An integrated clinical-public health informatics system for smoking cessation: Pragmatic clinical trial in 174 practices

Background: Increasing evidence supports web-assisted tobacco interventions, but few smokers take advantage of these tools. Objective: We evaluated an integrated clinical-public health informatics system where primary care providers “e-referred” patients to Decide2Quit.org, a web-assisted tobacco intervention. Methods: Practices (Internal Medicine and Family Practice) were randomized to one of two referral methods: 1) a paper brochure or 2) a point-of-care e-referral completed by entering smoker's email into the ReferASmoker.org web-based e-referral system. Smokers from e-referral practices received automated email links to Decide2Quit.org. All smokers who registered with Decide2Quit.org received access to interactive quit smoking calculators and decision tools. Using a within-practice fractional factorial design, smokers were further randomized to three conditions: (A), receiving only the calculators and tools above, or (B) “A” plus motivational emails tailored to their readiness to quit, or (C) “B” plus asynchronous communication with tobacco treatment specialist. At the practice-level, we evaluated number of referrals and subsequent smoker registrations. At the smoker-level, we compared self-reported smoking cessation at six months. Results: The 174 practices referred 4380 smokers to Decide2Quit.org, with a per-practice mean of 26.7(95% CI=21.0-32.3) among “brochure” practices, compared with 21.9 (range: 17.1-26.6) per e-referral practice (p = 0.3). The e-referral practices had a higher mean number of smokers registrations 6.7 (4.8-8.5) compared with 2.7 (2.0-3.4) in ”brochure” practices, yielding a statistically significant Incident Rate Ratio (IRR) of 2.5 (95% CI 1.8-3.5). At the smoker-level, the tools only (A) condition had a self-reported cessation rate of 17%, and the (B) condition was higher at 26%, as was the (C) condition 25%. Six-month cessation in the two groups that received motivational email messages (B and C) combined was significantly higher than the tools only (A) condition (Odds Ratio 1.6 (95% CI 1.02-2.8) p = 0.38). Conclusion: Clinical practices that used e-referrals had higher rates (more than double) of smokers using Decide2Quit.org, and augmented Decide2Quit.org tools increased patient quit rates.