Problematic substance use and digital interventions : researching intervention efficacy among Internet help-seekers, university students and patients in psychiatry

Abstract

Background: Digital technologies are ever more present in our lives, and are being applied in numerous areas of our existence including the area of health care and health information. Problematic alcohol and drug use impose a major global burden on health, and it is of interest to find new ways of reaching out to those affected by the problem. Interventions delivered over the Internet and in the form of smartphone apps have the potential to be readily available at all times and places. Many apps are publicly available, but only a tiny proportion have been evaluated. Also, there are many evaluated self-help applications for the Internet, but research on the addition of counselor guidance over the Internet is very scarce. There is a need for scientific scrutiny before being able to either recommend or implement these applications into regular care.

Aims: The aims of this doctoral project were: to carry out a preliminary evaluation of an Internet-based program for problematic use with guidance among Internet help-seekers; to investigate effects of three smartphone apps, two with estimated blood alcohol concentration (eBAC) feedback and a skills-training app, among university students; and to determine the prevalence of problematic substance use in a psychiatric population receiving treatment over the Internet, and to explore whether that substance use was associated with treatment outcomes.

Methods: An Internet-based eight-module program was tested in a randomized controlled pilot study with 80 participants. Participants were randomized to: no guidance, guidance over secure email messaging, or choice between guidance over asynchronous secure email messaging or synchronous chat messaging (STUDY I). Effects of two smartphone apps, Promillekoll and PartyPlanner, both with real-time eBAC feedback, and one, PartyPlanner, offering the possibility to simulate a drinking event beforehand and compare afterwards to the real event, were tested in a randomized controlled trial with 1929 university students having problematic alcohol consumption. Both app groups were compared to assessment-only controls (STUDY II). The effects of offering university students with excessive drinking, already enrolled in an ongoing study similar to STUDY II, an additional app focused on skills training were investigated (STUDY III). One hundred and eighty six students were randomly assigned to either an intervention group, receiving access to the app immediately, or a waitlist group given access to the app after six weeks. Both groups were compared to each other and also to a group of assessment-only controls (n=144) from the ongoing study. In the final study, data from 1581 patients having been treated at the internet-based cognitive behavioral therapy (ICBT) clinic in Stockholm, Sweden for major depression, panic disorder or social anxiety disorder, were analyzed to identify the prevalence of problematic substance use at beginning of treatment. The association of problematic substance use with therapy outcomes was also analyzed (STUDY IV)

Results: Study I: The guidance group (the secure email messaging and the choice of messaging mode groups were combined) had significantly lowered their alcohol consumption compared to the no-guidance group. Secure email messaging was the preferred mode of communication over chat for the choice group. Participants in the guidance group rated the program more favorably than participants in the no-guidance group. Study II: There were no differences between the apps and assessment-only controls on any alcohol related outcome, but one negative finding for the Promillekoll app, in that men in that group increased their drinking frequency. Study III: Both the intervention and the wait-list had lower proportions of excessive alcohol consumption compared to assessment-only controls at first follow-up. The intervention group lowered their drinking quantity at first follow-up and their drinking frequency at both follow-ups compared to controls. Study IV: Prevalence of problematic substance use was 32.4%. Exclusive problematic alcohol use was reported by 24.1%, 4.6% reported exclusive problematic drug use and 3.7% combined problematic alcohol and drug use. Problematic drug use among men and 25–34 year olds was associated with completing fewer treatment modules and combined substance use among women and 35–64 year olds negatively affected module completion. Baseline problematic substance use was not associated with worse therapy outcomes for patients treated for depression. However, hazardous alcohol use and probable dependence were associated with worse panic disorder treatment outcomes. Hazardous drug use as well as combined problematic alcohol use with either hazardous drug use or probable drug dependence were associated with worse social anxiety disorder outcomes.

Conclusion: Using novel communications technology is effective for reaching individuals with problematic substance use. Adding counselor guidance to a web-based relapse prevention program has beneficial effects on problematic drinking. Smartphone apps relying on eBAC feedback do not show much promise at the moment. Skills-training smartphone apps may have a beneficial effect on excessive alcohol consumption. The prevalence of problematic substance use is high in ICBT-treated psychiatric outpatients, but for a variety of levels of baseline substance use, the negative effect on therapy outcomes seems limited.