Cell Phone-Based Expert Systems for Smoking Cessation

Objective: This project examined the effectiveness of tailored text messages as an enhancement to a computerized tailored intervention (CTI) for smoking cessation with Veterans. It was theorized that the ubiquity and sophistication of today's wireless mobile technologies would represent new modes of delivery for empirically based smoking cessation and other behavioral health interventions. CTIs have shown increasing promise as useful behavior change programs for improved health behaviors; however, the ubiquity and sophistication of today's wireless mobile technologies represent new modes of delivery for evidence-based smoking cessation and other behavioral health interventions. The CTI was based on the Transtheoretical Model of Behavior Change (TTM) and developed by Pro-Change Behavior Systems, Inc.

Research Design: This randomized, prospective pilot employed a 2 (group) x 3 (timepoint) research design. Veteran smokers were recruited nationally from direct mailings, Facebook ads, community newsletters, posters, and flyers. In addition to the randomized control trial (RCT), a focus group (n=9) and individual usability tests (n=5) were conducted in order to assess the acceptability of the graphic user interface, questions, tailoring and program content to Veteran smokers. The pilot study tested the effectiveness of the CTI alone and CTI plus individualized text messaging enhancements in reducing smoking behavior in Veterans. In addition to smoking cessation, readiness to change other negative health behaviors � such as alcohol use, healthy eating, and exercise � was also assessed.

Summary of Research Methodology and Procedures: Usability Testing: Five separate usability tests were conducted for the CTI smoking cessation program using a combination of the Think Aloud protocol and Wizard of Oz approach. Usability testing software was used to take notes and record participant interactions with the program. Participants provided qualitative and quantitative feedback on overall presentation and usability of the program. All interview data were coded, categorized, and summarized in a feedback report so that revisions to the system could be adapted. Pilot Study: At all three timepoints, participants completed a brief Health Risk Intervention (HRI), two additional smoking assessments, and the smoking cessation (SC) intervention. The HRI assessment determined baseline readiness to change for smoking and several other related health behaviors, and the smoking cessation intervention provided assessment and feedback about participants' confidence to quit smoking, decisional balance, and use of various change strategies. Following the baseline HRI and SC intervention, the CTI system randomized participants into either the treatment group (Tx), who also received the CTI only, or the treatment-plus group (Tx-plus), who received the CTI and tailored text messages. The study was conducted as entirely anonymous. All participants were given three months' access to the web-based CTI system. The Tx participants received only the online feedback narratives/report that all participants received upon completion of the online Smoking Cessation intervention. Tx-plus participants also received individualized text messages on their cell phones 2-5 times per week (the number of messages correlated with the participant's stage of change) that provide expert guidance, encouragement, and reminders that support their smoking cessation efforts.

Findings: Four-hundred and forty-six Veterans were recruited at baseline. All participants were active smokers, distributed across the stages of change as 11.9% in Precontemplation (n=53; not intending to change in the next 6 months); 63% in Contemplation (n=281; intending to change in the next 6 months), and 25.1% in Preparation (n=112; intending to change in the next 30 days). Two-hundred and thirty-five individuals (52.7%) completed the one-month and three-month follow-ups: n=116 in the treatment group (CTI only; Tx) and n=119 in the treatment-plus group (CTI plus text messages; Tx-plus). Results indicated that the CTI-plus-text-messaging intervention outperformed the CTI-only intervention at an unprecedented quit rate. For those participants in the Tx group 32.8% had quit smoking at three-month follow-up; however, those in the Tx-plus group had quit rates of 43.2%. These findings are highly encouraging, and exceed those quit rates found for other CTI interventions for both groups by almost 15% (Krebs, Prochaska & Rossi, 2010).

Summary of Participant Experiences: While data are available only for those participants who completed the study, indications from the User Satisfaction Scale, as well as the open-ended questions, were that Veterans found the CTI and tailored text messages to be both appropriate and relevant. In addition, participants in the Tx-plus group also indicated that the text messages were helpful, timely and useful. There were no adverse events to report.