The reliability of sensitive information provided by injecting drug users in a clinical setting: Clinician-administered versus audio computer-assisted self-interviewing (ACASI)

Abstract

Research with injecting drug users (IDUs) suggests greater willingness to report sensitive and stigmatised behaviour via audio computer-assisted self-interviewing (ACASI) methods than during face-to-face interviews (FFIs); however, previous studies were limited in verifying this within the same individuals at the same time point. This study examines the relative willingness of IDUs to report sensitive information via ACASI and during a face-to-face clinical assessment administered in health services for IDUs. During recruitment for a randomised controlled trial undertaken at two IDU-targeted health services, assessments were undertaken as per clinical protocols, followed by referral of eligible clients to the trial, in which baseline self-report data were collected via ACASI. Five questions about sensitive injecting and sexual risk behaviours were administered to participants during both clinical interviews and baseline research data collection. “Percentage agreement” determined the magnitude of concordance/discordance in responses across interview methods, while tests appropriate to data format assessed the statistical significance of this variation. Results for all five variables suggest that, relative to ACASI, FFI elicited responses that may be perceived as more socially desirable. Discordance was statistically significant for four of the five variables examined. Participants who reported a history of sex work were more likely to provide discordant responses to at least one socially sensitive item. In health services for IDUs, information collection via ACASI may elicit more reliable and valid responses than FFI. Adoption of a universal precautionary approach to complement individually tailored assessment of and advice regarding health risk behaviours for IDUs may address this issue.

Keywords:

• reliability
• injecting drug use
• risk reporting/disclosure
• primary healthcare
• socially desirable behaviours
• stigma

Acknowledgements

The HAVIT study, through which some of the data reported here were collected, was supported by National Health and Medical Research Council (NHMRC) Project Grant # 510104 (CIA Maher). The authors would like to acknowledge the coordinated efforts and dedication of the clinical and research teams at the Redfern Harm Minimisation Clinic and the Kirketon Road Centre, and particularly Dr Rachel Deacon, the HAVIT study coordinator. Lisa Maher is supported by the award of an NHMRC Senior Research Fellowship. The first author's doctoral research is supported by a University International Post-graduate Award from the University of New South Wales. The last author is supported by NHMRC fellowships.