IMPORTANCE: Cardiovascular disease prevention, including lifestyle modification, is important but underutilized. Mobile health strategies could address this gap but lack evidence of therapeutic benefit.
OBJECTIVE: To examine the effect of a lifestyle-focused semipersonalized support program delivered by mobile phone text message on cardiovascular risk factors.
DESIGN AND SETTING: The Tobacco, Exercise and Diet Messages (TEXT ME) trial was a parallel-group, single-blind, randomized clinical trial that recruited 710 patients (mean age, 58 [SD, 9.2] years; 82% men; 53% current smokers) with proven coronary heart disease (prior myocardial infarction or proven angiographically) between September 2011 and November 2013 from a large tertiary hospital in Sydney, Australia.
INTERVENTIONS: Patients in the intervention group (n = 352) received 4 text messages per week for 6 months in addition to usual care. Text messages provided advice, motivational reminders, and support to change lifestyle behaviors. Patients in the control group (n=358) received usual care. Messages for each participant were selected from a bank of messages according to baseline characteristics (eg, smoking) and delivered via an automated computerized message management system. The program was not interactive.
MAIN OUTCOMES AND MEASURES: The primary end point was low-density lipoprotein cholesterol (LDL-C) level at 6 months. Secondary end points included systolic blood pressure, body mass index (BMI), physical activity, and smoking status.
Researchers looked at patients with coronary heart disease being treated at a tertiary care center in Australia. They had to be older than 18 years and have one of the following: a prior myocardial infarction, coronary artery bypass graft surgery, percutaneous coronary intervention, or at least 50% stenosis in at least 1 major epicardial vessel. They were randomized to two groups. The first got “usual care” The second got usual care, but also received semi-personalized test messages which were tailored to provide advice, motivation, and information that aimed to help them change their diet, be more active, or stop smoking (if they were).
They got 4 test messages a week for 24 weeks. They were sent on 4 of 5 randomly selected weekdays at random times during the day. The messages were not interactive, and they were told not to respond to them. They could stop getting the messages at any time. Of the 341 patients in the intervention group, only 7 asked to stop getting them: 4 didn’t like them, 1 developed comorbidities, 1 moved to another country, and 1 went on vacation.
All participants were assessed at baseline and at 6 months for LDL cholesterol, systolic blood pressure, BMI, physical activity, and smoking status.
Let’s take a pause here. If you had asked me to bet, I would have put all the money in my pocket on this being a negative trial. I mean, random text messages? That’s all? One-way communication? No way this would make a difference.
I was wrong. LDL cholesterol was 5 points lower in those in the intervention (79 mg/dL versus 84), and it started higher in the intervention group. Systolic blood pressure was 8 points lower (128 mm Hg versus 136). BMI was 1.3 points lower (28 versus 30.3). Physical activity was way up (936 metabolic equivalent task minutes per week versus 643), and the percentage of people smoking was way down (26% versus 43%). Before the trial, 53% of participants were smoking. I mean the NNT for this intervention just for smoking cessation is less than 6. It’s crazy.
They estimated the cost of this program to be about $10 for the 96 messages sent out. We could automate this easily. You don’t even need a smartphone to get a text message.
Granted, this was a reasonably motivated population, as they were already sick. But this program worked much more than usual care. It also focused on risk factors, not actual outcomes. We should do that study, too. It’s also possible that some of the self-reported measures might be inaccurate, but most of them weren’t self-measure. These are also only 6-month outcomes, and we don’t know how much they will stick.
This is promising. It really is. It’s low cost. It’s easy to do. It can work in resource-limited settings. If this was a drug, people would be investing like crazy.
