Customized recommendations and reminder text messages for automated, computer-based treatment during methadone

Abstract

The Recovery Line is an automated, computer-based intervention based on cognitive behavioral therapy (CBT) designed to provide real-time assistance by phone for patients in methadone maintenance. Preliminary efficacy findings were promising, however, as with other computer-based systems for substance use disorder, patient system use was less than recommended. Development and evaluation of system functions to increase patient engagement and use is needed. Thus, we conducted two randomized trials to evaluate system functions designed to increase patient use of the Recovery Line among methadone-maintained patients with continued illicit drug use. In Trial 1 (n = 60), patients received customized, system use recommendations or no recommendations on each Recovery Line call. Ratings of system usability were higher for customized recommendations (CR), but number of calls and total call time did not differ by condition. Trial 2 evaluated characteristics of reminder messages (message frame and reminder latency). Participants (N = 67) received gain- and loss-frame reminder messages, and were randomly assigned to immediate, short, or long term message latency. Although message framing had no effect, gender interacted with latency condition such that females did not differ by message latency, while males had significantly greater total contact time in the short latency conditions. Number of calls differed by condition over time such that the shorter latencies led to greater calls initially, but dissipated over time. Overall the study indicates that computer-based self-management systems can be adapted to increase patient engagement and use. (PsycINFO Database Record

Trial registration: ClinicalTrials.gov NCT02124980.

Figure 1

Trial 2 Call Minutes by Study Week and Reminder Latency Condition. *significant simple effects.

Figure 2

Trial 2 Total Call Minutes by Gender and Reminder Latency Condition. Although call minutes decreased weekly over the course of the study, time did not interact significantly with gender or condition and is not presented. Groups with different lettering were significantly different in post-hoc comparisons.