Comparing satisfaction, alliance and intervention components in electronically delivered and in-person brief interventions for substance use among childbearing-aged women

Abstract

Electronic delivery of Screening, Brief Intervention, and Referral to Treatment (e-SBIRT) may be a low-cost and high-reach method for screening and brief intervention in health care settings. However, its relative acceptability, ability to build a therapeutic alliance, and delivery of key intervention components compared to in-person SBIRT (SBIRT) is unclear. The association of these factors with intervention outcomes is also not known. We compared SBIRT and e-SBIRT on satisfaction, alliance, and receipt of intervention components, and evaluated the extent to which these intervention dimensions were related to later substance use. Data were collected as part of a randomized clinical trial (N = 439) examining SBIRT, e-SBIRT, and enhanced usual care for childbearing-aged women in two reproductive healthcare clinics (see Martino et al. (2018) for main trial findings). Participants receiving SBIRT or e-SBIRT (N = 270) rated satisfaction and alliance following a single-session, brief intervention, based on motivational interviewing that targeted hazardous substance use (tobacco, alcohol, illicit drugs and prescribed medications). Trained raters coded audio-recorded SBIRT sessions for the presence of six major intervention components, and evaluated the occurrence of these components in the e-SBIRT software. Overall, participants in both groups reported strong satisfaction (on average, “considerably” to “extremely” satisfied) and perceived working alliance (on average, “very often” to “always” allied). SBIRT participants provided higher overall alliance ratings, felt more encouraged to make their own decisions, and rated the intervention's likely helpfulness to other women higher. Fewer e-SBIRT participants received intervention components focusing on personalized feedback, developing importance of and confidence in making changes to substance use, and developing a plan to change, compared to SBIRT participants. However, e-SBIRT participants were equally or more likely to receive components seeking to help them understand their use, discussing reasons for use, and summarizing and supporting what the patients elected to do. Notably, satisfaction, alliance, and number of intervention components received were not associated with total days of substance use. Although we found no evidence that the intervention characteristics evaluated in this study were associated with outcomes, acceptability and alliance may have other important implications. Findings suggest areas for improvement with respect to e-SBIRT satisfaction and alliance formation.ClinicalTrials.govregistration number: NCT01539525