Improving Type II Diabetes Education and Self-care Management in Older Adults Utilizing a Two-way SMS Coaching and Education Program (Care4Life)
HealthInsight – Utah Beacon Community
This program supports diabetes self-management by connecting patients to a low-cost, convenient, and personalized coaching and education text messaging tool, Care4Life. Care4Life sends text messages through a patient’s cell phone with education tips and reminders that support blood glucose monitoring and encourage progress towards personal health goals. Working with 19 local primary care clinics through the Utah Beacon Community Program, 450 adult patients were enrolled in the 6-month Care4Life program. One third of these patients were enrolled in a randomized control study and the remainder were enrolled in a non-study group as part of a larger quality improvement effort.
The project demonstrated that an older, chronically ill patient population with lower technology proficiency (versus the U.S. population as whole) can successfully use a two-way SMS behavioral support program for Type II Diabetes self-management. Moreover, users of Care4Life reported very high satisfaction with the program. The program evaluation assessed changes in diabetes control (primarily HbA1c control) as well as patient satisfaction and improvement in diabetes knowledge using mid-point and exit surveys. Patients with high HbA1c levels showed greatest improvement, while all patients reported high satisfaction with the program. The program results indicate that a diabetes text-messaging intervention can be effective in a targeted population.
WHO WAS INVOLVED
Voxiva, Inc.; HealthInsight, a Beacon Community Program; 19 primary care clinics in the Salt Lake City Metropolitan Area
WHAT THEY DID
Problem Addressed: Approximately 14.1 percent of adults over age 50 in the Salt Lake City Metropolitan Statistical Area have diabetes. The key to living well with diabetes is careful and consistent management. Day-to-day control of diabetes symptoms prevent other serious health complications down the road and enhances independence, particularly for older adults. However, optimal control of diabetes is difficult to attain, as evidenced by the high morbidity and mortality associated with the disease. Diabetes is the leading cause of blindness among Utah adults ages 25 to 74, and over half of all non-traumatic lower extremity amputations and dialysis treatments occur in persons with diabetes. While 61 percent of Utah adults with diabetes perform daily self-monitoring of blood glucose, this key self-management practice varies by age. A study conducted by the Utah Department of Health found that older adults are much less likely to engage in self-management activities in general. Results showed that advanced age (50 or more years) lowers the odds that a patient will engage in daily blood glucose monitoring by more than one-half. This study also found that diabetes education plays a positive role in increasing the likelihood that patients will engage in self-monitoring: diabetes education nearly triples the odds of daily blood glucose monitoring and nearly doubles the odds of daily feet checking. These statistics speak to the need for innovative and more widely distributed approaches to diabetes education and coaching.
Nurse care management has been shown to improve patient self-management and engagement, but this strategy is costly and requires practices or health plans to invest in a disease management program. Given the prevalence of diabetes and limited resources, there is an ongoing need for cost-effective alternatives for diabetes care that are capable of reaching large numbers of patients.
Patient Population: The patient population for this intervention was drawn from 19 primary care clinics who serve patients from diverse socio-economic backgrounds in the greater Salt Lake County area.
Description of the Program:
Care4Life is a two-way text messaging program that sends text messages through a patient’s cell phone with reminders that support blood sugar monitoring and education tips that encourage progress towards personal health goals. Working with 19 local primary care clinics through the Utah Beacon Community Program, 450 adult patients were enrolled in the 6-month Care4Life program. Participants received 1-7 texts per day, depending on what protocols they selected at enrollment. Text message topics include education tips and reminders that support blood sugar monitoring, healthy eating, exercising, and encouraging progress towards health goals.
Participants were identified via a customized query of each clinic’s electronic medical system with inclusion criteria:
- 18 to 65 years of age
- diagnosed with Type II diabetes
- treated at one of 19 primary care clinics in the Beacon Community Program
- HbA1c >8% in the past year
- Not pregnant
- Own a cell phone with text messaging capability
- English or Spanish speaking
Using this method, HealthInsight identified ~2,100 patients eligible for inclusion in the program. Patients were mailed a personal letter from their physician or clinic inviting them to join the project and consent via a website. A randomization function was built into the consent website so that patients selected for the intervention group were immediately forwarded to the Care4Life enrollment form. Those in the control arm had no further obligations. Both groups received a gift certificate as an incentive. HealthInsight staff conducted follow-up calls to patients who did not respond to the invitation letter and assisted those who needed help over the phone.
Day-to-day operations of the Care4life program can be broken out into four distinct phases:
Provider recruitment and enrollment – This phase consists of in-person meetings with clinic staff to explain the technology tool, the program design, data needs, and clinic staff commitments required. Provider recruitment and enrollment is conducted by the program manager, clinic facilitators, and medical director.
Patient recruitment and enrollment – This stage includes acquiring patient data from the clinics’ electronic medical records (which varied in type and functionality between settings), generating mass mailing communications, triaging patient telephone inquiries, conducting follow-up calls, and assisting patients with the enrollment process (in-person and by phone). Patient recruitment and enrollment is conducted by the study manager, administrative support staff, and clinic staff.
Implementation – This stage consists of monitoring patient engagement and attrition, as well as generating and delivering quarterly reports to clinics. These reports are prepared by the study manager, administrative support staff, and clinic facilitators and show how patients are interacting with the program and if their self-reported blood sugars are changing.
Evaluation – This stage focuses on acquiring, compiling, and analyzing data on clinical measures drawn from patient electronic medical records; aggregating and analyzing the self-reported blood sugar and satisfaction data from the Care4Life system; and following up with clinic staff to acquire missing data. Data acquisition and analysis was conducted by three healthcare analysts and the project manager.
Overall Results: The project demonstrated that an older, chronically ill patient population with lower technology proficiency (versus the U.S. population as whole) can use a two-way SMS behavioral support program for Type II Diabetes self-management to improve their HbA1C levels. Moreover, users of the program reported very high satisfaction with the tool.
Reduction in HbA1c in Subset of Patients. Program results showed that HbA1c reduction and success with the Care4Life program are more likely for persons with high HbA1c.
High Patient Satisfaction. Patients reported high satisfaction with the program after 90 and 180 days.
Provider Satisfaction. Staff from participating clinics were very enthusiastic about the Care4Life program, both at the outset and after viewing early results.
Limitations: This project was a randomized controlled trial with a relatively small sample conducted over a short duration. Many patients were lost to follow-up due to gaps in clinical data supplied by participating clinics.
ADOPTION CONSIDERATIONS AND LESSONS LEARNED
Context of the mHealth Program: HealthInsight is a community-based non-profit organization based in Salt Lake City whose mission is to improve health and health care in Utah. HealthInsight is also the recipient of a Beacon Community grant, the focus of which is to improve care and outcomes for patients with Type II diabetes in the Salt Lake Metropolitan area.
The Utah Beacon Community consists of 60 primary care clinic partners that have committed to demonstrating measurable improvement in diabetes outcomes over a three-year period. This project was implemented among a group of engaged primary care providers interested in improving care delivery for adults with diabetes and in the context of a larger quality improvement effort. Participating clinics recognized that improving blood sugar control would require changes in patient behavior outside of the clinic setting. This program aligns with that goal, and therefore the reception from clinics approached for this project was highly positive.
Several local and national initiatives began during this same period. Utah passed legislation creating a capitated, accountable care structure in the Medicaid program. In addition, several local health plans began experimenting with medical home projects. Finally, in order to financially reward improvement in diabetes care, the Utah Beacon program implemented a pay for performance incentive that rewarded providers for achieving measurable improvements in process and outcome measures, including blood sugar control. All of these factors contributed to a fertile environment for this project, but also created challenges in that practices were occupied with competing priorities that dominated their time and required changes in workflow and data reporting during this period. Minimizing the impact on clinics therefore became a high priority when developing the implementation strategy. For this effort, HealthInsight handled recruitment and enrollment for the clinics participating in the study rather than requiring clinic staff to enroll patients during office visits, which most felt would be unfeasible given existing demands and reimbursement models. However, in a different payment environment, working the intervention into the workflow will likely be more feasible.
Planning and Development Process: After meeting with clinic staff, HealthInsight developed a program design that would maximize enrollment, minimize the impact on clinic staff, and provide a sound basis for evaluating the intervention. The planning and development process required Institutional Review Board application and approval. HealthInsight created a protocol, recruitment materials, an online consent process, and established an 8-person internal team with human subjects training. Recruitment occurred through in-person meetings with clinic managers and in some cases, through presentations to the entire clinic staff at staff meetings. Discussion centered on the logistics of how the project would be operationalized and what learnings HealthInsight hoped to generate from the project. HealthInsight presented this project at community meetings and at a “Learning Session” for the Beacon Program clinics prior to the project launch to answer questions, as well as to publicize and further describe the project. Some clinic staff requested a test period to try out the technology prior to committing to participation. These partners were provided test codes and allowed to use the program prior to launch, in order to gain buy-in.
Getting Started with This Innovation: Critical initial steps to getting this program started included:
- Primary Care Clinic buy-in and support. For this project, HealthInsight had the advantage of an already engaged cohort of clinics that were actively participating in an ambitious quality improvement effort, with the stated goal of improving outcomes for patients with Type II diabetes on a 3-year timeline. Clinic buy-in is absolutely essential given the need to interact with patients and access their personal health information.
- Use of IRB-approved staff. Clinics were too short-staffed to have done the medical record queries on their own without the help of IRB-approved staff, making this step essential. IRB-approved staff directed the IRB approval process for the project, allowing project staff to access patient data for recruitment and evaluation.
Business Model, Program Costs, Resources Used, and Skills Needed
- Study Manager (.4 FTE)
- Medical Director (.05 FTE)
- Administrative Assistant (.4 FTE)
- Healthcare Analysts (3) (.1 FTE each)
- One analyst for the evaluation
- Two analysts to acquire and structure the data
- Beacon Program Director (.01 FTE)
- Project Coordinator (.1 FTE)
- Programmatic funding totaled $100,000, 50% of which went to personnel costs; 20% went to operating expenses, IRB costs, travel and meeting hosting expenses; 20% went to patient incentives; and10% went to contract services for Web programming and consulting services.
- The per patient per month cost for the Care4Life program is $8.30, or $24.90 per patient for the program for three months.
Sustaining and Scaling Use of the Technology:
Patients in the 19 Beacon clinics in the Care4Life program continued to use the program 12 months after the intervention concluded. With 19 primary care clinics participating in the project, each site has differing levels of readiness to adopt mHealth technologies for patient support and varied the integration of the program into clinical practices. For example, one of the participating clinics devised a strategy for fully implementing the recruitment and enrollment process into the workflow by incentivizing medical assistants to sign patients up over the phone during scheduled outreach calls, resulting in the enrollment of 40 new patients.
Project findings highlight the potential synergy of the Care4Life program in particular, and mHealth programs more generally, with new models of care and reimbursement such as the Patient-Centered Medical Home and Accountable Care Organization.
Other Considerations and Lessons Learned:
The process of implementing the program yielded many insights as did the program outcomes. Moreover, working with providers for the purposes of this program helped advance the conversation around the need for more patient engagement and the promise of new technology tools to help support chronic disease patients in the difficult task of day-to-day management. For the project team, valuable insight was gained into the practical barriers that prevented wider adoption of such innovations as well as the overlooked opportunities that may accelerate it in the near future. The project team also learned a great deal from the patients in the program, who provided a window into the myriad of challenges associated with managing a complex health condition, particularly with limited financial resources and/or care management support.
Maximize Patient Recruitment through Multiple Enrollment Processes. The project utilized multiple forms of recruitment and enrollment processes for participants, including mass mailing communications, provider referral, in-person enrollment fairs, follow-up calls, web-based enrollment, with varying degrees of success. In-clinic sign-up events, called Healthy Living Days, proved highly successful in recruiting patients. Often one-on-one time, either in-person or over the phone was needed to assist participants with enrollment.
Leverage Existing Programs and Partnerships to Take Advantage of Previous Patient-Provider Relationships. For this project, HealthInsight had the advantage of an already engaged cohort of clinics that were actively participating in an ambitious quality improvement effort through the Beacon Community Program, with the stated goal of improving outcomes for patients with Type II diabetes on a 3-year timeline. These relationships and common goals not only increased provider and staff engagement and project support, but also improved the introduction, acceptance, and long-term sustainability within the clinics.
Minimize Impact on Clinicians – Build Enrollment and Program into Workflow. This project occurred during a tumultuous time of transition for primary care clinics in the region. Several competing initiatives (e.g., meaningful use, EMR adoption, changing payment models such as the patient centered medical home) occupied their attention. Utilizing HealthInsight staff for EMR queries, patient recruitment, follow-up calls, and evaluation minimized the impact on providers. In the long term, such interventions should be integrated into the workflow to succeed. One of the participating clinics devised and implemented a plan that utilized medical assistants to find, enroll, and follow-up with patients, which may provide a blueprint for other primary care providers.
Use Data to Motivate and Empower Clinicians. The generation of patient data from the Care4Life program, as well as patient satisfaction data, can provide meaningful value add to clinicians’ practices. Quarterly reports to providers showed how patients were interacting with the program and if their self-reported blood sugars were changing.
Incentives Not Primary Motivator for Patient Participation. The project team observed that the amount of the incentive for patient enrollment in the program did not seem to impact the enrollment return. Seven weeks into the recruitment process the incentive was increased from $20 to $35 dollars in response to a lower than expected return, but the enrollment rate did not alter. The monetary reward was not necessarily the primary motivator for participation in the program or study. (Or that the incremental increase in the incentive was not large enough to attract patients who would have only joined if a larger incentive was offered.)
Additional Need for Robust Care for Patients with Poorly Controlled Diabetes. While recruiting for the project, it became apparent that there is a need for more robust care and population management for patients with poorly controlled diabetes in the independent physician practice setting. A large proportion of patients enrolled in the project did not receive HbA1c testing at recommended intervals during the time period of this study (nine months).