Continuous Monitoring Outperforms Fingerstick Testing in Patients With T1D, High Glycated Hemoglobin

New evidence supports the use of continuous glucose monitoring as a means to help control glycated hemoglobin levels in people with type 1 diabetes (T1D).

The study, which compared continuous monitoring and automated alarms to patient-initiated finger-stick monitoring, found the former led to significantly lower glycated hemoglobin levels. The report, called the FLASH-UK Study, was published in the New England Journal of Medicine.

Corresponding author Lalantha Leelarathna, PhD, of the Manchester University NHS Foundation Trust, in England, and colleagues, said many patients with T1D struggle to monitor their blood glucose levels and adjust their insulin dosage accordingly.

“Consequently, in most persons with type 1 diabetes, the glycated hemoglobin level, a marker of glycemia, remains above target levels and increases the risk of major complications,” they wrote.

Continuous glucose monitoring systems were designed to eliminate the need for constant fingerstick testing by monitoring a patient’s glucose levels either intermittently or in real-time, and giving patients the option to set alarms if they reach certain high or low blood glucose benchmarks.

In the new study, Leelarathna and colleagues sought to find out whether such a system would have an impact on the risk that a patient experienced glycemia. They focused on patients with high glycated hemoglobin levels, as those patients are at a particularly high risk of complications.

The investigators recruited 156 people with type 1 diabetes glycated hemoglobin levels between 7.5% and 11.0%. Those patients were split into 2 equal cohorts of 78 patients. Half of the patients were given continuous glucose monitoring devices that intermittently scanned their glucose levels. The other half were asked to continue monitoring their own blood glucose levels using fingerstick testing. The patients had an average age of 44, had an average duration of disease of 21 years, and 44% were women.

At baseline, the mean glycated hemoglobin levels of the 2 groups were similar, 8.7% (0.9%) in the continuous monitoring group and 8.5 (0.8%) in the fingerstick group. After 24 weeks, however, the levels dropped to 7.9% and 8.3% (0.9%), respectively, Leelarathna and colleagues found.

In addition to reducing glycated hemoglobin levels, the investigators also found that patients in the continuous monitoring group spent more time within their targeted glucose ranges, and had a reduced hypoglycemia burden.

In terms of safety, the authors said 2 patients in the fingerstick group experienced episodes of severe hypoglycemia, and 1 patient in the continuous monitoring group experienced a skin reaction to the device’s sensor.

The investigators said that their data suggest that not only is intermittent scanning with a continuous glucose monitor more effective than fingerstick monitoring; it also performs similarly well to systems that monitor glucose in real-time.

Another question facing patients and physicians is whether alarms provide a benefit to patients. The authors said previous studies have suggested that the answer is “yes,” although such research was done prior to the invention of newer devices that offer continuous scanning.

While questions of the strengths and benefits of different types of continuous modeling systems linger, the authors said their study makes a strong case for automated systems.

“The benefit with respect to glycated hemoglobin levels was robust in a range of sensitivity analyses, including imputations under plausible missing-at-random and missing-not-at-random assumptions,” they wrote.

One remaining question, however, is how such systems affect the economics of monitoring people with T1D. The investigators said a cost-effectiveness analysis of the data is currently underway.

Reference

Leelarathna L, Evans ML, Neupane S, et al. Intermittently Scanned Continuous Glucose Monitoring for Type 1 Diabetes [published online ahead of print, 2022 Oct 5]. N Engl J Med. 2022;10.1056/NEJMoa2205650. doi:10.1056/NEJMoa2205650