The Challenges of Implementing a Text Message Intervention to Promote Behavioral Change in Primary Care Patients With Hypertension and Diabetes

Abstract Background Controlling blood pressure and glycemic levels is a challenge that requires innovative solutions. Objective To assess the feasibility of implementing a text message intervention among low-income primary care patients, as well as to assess self-reported behavioral change. Methods A set of 200 text messages was developed on healthy eating, physical activity, adherence, and motivation. Participants from Vale do Mucuri, MG, Brazil diagnosed with diabetes or hypertension or undergoing screening for those diseases, received 5 to 8 messages per week for 6 months. They answered a questionnaire to report their satisfaction and behavioral changes. Results Of the 136 patients, 117 (86.0%) answered the questionnaire. Most reported that the messages were very useful (86.3%), easy to understand (90.6%), and were very helpful for behavioral change (65.0%); 84.6% reported that they had started eating healthier. The most frequent reported lifestyle changes were: improved diet quality (85.5%), reduced portions (65.8%), and weight loss (56.4%). The majority of patients shared the messages (60.7%) with family or other acquaintances, considered the number of messages to be adequate (89.7%) and would recommend the program to others (95.7%). Conclusion An intervention based on text messages to promote behavioral change in patients with hypertension or diabetes in primary care is feasible in low-resource settings. Future studies are needed to assess the program’s long-term effects on clinical outcomes.


Introduction
The burden of non-communicable diseases such as hypertension and diabetes is growing, especially in developing countries.This is largely attributable to behavioral factors, overconsumption of unhealthy foods and energy dense diets, obesity, social stress, and sedentary lifestyle, [1][2][3] which are a consequence of urbanization, industrialization, and globalization. 3,4ontrolling blood pressure and glucose levels in patients with hypertension and diabetes is challenging, especially in resource-constrained regions.²Strategies are required that empower and encourage behavioral change and promote treatment adherence. 3,5,6Mobile phone ownership has spread worldwide, allowing mobile-technology-based administration, capillary blood glucose monitoring, and hypoglycemia.The messages were based on evidence-based guidelines [12][13][14][15][16][17][18][19][20][21][22][23] and the Brazilian National Nutritional Guide. 24e messages were designed to provide information, motivation, advice, and positive reinforcement, to promote medication adherence, and to encourage patients to adopt healthy dietary habits, exercise, and quit smoking (where applicable).The messages, which contained up to 145 characters, were reviewed by an expert panel consisting of 1 family physician, 1 cardiologist, 1 endocrinologist, 2 psychologists, 1 nurse, 1 nurse technician, 1 physical therapist, 2 pharmacists, 2 nutritionists, 1 physical education specialist, and 3 patients.Health care professionals were asked to assess the messages' clinical benefits, readability, usefulness, and language, while patients were asked to assess their comprehension and usefulness.All panel members could suggest changes and new messages.
[27] The final set included 198 messages.Information on the groups of patients who would receive the specific messages was obtained from the medical record system.The set included both an introductory and a concluding message, which explained that the intervention was over but encouraged the continuance of healthy practices (totaling 200 messages).The messages were semi-personalized.Some examples are shown in Box 1.

Field study
The patients were recruited by primary care professionals.They received text messages 5 to 8 times a week on weekdays for 6 months (February to August, 2018).The messages were always sent at the same time (10:00 AM).Patients with diabetes and/or smokers also received extra text messages, sent at 4:00 PM.These times were based on general Brazilian eating habits, especially those of people in the study region. 28e unidirectional messages, with content predetermined according to the day of the week, were sent free of charge to the patients by an automated program developed and tested by our group. 28terventions (mHealth) in patient education and chronic condition management. 7recent systematic review assessed the effectiveness of mHealth interventions to promote physical activity and healthy diet in low-and middle-income countries.The results were heterogeneous, and no clear patterns could be determined among interventions that succeeded and those that did not. 4Thus, further studies on the topic are warranted and timely.
Since text messaging, an easy strategy based on a widespread and low cost communication system, is considered an appropriate intervention for a variety of health behaviors, 8 our aim was to assess the feasibility of a text message intervention among primary care patients with hypertension or diabetes in a low-resource setting, evaluating self-reported behavioral change and user satisfaction.

Study design and setting
This is a substudy of HealthRise Brazil Teofilo Otoni (HRTO) Project.The project is part of a previously described multinational program. 9,100][11]

Participants
Patients aged 30 to 69 years, with no concomitant diagnosis that would limit life expectancy and no severe cognitive impairment, being neither bedridden nor unable to attend appointments at the primary care unit, were eligible for the HRTO Project.For the present study, participants in HRTO Project were eligible if they agreed to participate and had a mobile phone capable of receiving short text messages.Informed consent was obtained from all participants.

Satisfaction and behavioral change survey
Based on a previous intervention, a questionnaire was developed about patient satisfaction and selfreported behavioral change. 28It contained 15 questions (14 multiple choice and 1 open-ended).
A Likert scale was used to ask the patients' opinion about the messages' impact on behavior, whether they would like to continue receiving the messages, (if so) with what frequency, and if they would recommend the intervention to a friend or a relative.In a pilot study, the 3 patients had great difficulty understanding differences among the alternatives, so we reduced them to 3 for the sake of clarity.
The survey was applied by telephone.We attempted to call each registered number at least 3 times.When there was no response, a text message was sent describing the topic of the call and the day and time of the next attempt.

Data analysis
We used descriptive methods to assess patient characteristics and survey responses.Continuous variables were expressed as mean and standard deviation (SD) or median and interquartile range (IQR).The normality of data distribution was assessed using the Kolmogorov-Smirnov test, with a 5% significance level.Categorical variables were expressed as absolute and relative frequencies.All statistical analyses were performed using IBM SPSS Statistics 21 (IBM, Armonk, NY, USA).

Results
Of the 136 patients who received the messages (median age 57 [IQR 51-64] years, 67.6% female) (Table 1), 17 did not read them due to the following reasons: they were illiterate (n = 2), they had no interest in their content (n = 6), they were not the intended recipient (n = 3) because the participant changed telephone numbers during the study without informing the research team or health unit, or for unknown reasons (n = 6).Two other participants refused to answer the survey, leaving a total of 117 (86.0%) survey respondents.
The questionnaire responses are shown in Table 2.The majority of patients considered the messages very useful (86.3%), easily understandable (90.6%), and very helpful for behavioral change (65.0%).Regarding behavior, 84.6% reported a healthier diet, although but most reported no changes regarding exercise (52.1%).The most frequently reported lifestyle changes were improved nutrition quality (85.8%), greater energy for daily activities (69.2%), and reduced portions (65.85).Most respondents reported having shared the messages with friends or relatives (60.7%), considered the volume of messages appropriate (89.7%), and would recommend the program to a friend or a relative (95.7%).

Discussion
This study, developed in a resource-constrained area, found that most participants were pleased to receive text messages promoting healthy lifestyle and self-care regarding hypertension, diabetes, or both.A total of 86%, 91%, and 65% of patients reported that the messages were very useful, easy to understand, and were very helpful for lifestyle change, respectively.The most frequently reported lifestyle changes were improved diet (86%), greater energy for daily activities (69%), and reduced portions (66%).

Box 1 -Examples of text messages sent to the participants Introductory message
Hello, [name]!Beginning today, you will receive messages from the HealthRise Project with important information about your health.Read them carefully.

Messages about good nutrition
Hello, [name]!Processed food (such as sodas, canned food, and noodles) have a high sodium content and can lead to hypertension.

Messages about physical activity
Hello, [name]!Have you considered using the stairs instead of the elevator?Climbing the stairs uses more calories!Try it!

Motivational messages
Hello, [name]!Take care of your health, believe in yourself.Go to your scheduled appointments with the doctor or nurse.

Messages containing general information
Hello, [name]!Complications related to high blood pressure may affect the heart, eyes, brain, and kidneys.Talk to your doctor about this.

Final message
Hello, [name]!This message program ends today, but keep up your efforts to maintain healthy habits.They are crucial for your treatment.

The message limit was 145 characters in Brazilian Portuguese.
Int J Cardiovasc Sci.2023;36:e20220050 heart disease.They found mean reductions of 5 mg/dL in low-density lipoprotein cholesterol (95% CI, -9-0), 7.6 mmHg in systolic blood pressure (95% CI, -9.8--5.4),and 1.3 kg/m 2 in BMI (95% CI, -1.6--0.9), in addition to a lower risk of smoking (relative risk 0.61, 95% CI 0.48-0.76)and increased physical activity. 36,37Results about the impact Our results demonstrate that the text message intervention was feasible and well appreciated by primary care patients in a resource-constrained and mostly rural region, indicating that it could positively impact their hypertension/diabetes care.Saleh et al., 29 assessed the perception of 1000 rural dwellers and refugees in the Middle East diagnosed with hypertension, diabetes, or both, who received text messages to promote lifestyle change.Their results corroborate our findings, ie, the majority of the participants found the messages useful, easy to understand, and that they adhered to the lifestyle change suggestions.
Previous studies in other countries have shown the benefits of text message interventions aimed at behavioral change to help control diabetes.A meta-analysis of 13 clinical trials (1164 diabetes patients) found a 0.62% reduction in A1c levels. 30Glycohemoglobin and blood pressure are surrogates for important clinical outcomes and complications in these patients.In a prospective study on the association between A1c levels, blood pressure, and vascular complications in patients with type 2 diabetes, for each 1% decrease in mean HbA1c, there was a 37% reduction in microvascular complications and a 21% reduction diabetes-related complications, including death. 31There was a direct association between blood pressure and complication risk: for each 10 mmHg reduction in systolic blood pressure, there was a 12% decrease in any complications and a 15% decrease in deaths. 32her benefits have been increasingly reported for text message programs that promote lifestyle changes, physical activity, smoking reduction or cessation, and medication adherence among patients with chronic noncommunicable diseases. 7,33In the TELEFIT study, 4 months of educational and motivational text messages promoting weight reduction were sent to 46 patients followed up in primary care.The intervention led to significant reductions in body mass index (BMI) (31.3 kg/m 2 vs 29.9 kg/m 2 , p < 0.001) and systolic and diastolic blood pressure (125 mmHg vs 120 mmHg, p = 0.013, and 80 mmHg vs 80 mmHg, p = 0.006, respectively). 28 meta-analysis assessing the impact of text messages on smoking found a 1.37 times (95% CI, 1.25-1.51)higher odds of smoking cessation. 34In another meta-analysis, Thakkar et al., 35 found a 2-fold increase in medication adherence among patients with chronic diseases after a text-message intervention (OR 2.11, 95% CI, 1.52-2.93).In a randomized clinical trial, Chow et al., 36 evaluated a 6-month motivational text-message intervention for lifestyle modification in secondary prevention of coronary of text message interventions on systolic and diastolic blood pressure in hypertensive patients are still inconclusive.Some studies have demonstrated benefits, but their quality of evidence is low. 38,39 2018 in Brazil, 82.9% of urban dwellers and 57.3% of rural dwellers were mobile phone users. 40At the end of 2019, 51.6% of all numbers were on a prepaid basis.This method has more controllable costs, and defaulting on payment has no major consequences, 41 which facilitates cancelling and changing numbers.Thus, it is of the utmost importance to follow the World Health Organization's recommendation to monitor message reception by patients in text message interventions. 42In Box 2, we describe the lessons we learned from the intervention.
This study suggests that a lifestyle intervention based on behavioral change principles that delivered self-care advice through a well-accepted information and communication modality (text messaging) effectively promoted lifestyle change in a group of 117 Brazilians adults.Such changes help control non-communicable diseases, such as hypertension and diabetes.The fact that a relevant proportion of the participants shared the messages with relatives and/or friends is suggestive of the intervention's reach, ie, that behavioral change can spread beyond participants to the people in their environment.
This study included the collaboration of a psychologist with expertise in primary care, ensuring that the messages were not negatively framed.We believe that future studies investigating text messages as a health intervention should focus on positively framed messages.Patients play a fundamental role in their own health care and must be encouraged to participate in decisions about their therapy and to understand their options and the consequences of their actions.An imperative tone can be a barrier to behavioral change. 43Positively framed messages tend to be more acceptable, empowering patients and including them as an important partner in their own care.
Although the results of this study hold promise, including the high rate of engagement and the remarkable lifestyle changes, it is important to recognize some limitations.It is possible that people who might have benefitted from the intervention were excluded due to lacking a mobile phone or being illiterate.Moreover, the intervention's duration was relatively short, and the sample size was small.Future studies are needed to assess its long-term impact.Several factors could be altered or enhanced to further improve clinical outcomes.It is known, for example, that self-monitoring data are subject to reporting bias.Additionally, potential barriers should be taken into consideration when a text messaging protocol is implemented in low resource settings, including inadequate health literacy, which could lead to misinterpretation of message content, vision problems (especially considering that most of our participants were over 60 years old), and language barriers.To overcome these problems, we assessed message comprehension in primary care patients prior to implementing the intervention.Other barriers to such interventions include privacy and disclosure issues, as well as logistical issues, such as a lack of network coverage or the inability to charge a phone due to lack of electricity, although the latter has been decreasing with expanding national infrastructure.
Although telehealth solutions for patients with chronic diseases are becoming a trend, such initiatives are generally isolated efforts and are not integrated with other information and communication technology solutions, such as electronic health records.Our service (Telehealth Center of the Universidade Federal de Minas Gerais University Hospital) is currently involved in a joint program with the Brazilian Ministry of Health in an attempt to resolve this problem, integrating text message delivery with the medical record system.Policy makers must support the development and implementation of mHealth interventions, considering that such an investment will lead to better care for those suffering from chronic diseases, thus strengthening health services.The importance of investing in e-health and telehealth interventions has become more evident since the COVID-19 pandemic, forcing us to reassess the conduct of social organizations and health practices. 44

Conclusion
Our study showed that a text-based intervention to promote behavioral change in patients with hypertension or diabetes is feasible in resource-scarce settings, reinforcing what appears to be a current trend in health care.The participants were pleased to receive text messages and were willing to continue receiving them.In light of the growing rates of non-communicable diseases worldwide, future studies should assess the limits, reach, and long-term effectiveness of motivational and health educational text messages on clinical outcomes.

Table 1 -Clinical characteristics of the participants (n = 136)
A1c: glycated hemoglobin; BMI: body mass index; DBP: diastolic blood pressure; IQR: interquartile range; SBP: systolic blood pressure; SD: standard deviation.a Extracted from the medical record system.b Only patients with diabetes.Number of patients with valid A1c values who did and did not receive the messages at baseline and after 6 months were n = 7 and n = 12, respectively.c Only patients with hypertension.The number of patients