Skip Navigation
Service Delivery Innovation Profile

Daily Text Messages and Nurse Followup Improve Self-Management Behaviors in Patients with Diabetes, Leading to Better Glycemic Control and Lower Costs


Tab for The Profile
Comments
(0)
   

Snapshot

Summary

University of Chicago Medicine (a large academic medical center) developed and implemented a program called CareSmarts in which patients with diabetes received several text messages each day providing education on how to manage the disease and reminders to follow recommended self-care measures. Some messages asked patients to send a reply text indicating whether they were heeding the reminders. A nurse subsequently called those who consistently failed to respond and those whose responses indicated a need for assistance, working with them and their physicians to overcome obstacles to proper self-management of the disease. The program improved self-management capabilities, leading to better glycemic control, fewer medical visits, lower costs, and high patient satisfaction. The program operated successfully for six months beginning in May 2012, after which time it was discontinued due to a lack of ongoing funding.

Evidence Rating (What is this?)

Moderate: The evidence consists of a comparison of trends in blood glucose levels among program participants and a control group of similar patients who decided not to participate in the program; pre- and post-implementation comparisons of adherence to various recommended self-management behaviors, medical visits, and costs; and post-implementation reports from participating patients on their satisfaction with the program.
begin do

Developing Organizations

University of Chicago Medicine
end do

Date First Implemented

2012
begin ppxml

Patient Population

During the trial, the typical participant was 53 years old and had been dealing with diabetes for 8 years. Approximately two-thirds were African American, with the remainder being white (26 percent) or Asian/Pacific Islander (7 percent). With respect to diabetes control, participants fell into three categories, each of roughly the same size: those with the disease well under control (defined as glycated hemoglobin, or HbA1c, of 7 percent or less), those who had the disease moderately well controlled (between 7 and 8 percent), and those who had the disease poorly controlled (8 percent or more).
 Race and Ethnicity > Black or african american; Vulnerable Populations > Racial minorities; Urban populationsend pp

Problem Addressed

Diabetes is a common, costly condition that can have devastating and life-threatening consequences. Although effective treatments are available, many people with diabetes experience complications because they fail to follow recommended self-management strategies. Educational information and reminders via text messages have the potential to increase adherence, but few providers communicate with patients in this manner.
  • A common, costly, and devastating condition: According to the American Diabetes Association, diabetes affects 8 percent of the U.S. population, or 23.6 million children and adults. Compared with the general population, those with diabetes face twice the risk of heart disease and stroke. Other frequent complications include damage to eyes, kidneys, feet, and nerves. In 2004, diabetes was the sixth leading cause of death in the Nation, and it remains one of the 10 most costly medical conditions.1 Diabetes has been a major public health issue in Chicago for many years, and many patients visit University of Chicago Medicine for diabetes treatment.
  • Poor adherence: Many patients with diabetes do not adhere to recommended self-management behaviors or prescribed treatment regimens. As a result, they do not have the disease under control and remain at high risk of complications. Barriers to adherence include time constraints, knowledge deficits, denial, limited social support, inadequate resources, and low self-efficacy.2 Before implementation of this program at the University of Chicago Medicine, physicians in outpatient clinics frequently treated patients whose diabetes had worsened due to their failure to regularly check blood sugar levels, eat healthy foods, exercise, or take their medications as prescribed.
  • Unrealized potential of text messages: Patients of all socioeconomic levels use text messaging frequently,3 making it a potentially effective method for providers and patients with diabetes to communicate outside of standard office visits. Programs that deliver reminders and educational information via text message have the potential to improve adherence among patients with diabetes, especially when coupled with other care management activities. However, few providers currently use this approach.4

What They Did

Back to Top

Description of the Innovative Activity

University of Chicago Medicine developed and implemented a program called CareSmarts in which patients with diabetes received several text messages each day providing education on how to manage the disease and reminders to follow recommended self-care measures. Some messages asked patients to send a reply text indicating whether they were heeding the reminders. A nurse subsequently called those who consistently failed to respond and those whose responses indicated a need for assistance, working with them and their physicians to overcome obstacles to proper self-management. A detailed description of key program components follows:
  • Nurse-led program enrollment: The program served members of the medical center's employee health plan (University of Chicago Health Plan) who had diabetes and received treatment for it at a primary care or endocrine clinic affiliated with the medical center. Two registered nurses employed by the plan called members who had diabetes to explain the program and enrolled those who expressed interest through a Web-based form. As part of this process, the nurse collected and entered the patient's mobile phone number, diabetes care plan, and preferred times for receiving messages.
  • Daily educational messages and reminders, tailored to individual circumstances: Patients typically received three or four messages a day, drawn from a database of more than 1,200 specific messages. Messages generally fell into one of six categories: education about the disease, prompts/reminders related to appropriate self-management, assessments (requests for information from patients, via a reply text), feedback based on the assessments, tips to promote adherence, and encouragement. The mix and content of messages changed every 2 weeks based on a patient's interactions with the system. For example, if a patient's responses indicated he or she was having difficulty managing blood sugar but was exercising regularly, the patient would receive more messages about glucose monitoring and fewer about exercise. Descriptions and examples of each type of message follow.
    • Educational messages: For the first 10 weeks, all participants received a series of educational messages about self-care, broken into five 2-week modules. Messages covered various aspects of self-management (e.g., medications, nutrition, glucose monitoring, foot care, exercise) and living with chronic illness (e.g., navigating the health care system, managing stress). Once participants completed the 10-week curriculum, they had the option of continuing or stopping the educational messages. Examples of educational messages include the following:
      • Glucose monitoring: "A good blood sugar within 2 hours of eating is less than 180 mg/dL. A good fasting (before breakfast) blood sugar is 80 to 125 mg/dL."
      • Living with chronic illness: "Did you know that stress increases your blood sugars? In fact, not only can stress increase your sugars, but high sugars can also increase your stress."
    • Prompts and reminders: These messages reminded patients about the need to take their medication as prescribed, monitor glucose levels, and engage in proper foot care, with the frequency of delivery determined by participants. Examples of prompts included: 
      • Medication: "Time for your medicine!"
      • Foot care: "Check your feet every day. You should look between the toes and bottoms of your feet for cuts, cracks, or anything else out of the ordinary."
    • Assessments: Assessment messages request information via reply text from patients related to self-management behaviors. For example:
      • Medication: "In the last 7 days, how many days did you take all your diabetes medications?" or "Do you need refills of any of your medications?"
      • Glucose monitoring: "Did you test your blood sugar today?" (yes/no)
    • Feedback: These messages reinforced positive behaviors and discouraged negative ones based on responses to the assessment texts, such as:
      • Medication: In response to a patient not taking medications as prescribed, a text might read: "Think about the last time you didn't take your medications. What happened? Think about what you can do to prevent it from happening again."
      • Glucose monitoring: In response to a patient reporting perfect adherence to glucose monitoring for a week, a text might read: "7 for 7, perfect job!"
    • Tips: Tips promoted adoption of a particular behavior. For example:
      • Nutrition: "If it's not in your kitchen, you probably won't eat it. Avoid temptation by not keeping desserts or unhealthy snacks in the house."
      • Exercise: "Lifting small weights at home or while you jog can build muscle and lower your blood sugars. No weights? Use a can of vegetables!"
    • Encouragement: These messages were intended to increase perceived support. For example:
      • Glucose monitoring: "Monitoring blood sugars is not just so your doctor knows how you are doing. Glucose monitoring is a tool for YOU to know how you are doing."
      • Living with a chronic illness: "Everyone feels bad about their diabetes from time to time, even if they have had it for a while. It's what you do with those feelings that counts."
  • Followup calls to nonrespondents and those facing challenges The two health plan nurses received alerts whenever patients consistently failed to respond or when responses fell outside established parameters, with alerts falling into one of two categories ("low" and "high") based on the urgency of the situation. The average participant generated approximately one alert per month, with roughly two-thirds being low-level alerts. Nurses called patients who triggered alerts to help address the situation, as outlined below: 
    • Low-level alerts: These alerts often had to do with care coordination issues, such as a patient indicating the need for a medication refill or a referral to a diabetes educator. A nurse addressed the issue and responded by the next business day; for example, the nurse might coordinate with the patient's primary care team to process a refill or referral and then notify the patient with the relevant details.
    • High-level alerts: These alerts related to more serious problems, such as low self-reported medication adherence or patients consistently not responding to requests for information. Nurses called these patients by the next business day to discuss the problem and offer help in resolving it and then communicated a summary of the call to the primary care team by e-mail. Typically, a member of this team also contacted the patient, scheduling an inperson appointment to address the problem if necessary.

Context of the Innovation

University of Chicago Medicine is a large academic medical center located on the south side of Chicago, a predominantly urban, working-class, African-American community. The medical center includes an acute care hospital and a comprehensive ambulatory care center with an internal medicine practice and an endocrinology practice with substantial expertise in treating patients with diabetes. The University of Chicago Health Plan is a 10,000-member plan comprising primarily employees of the university and the medical center and their dependents. The health plan employs two registered nurses who function primarily in an administrative capacity.

The impetus for CareSmarts came from two medical center physicians who became increasingly aware of the problem of diabetes patients not following recommended self-care measures. The physicians also recognized the potential for text messaging to help them do a better job in monitoring and supporting these patients. Inspired in part by the use of text messages for patients with chronic diseases by health clinics in Uganda and India, the two physicians and several colleagues formed a research team in late 2009 to develop and implement a month-long pilot study. During this pilot test, patients with diabetes received automated text messages (from a set of 20) reminding them about their medications, the need for foot care, and upcoming appointments, with patients sending reply texts indicating whether they were following these reminders. This pilot test showed favorable results related to patient response and medication adherence, which in turn led to the decision to conduct a larger, longer term study to evaluate the impact of a more comprehensive program that included many more messages (including educational material).

Did It Work?

Back to Top

Results

The program improved self-management capabilities among patients with diabetes, leading to better glycemic control, fewer medical visits, lower costs, and high levels of patient satisfaction.
  • Better self-management capabilities: Participants improved adherence to various self-management behaviors known to improve diabetes control, including following a healthy eating plan (with the average participant adhering to such a plan 4.5 days a week before enrolling in the program and 5.2 days a week afterward), monitoring blood glucose (4.3 to 4.9 days a week), and practicing foot care (3.6 to 4.3 days a week). Participants also reported increases in the proportion of days they adhered to their prescribed medication regimen, which rose from 83 percent to 91 percent.
  • Improved glycemic control: Participants' average HbA1c levels fell from 7.9 percent to 7.2 percent. Those with the poorest control at the program's onset experienced the largest average decline, from 10.3 percent to 8.5 percent. By contrast, no change occurred in HbA1c levels in a control group of similar patients who did not receive text messages, including among those with poorly controlled diabetes.
  • Fewer medical visits: Participants experienced a decline in total outpatient visits (for both diabetes and nondiabetes reasons), from an average of 6.37 visits during the 6-month period before the program to 5.04 visits during the 6-month period afterward. Emergency department (ED) and hospital use also trended downward, but these changes did not reach the level of statistical significance.
  • Lower costs: For the average participant, total health care costs fell by $812 during the 6-month period after program participation (compared with the 6-month period before). This figure includes a drop of $1,332 in the costs of outpatient, ED, and inpatient care, which was partially offset by a $520 increase in prescription drug costs. Given that the program cost an estimated $375 per participant, the program resulted in a net cost savings of $437 per participant.
  • High patient satisfaction: Seventy-three percent of participants were satisfied with the program, and 77 percent said that they would participate in a similar program in the future.

Evidence Rating (What is this?)

Moderate: The evidence consists of a comparison of trends in blood glucose levels among program participants and a control group of similar patients who decided not to participate in the program; pre- and post-implementation comparisons of adherence to various recommended self-management behaviors, medical visits, and costs; and post-implementation reports from participating patients on their satisfaction with the program.

How They Did It

Back to Top

Planning and Development Process

Key steps included the following:
  • Securing funding and staff from health plan: In 2011, the research team contacted health plan leaders to discuss the proposed program. These leaders expressed interested in helping with an expanded version of the pilot program described earlier and agreed to provide financial support and assign two existing nurses to handle enrollment and day-to-day program administration.
  • Partnering with software company: One of the lead researchers formed a private company, mHealth-Solutions, to enhance the program's technological capabilities. Under a contract with the medical center, the new company created software programs to make it easier to enroll patients, send text messages, and track patient responses.
  • Expanding and testing content: After reviewing several existing sources of information about diabetes, the researchers drafted hundreds of additional text messages, most of them for the educational modules. They also held several focus groups with clinic patients to gauge their response to the messages, with refinements made based on this feedback (primarily to make the messages more accessible).
  • Training nurses: Before the May 2012 program launch, the research team trained the nurses on their role in the program.
  • Discontinuing program due to lack of health plan funding: Following the 6-month study period, the health plan experienced a change in leadership and declined to continue funding the program.

Resources Used and Skills Needed

  • Staffing: Two existing health plan nurses handled most of the program's day-to-day administration, spending roughly 5 to 10 hours a week on related activities. Additional contributors included members of the research team, physicians and nurses at the two medical center clinics that treat diabetes patients, and programmers at the software company.
  • Costs: The most significant expenses included software development, telecommunication charges, and nurse time. The 6-month program cost an estimated $375 per participant ($150 for technology and $225 for staff).
begin fs

Funding Sources

University of Chicago Health Plan
end fs

Tools and Other Resources

More information about mHealth Solutions is available at: http://www.mhealth-solutions.com.

Adoption Considerations

Back to Top

Getting Started with This Innovation

  • Start small: Providers sometimes avoid implementing technology-oriented programs because they assume such programs will be expensive and time consuming. To address these concerns, providers should consider starting with a very basic program, such as a nurse manually sending out 5 or 10 text messages to a small group of patients. If patients are receptive, program leaders can add additional messages and make the transition from manual to automated messaging.
  • Solicit patient input: Prior to implementing this type of program, providers should ask potential participants what sort of information they are most interested in and tailor the messages accordingly. Through focus groups, leaders of this program found that patients were very interested in information about foot care and expanded the content on this subject in response to this feedback.
  • Tailor content to local interests: To the extent possible, messages should occasionally include information specific to the local area. For example, instead of just advising patients to buy healthy foods at the supermarket, a message could let them know about a particular store in the area that is having a sale this week on broccoli.
  • Make personal connection at enrollment: While participants could have enrolled on their own, program leaders decided to have the nurses complete the online form with patients, either in person or over the phone. This approach allowed patients to develop a relationship with the nurse who would later be monitoring their responses and supporting them in managing diabetes. Establishing this relationship may increase the likelihood that patients will heed text reminders, read educational material, and respond to the nurse's phone calls.

Sustaining This Innovation

  • Improve technology over time: Ongoing improvements in technology may present opportunities for program enhancements. Although CareSmarts was not integrated with the health plan's electronic health record (EHR) system, program leaders believe that such integration could have been beneficial. For example, appointment information in the EHR could have automatically triggered reminder texts, and patient responses to texts could have been made available to primary care doctors through the EHR.
  • Consider expansion to other chronic diseases: Assuming the program produces positive results with diabetes patients, providers could consider expanding it to patients with other chronic diseases. The same general framework (e.g., education, medication reminders, self-care tips) can be applied to any chronic condition.

More Information

Back to Top

Contact the Innovator

Shantanu Nundy, MD
Managing Director
Evolent Health
800 N. Glebe Road, Suite 500
Arlington, VA 22203
(571) 389-6000
E-mail: snundy@evolenthealth.com

Innovator Disclosures

Dr. Nundy cofounded mHealth Solutions, the company that provided technological services for CareSmarts under a contract with University of Chicago Medicine, as described in the Planning and Development Process section. While the program was in operation, Dr. Nundy was not employed by mHealth Solutions and had no formal relationship with the company.

References/Related Articles

Dick JJ, Nundy S, Solomon MC, et al. Feasibility and usability of a text message-based program for diabetes self-management in an urban African-American population. J Diabetes Sci Technol. 2011;5(5):1246-54. [PubMed]

Nundy S, Dick JJ, Chou CH, et al. Mobile phone diabetes project led to improved glycemic control and net savings for Chicago plan participants. Health Aff (Millwood). 2014;33(2):265-72. [PubMed]

Nundy S1, Dick JJ, Goddu AP, et al. Using mobile health to support the chronic care model: developing an institutional initiative. Int J Telemed Appl. 2012;2012:871925. [PubMed]

Footnotes

1 Cohen M. Kaiser Commission on Medicaid and the Uninsured: an overview of Medicaid enrollees with diabetes in 2003. October 2007. Available at: http://www.kff.org/medicaid/upload/7700.pdf (If you don't have the software to open this PDF, download free Adobe Acrobat ReaderĀ® software External Web Site Policy.).
2 Gazmararian JA, Ziemer DC, Barnes C. Perception of barriers to self-care management among diabetic patients. Diabetes Educ. 2009;35(5):778-88. [PubMed]
3 Pew Internet & American Life Project. Mobile access 2010. Available at: http://www.pewinternet.org/Reports/2010/Mobile-Access-2010.aspx
4 Dick JJ, Nundy S, Solomon MC, et al. Feasibility and usability of a text message-based program for diabetes self-management in an urban African-American population. J Diabetes Sci Technol. 2011;5(5):1246-54. [PubMed]
Comment on this Innovation

Disclaimer: The inclusion of an innovation in the Innovations Exchange does not constitute or imply an endorsement by the U.S. Department of Health and Human Services, the Agency for Healthcare Research and Quality, or Westat of the innovation or of the submitter or developer of the innovation. Read more.

Original publication: July 30, 2014.
Original publication indicates the date the profile was first posted to the Innovations Exchange.

Last updated: August 13, 2014.
Last updated indicates the date the most recent changes to the profile were posted to the Innovations Exchange.