Internet Interventions (Apr 2022)

Text based internet intervention of Binge Eating Disorder (BED): Words per message is associated with treatment adherence

  • Jakob Linnet,
  • Esben Skov Jensen,
  • Eik Runge,
  • Marina Bohn Hansen,
  • Søren Peter Thygesen Hertz,
  • Kim Mathiasen,
  • Mia Beck Lichtenstein

Journal volume & issue
Vol. 28
p. 100538

Abstract

Read online

Background: Some evidence suggests that in internet-based cognitive behavioral therapy (iCBT) the likelihood of adherence is increased when patients write longer messages to the therapist in the program. This association has not previously been investigated in iCBT for Binge Eating Disorder (BED). Objective: In this study, we hypothesized that the number of words written by patients with mild to moderate BED was associated with increased likelihood of treatment completion in a text-based iCBT program. Material and methods: We compared 143 BED patients (92 completers and 51 non-completers) on the number of messages and words written to their therapist during the treatment. Results: Completers wrote significantly more words per message (words/message) than non-completers. The results remained significant after controlling for gender, age, educational level, marital status, children, source of income and intake measures of BED, BMI and depression symptoms (Wald = 14.48, p < .001). The odds ratio of completion increased by 1.5% for each additional word patients wrote per message (OR = 1.015). The model showed a 72.4% classification accuracy, and an optimal cut-off point of 68.99 words/message for differentiating completers and non-completers. The model accurately identified 80.9% of completers (sensitivity) and 54.9% of non-completers (specificity). Conclusions: The number of words/message patients write may have important implications for ascertaining likelihood of adherence and improving adherence rates. From a clinical perspective, therapists should encourage patients to use the option of writing messages to the therapist. Words/message may prove to be a transdiagnostic predictor of treatment adherence in text based iCBT.

Keywords