Texting to Recruit Unassisted Tobacco Users at a Large Safety Net Health System Into Quitline Service for a Medicaid Value-Based Care Program: Pragmatic Randomized Controlled Trial.

Background: A growing body of research supports the efficacy of text messaging programs to help tobacco users quit, but texting as a strategy for recruiting tobacco users into other evidence-based cessation services, such as quitline coaching, is less well understood. Texting to offer treatment could increase use of cessation resources, an important consideration for health systems trying to improve their quality metric performance on tobacco assessment and counseling. Objective: The aim of this study is to compare the effects of text messages offering free quitline coaching or free nicotine patches on engagement with quitline services by patients identified in electronic health records as unassisted tobacco users. Methods: Participants (N=4171) were adult patients of Los Angeles County Department of Health Services who had had a clinical visit in the past 12 months, were willing to receive text messages, and were identified as unassisted tobacco users (ie, those with no documented advice or assistance to quit in the past 24 months). They included 3139 English speakers and 1032 Spanish speakers. Participants were stratified by language, then randomly assigned to one of two groups. Group 1 received a text offering free quitline coaching. Group 2 received a text offering free nicotine patches. The texts were sent in April and May 2022. Outcome measures were the proportions calling the quitline and receiving evidence-based cessation treatments. Data were analyzed in 2025. Results: Overall, 1.5% (61/4171) of participants called the quitline, including 1.3% (28/2086) in Group 1 and 1.6% (33/2085) in Group 2, and 0.5% (21/4171) received treatment, including 0.4% (8/2086) in Group 1 and 0.6% (13/2085) in Group 2. There was no significant difference either in calls (P=.52) or receiving treatment (P=.29). However, Spanish speakers were significantly more likely to call than English speakers, 2.3% (24/1032) vs 1.2% (37/3139), respectively (P=.008), and engaged in treatment at approximately twice the rate of English speakers, 0.9% (9/1032) vs 0.4% (12/3139), although the latter difference was nonsignificant (P=.07). Conclusions: A single text was effective in connecting unassisted tobacco users with evidence-based quitline services for both English and Spanish speakers, but especially the latter. A population health approach of reaching out to tobacco-using patients outside clinical visits can meaningfully supplement provider referral. Research is needed on ways to increase the population health impact of this strategy.