brief Symptom Inventory (BSI) relative to a paper-and-pencil version. The study was adequately powered. Results showed that patients gave similar responses to the IVR and paper-and-pencil surveys; in addition, patients were generally equally satisfied with both experiences. We conclude that, while more large-scale research is needed, IVR can be a useful survey administration tool.

Conversion and validation of the Teen-Addiction Severity Index (T-ASI) for Internet and automated-telephone self-report administration.

This study converted the Teen-Addiction Severity Index (T-ASI) into self-report formats using Internet (Net) and interactive voice response (IVR) automated-telephone technologies. Reliability and convergent validity were assessed among 95 inpatient adolescent participants. Current functioning scores obtained by clinician interview correlated well with self-report Net (mean r=.74, SD=.14) and IVR (mean r=.72, SD=.16). Lifetime history items obtained by clinicians were consistent with self-report Net (mean r=.60, SD=.32; mean kappa=.67, SD=.24) and IVR formats (mean r=.60, SD=.30; mean kappa=.64, SD=.26). Participants rated "ease of use" as being high for both Net and IVR formats. These findings suggest that automated T-ASI administration is a valid and potentially less expensive alternative to clinician-administered T-ASI interviews.

Validation of the Addiction Severity Index (ASI) for internet and automated telephone self-report administration.

This study assesses the convergent validity of Internet (Net) and interactive voice response (IVR) automated telephone self-report versions of the Addiction Severity Index (ASI) relative to the established, clinician-administered (CA) ASI. Eighty-eight subjects were recruited from an addiction treatment program to complete three ASI assessments. The mean correlation between composite scores obtained by Net and IVR and those obtained via clinician interview was.91 (range.81-.95). For History items, the mean correlation was.77 (range.14-1.00) and the mean kappa coefficient was.75 (range.46-1.00). The results demonstrated the validity of these self-report Net and IVR versions of the ASI. Self-report Net and IVR were rated as "very satisfactory" or "extremely satisfactory" by a majority of respondents for ease of use. Automation can reduce the labor costs associated with ASI administration and may facilitate longitudinal tracking of subjects from home.