A Direct-to-Patient Alert for Glycated Hemoglobin Screening Using Prediction Modeling and Mobile Health (mHealth)

Project Final Report (PDF, 495.41 KB) Disclaimer

While “cold” texting of patients to suggest screening for diabetes is feasible from a technological and governance perspective, more research is needed to inform how to better engage patients.

Project Details -
Completed

Type 2 diabetes (T2D) is a pervasive public health issue in the United States. Of those adults with T2D, approximately a quarter of cases are undiagnosed. Onset can be difficult to predict due to a lack of symptoms early in the disease course. Hyperglycemia (elevated blood sugar) is the defining characteristic of T2D and can be present for years before a diagnosis. Screening, diagnosis, and management of T2D may be done via hemoglobin A1c (HbA1c), a blood test that is reflective of an individual’s average blood sugar over 3 months. It is thus an easy and accurate method for determining if a patient has experienced hyperglycemia regularly, possibly indicating T2D.

In a previous study, the researchers created a clinical decision support (CDS) calculator for predicting current HbA1c levels in patients without any previous evidence of diabetes or hyperglycemia. This calculator was created using data on adult patients who have undergone a serum HbA1c test prior to evidence of hyperglycemia or diabetes recorded in the electronic health record (EHR). In the current research, the investigators studied the “direct-to-patient alerts” method for targeting high-risk patients via text messages and ordering HbA1c in patients who are interested.

The specific aims of the research were as follows:

  • Develop an mHealth text message intervention to improve the early identification of patients with hyperglycemia.
  • Evaluate the impact of a text message, mHealth glycated hemoglobin screening program in the Department of Family Medicine.
  • Perform telephonic surveys on 100 patients (50 who completed screening and 50 who did not complete screening).

To develop the content of the text messages, researchers conducted six focus groups with adults without diabetes. The researchers displayed a series of text messages and asked the participants to react to the message content and suggest improvements. The resulting text message intervention was evaluated by examining the number of individuals who responded to the text, had an HbA1c test ordered, completed the test, and ultimately had an increased risk of diabetes (HbA1c greater than or equal to 5.7%). Over the course of several months, the team sent text messages to 500 individuals and the majority either opted out to receive messages or did not respond. Overall, HbA1c tests were ordered on 40 patients, with 13 individuals completing the test and, of those, 1 person was found to have an HbA1c greater than or equal to 5.7%.

While the research team demonstrated that “cold” texting of patients for quality improvement is feasible from a technological and governance perspective, the percentage of patients interested in HbA1c was disappointing, especially since only 13 individuals completed testing. The researchers identified several reasons for low engagement: the study occurred during the COVID-19 pandemic, and people may have been reluctant to go to a provider office for testing; SPAM sent via text messaging has increased dramatically and likely made patients reluctant to participate; and many patients would like their physicians to be involved in some capacity. Their findings suggest that patients accept text messages that alert them to a higher risk for elevated HbA1c but need clarification of the new processes in which they are asked to engage. Future research is necessary to determine whether these text messages lead to meaningful health behaviors in practice and whether this could apply to other opportunities.

Using Direct-to-Patient Technology and Clinical Decision Support to Increase Type 2 Diabetes Screening

Using Direct-to-Patient Technology and Clinical Decision Support to Increase Type 2 Diabetes Screening

A low-cost, novel direct-to-patient CDS tool that identifies patients at high risk of type 2 diabetes and offers them a screening test could increase the number of patients screened and save physicians’ time.

Principal Investigator: Wells, Brian J.
Organization: Wake Forest University Health Sciences
Research Profile: A Direct-to-Patient Alert for Glycated Hemoglobin Screening Using Prediction Modeling and Mobile Health (mHealth)
Funded Amount: $295,774

Identifying patients with elevated blood glucose levels

Type 2 diabetes (T2D) and pre-diabetes are significant public health problems. Characterized by elevated blood glucose levels, T2D and pre-diabetes, with the resulting complications that occur when untreated, can result in significant medical costs. Those at risk for T2D may be screened with serum hemoglobin A1C (HbA1c) testing, which reflects a patient’s average blood glucose level over the prior 3 months. Despite current screening guidelines, many high-risk patients do not get screened, contributing to approximately a quarter of T2D cases going undiagnosed. To assist physicians in identifying patients at risk, Dr. Brian Wells and his Wake Forest research team previously developed an HbA1c risk calculator for predicting which patients without previous symptoms of diabetes or hyperglycemia would have an elevated HbA1c. One method to increase screening is to alert providers when patients are at risk as identified by the risk calculator.

In primary care, one of the best things that we can do to improve patient outcomes is identify and manage chronic conditions early.”
- Dr. Wells

Offering screening to patients through text messaging

However, as a family physician Dr. Wells understands that busy PCPs may experience “alert fatigue” from alerts such as these. As such, he and his research team are developing a direct-to-patient CDS tool to identify at-risk patients and directly offer them screening. The tool will access patient data in the EHR and identify patients at risk using characteristics validated in the calculator. Once a patient is identified, they will receive a text message offering them testing through their primary care physician’s office. If the patient agrees to the screening, a lab order will be automatically placed. Results from the screening will be sent to both the patient and the patient’s physician.

Changing the way services are delivered

Using direct-to-patient technology is a novel strategy for reaching patients for screening recommendations. Dr. Wells thinks that by bypassing the clinician, more patients with be screened; those identified as hyperglycemic will be offered earlier treatment and thus better outcomes. Dr. Wells hopes that by “going directly to the patient and empowering them to get screened” the tool will help reconnect high-risk patients to the healthcare system. This low-cost, low-risk tool has the potential to be adapted for use with other chronic health conditions.