番茄社区app

ORLANDO -- In a three-way head-to-head comparison of continuous glucose monitors (CGMs), all performed well in an outpatient setting in type 1 diabetes patients, researchers reported here.

Comparing the Dexcom G5, Abbott Freestyle Libre Pro, and the recently approved Senseonics Eversense system, all CGMs produced higher than average mean absolute relative differences (MARD) compared with averages reported during in-clinic studies, said Rabab Jafri, MD, of Massachusetts General Hospital in Boston, and colleagues.

However, the Eversense system achieved the lowest nominal MARD in the three-way comparison, the group reported at the American Diabetes Association annual meeting (Eversense vs Libre Pro P=0.004; other comparisons P=NS):

• Eversense: MARD of 14.8%
• Dexcom G5: 16.3%
• Freestyle Libre Pro: 18.0%

The Eversense system also performed the best in hyperglycemia, while the G5 performed the best in hypoglycemia. The Libre Pro performed particularly inaccurately in the hypoglycemic range:

• Eversense MARDs: 24.9 &plasma glucose <70 mg/dL); 16.1 (70-120 mg/dL); 12.9 (>120 mg/dL); 12.8 (>180 mg/dL)
• G5 MARDs: 23.6; 18.5; 14.2; 13.3
• Libre Pro MARDs: 36.1; 17.8; 16.3; 17.3

"Eversense has accuracy that is at least as good as the G5, which has been used to drive the bionic pancreas," Jafri noted. "The Eversense may be used to provide glucose values to artificial pancreas devices."

Further assessed in a two-way comparison, the Eversense CGM still came out with a significantly lower MARD versus the G5 system (MARD 15.1% vs 16.9%, P=0.008).

The study included 23 adults with type 1 diabetes who wore all three CGMs at over a 6-week study period. Patients conducted fingersticks twice per day to calibrate the G5 and Eversense systems, while the Freestyle Libre Pro doesn't require fingersticks. All participants used an insulin pump for at least 6 months, and all had used a CGM for at least 1 month prior to inclusion.

The three-way analysis assessed 829 sets of plasma glucose (PG) values measured with the Nova Biomedical StatStrip Xpress meter, which has a MARD of 6.3%, that were compared against the CGM readings that were given 5 minutes prior to the PG values. The two-way analysis between the Eversense and G5 systems compared a larger dataset of 2,277 readings against PG values.

Meeting attendee Irl Hirsch, MD, of the University of Washington in Seattle asked what providers should back to their clinics from these findings.

"From the artificial pancreas perspective, I think it's important that patients have a choice in the CGM that they use," Jafri stated. Ideally, there could be more than one CGM option as part of the artificial pancreas algorithm in the future, she added.

Comment