Continuous glucose monitoring gives a person repeated glucose readings across the day. Insulin use is one reason to consider it. A new trial asks what it can offer adults with type 2 diabetes who are not using insulin.
The answer is worth a practice's attention. So is the work that comes after a sensor is prescribed.
What CONNECT found
The CONNECT trial, published October 14 in the New England Journal of Medicine, randomly assigned 283 adults at 22 U.S. primary care practices to CGM or routine care for 26 weeks.
Average A1c fell from 8.8% to 7.2% in the CGM group and from 8.7% to 8.0% with routine care. The adjusted difference between groups was 0.9 percentage points in favor of CGM.
The CGM group also spent 21 percentage points more time in the study's glucose range of 70 to 180 mg/dL. That is about five more hours per day. It is a difference between groups, not five hours added to every participant's day.
These are glucose outcomes. They do not establish that CGM prevented heart attacks, kidney disease or deaths in this trial.
Who the finding describes
Participants were adults with type 2 diabetes who were not using insulin. Entry required an A1c of at least 7.5%, according to the study's clinical reporting. This was not a trial of every person with type 2 diabetes, regardless of glucose control.
Dexcom funded the study, which evaluated its G7 device. That funding matters when reading the evidence. It does not erase the randomized comparison, and it should not disappear from a summary of it.
Trial participants also had access to sensors and study follow-up. Reported adherence was high. The clinical report notes that sensor provision and visit frequency may have helped, and that use and benefit could be lower in routine care.
A practice cannot assume it will reproduce the result by writing the prescription alone.
Start with the question the patient needs answered
For an appropriate patient, a useful discussion might begin with a question: what do we hope these readings will help us understand or change?
The answer could concern glucose patterns around meals, activity or a change in treatment. That question should be agreed with the clinician rather than inferred from a single high reading.
We think this is a better starting point for a workflow than asking how many patients can be given a device. A device count tells a practice how much it distributed. It does not tell the practice whether anyone understood the information or used it in a care plan.
This is an operational proposal, not a treatment recommendation for an individual patient.
Decide what happens after the readings arrive
Before a practice broadens CGM use, it can make the review plan concrete.
Name the person who helps with setup and confirms that the patient knows how the information will be used. Explain how the patient should ask for help, including what to do with an urgent concern. Do not leave them to assume that a person is watching the data continuously.
Agree when the clinician will review the information and what will be reviewed at that visit. Record the question, the interpretation and the next step together. If the plan changes, the patient should be able to understand what changed and why.
These are suggestions for designing the service around the device. CONNECT did not test this checklist against another workflow, and it does not prove that a particular software arrangement is necessary.
Measure more than whether the feed connected
A successful connection is a technical result. A patient understanding the plan is a different result. A change in glucose measures is another.
A small evaluation can keep those questions separate. Did the patient receive and use the device? Could the clinician access the intended reporting period? Did the planned review happen? Was the resulting care plan communicated?
Glucose outcomes belong beside those process measures. So do the practical barriers a trial may soften, such as affordability, supply continuity and help with setup.
The trial's findings also do not decide what an individual insurance plan covers. Evidence, eligibility and the patient's actual access need their own checks.
At Thyra, we think a workflow should make the path from information to a clinical decision clear. We are not claiming that CONNECT evaluated Thyra or that every device connection is available in every practice environment.
CGM before insulin deserves a serious conversation. That conversation should include who will help the patient use what they learn.
Questions before expanding CGM use
Does CONNECT make CGM a universal recommendation before insulin?
No. It provides randomized evidence in the population studied, including adults with elevated A1c. Individual use remains a clinical decision, not a conclusion from a headline.
Did the CONNECT trial establish long-term complication prevention?
The reported results concern 26-week glucose outcomes. They should not be presented as proof of fewer cardiovascular events, kidney complications or deaths.
What should a practice settle before adding a new CGM patient?
The purpose of use, setup support, the review plan and the route for questions. Those are proposed workflow questions, not a validated CONNECT implementation checklist.
Sources
NEJM primary paper, October 14