We recruited smokers receiving primary care at Massachusetts General Hospital-affiliated community health centers in Charlestown and Revere, Massachusetts. We identified patients receiving primary care using a validated algorithm used by the Massachusetts General Hospital Practice-Based Research Network [32]. Other eligibility criteria based on EHR data included a primary language of English, age 18 years and older, current smoker, and with a mobile telephone number.

We developed the GetReady2Quit (R2Q) text messaging program with content for smokers ready to quit in the next 30 days and content for smokers not ready to quit (Table 1). For smokers who were ready to quit in the next 30 days, we used the downloadable QuitNowTXT library [33]. QuitNowTXT includes 118 messages delivered over 6 weeks tailored to a user-entered quit date. These messages include behavioral advice and motivational and educational messages about the harms of tobacco and the benefits of quitting. The program has limited two-way communication including keywords to request help by texting “CRAVE,” “MOOD,” or “SLIP.” There were also weekly smoking status assessment messages that invited a user response. Unlike other text messaging programs [34], if a user did not respond to an assessment, no further messages were sent. Maximum message volume was 25 messages per week with the highest volume in the 2 weeks before and after the quit date. For smokers who were not ready to quit in the next 30 days, 31 novel messages were developed by an expert team of primary care physicians (PCPs), a tobacco cessation counselor, a mobile health manager, and a behavioral scientist.

Content included 16 motivational messages to encourage individuals to identify personal reasons for change and internal motivations to quit [35,37]. Fifteen messages encouraged smokers to try a practice quit attempt explained as an attempt to not smoke for hours or days without a commitment to stop for good [36]. Practice quit attempts can increase motivation and self-efficacy [36,38]. Smokers not ready to quit were sent three to five messages per week. At the end of this message campaign, users were asked again if they were ready to quit in the next 30 days. Those that were ready were sent the QuitNowTXT messages. Those that were not ready were sent a final recommendation to contact their doctor or the state quitline.

Between March 2014 and June 2015, PCPs were asked to screen potentially eligible patients. The PCP-approved patients were sent an opt-out letter (Multimedia Appendix 1), informing them about the purpose of the feasibility study, content of the R2Q text messaging program, and that they would be sent a text message in the next week unless they called to opt out. Patients who did not opt out were sent a single text message inviting them to opt in to the R2Q program (Table 1). Opting in implied consent. Participants were sent four text message queries assessing nicotine dependence, readiness to quit, and quit date. Ethical approval was obtained from the Partners Healthcare, Inc Institutional Review Board.

We compared R2Q enrollees, those who opted in following the proactive text message, with patients who were eligible and invited but who did not enroll in terms of demographics, primary insurance, and comorbidities. We used portions of the Reach Effectiveness Adoption Implementation Maintenance (RE-AIM) methodology [31] to measure program reach, engagement, and adoption. Reach was the proportion of users sent a proactive text invitation who opted in. Engagement was measured as sending one or more keywords to the server. Adoption was defined as days in the text messaging program before the participant texted “STOP” or failed to respond to an assessment message. We compared engagement and adoption by readiness to quit using unadjusted t tests, Wilcoxon rank sum tests, and chi-square tests.

This study evaluated a proactive tobacco cessation intervention that reached out to patients by text message. It shows promise as a low-cost, scalable intervention for primary care populations. Program reach at 9.3% was comparable to other proactive care models for smokers that used more intensive outreach methods, including up to 15 outreach telephone calls [24]. Similar to telephone outreach programs, both smokers ready to quit and those not ready to quit enrolled [27]. These results support the feasibility of future work to design and test a proactive text messaging intervention targeting primary care patients.

Text messaging programs originating from the physicians’ office may leverage the influence physicians have on smokers [18]. Individuals most often look to their own health care systems for online health information [39]; therefore, trust in their health care providers may make health-promoting advice more potent if it is coming from their physicians’ office. This trust may also encourage even unmotivated smokers to engage in health-promoting activities sent to them by their physicians’ office.

Our single message enrollment process was simpler than recruitment used by other proactive care models [20-26]. The low intensity may have been appealing to both smokers not ready to quit and those busy managing other chronic diseases, who may not have time or interest in more complex interventions. Indeed, except for cardiovascular disease, patients with comorbid chronic diseases were no less likely to opt in. Integrating the program with other optional cessation services, such as pharmacotherapy, may increase the program’s appeal and improve reach and effectiveness. Future work will need to explore ways for text messaging to be integrated with other cessation services available to primary care patients who smoke.

In this pilot study, we did not have enough resources to assess smoking outcomes or receipt of text messages. Therefore, we could not account for invalid telephone numbers or failed message delivery. If these are considered, the uptake of the program following receipt of the proactive text may have been even higher.

A proactively delivered text messaging program targeting primary care patients who smoke reached as many smokers with a single text as more intense and costly telephone call- or mailed-based proactive outreach methods. This method engaged both smokers ready to quit and those not ready to quit and shows promise as part of a proactive care model for addressing smoking in primary care populations.