Personal CGM use in hospitals feasible, preferred by nurses and patients
Click Here to Manage Email Alerts
Key takeaways:
- A new policy implemented by Stanford Health Care allowed people to use their own CGMs in the hospital.
- Most nurses and patients preferred the new policy to mandatory finger-stick blood glucose monitoring.
A change to hospital policy allowing patients to use their own continuous glucose monitors instead of finger-stick blood glucose monitoring was met favorably by a majority of patients and nurses, according to study findings.
In November 2022, Stanford Health Care put a new policy into effect allowing the use of personal CGM for inpatient care. A subcommittee of the system’s inpatient diabetes task force created a policy addressing nursing protocol for validating CGM with finger-stick blood glucose testing and the use of inpatient CGM.
“Our patients asked if they could use their CGMs, as they were frustrated with having to do frequent finger sticks, even though they had these devices,” Michael Hughes, MD, instructor in the division of endocrinology, gerontology and metabolism at Stanford University School of Medicine, told Healio. “Some of our nurses noticed this issue too, so we sought to make it happen as a team. We found a way to reduce the number of finger sticks while still keeping an eye on the CGMs to make sure they were accurate. In the end, patients were glad to see fewer finger sticks and have their CGMs become a part of their care.”
Validating CGM measures
In a study published in Diabetes Care, Hughes and colleagues collected attempted validations of paired CGM and finger-stick blood glucose readings from the start of the policy change in November 2022 to August 2023. In order for a patient to use CGM, the device must meet the validation criteria. The criteria stated a CGM reading needed to be within 20% of a finger-stick measure if the finger-stick blood glucose was 70 mg/dL or higher or within 20 mg/dL if the finger-stick measure was less than 70 mg/dL. Mean absolute relative difference was calculated between the CGM and finger-stick measures. Accuracy was also assessed by placing measurements on a Clarke error grid. Nurses caring for patients using CGM and patients who used CGM were invited to complete a survey on the policy.
Hughes said the new protocol expands on policies implemented at other institutions by permitting the use of multiple types of CGMs.
“We also aimed to create a system that would be easy for a new intern to order, even during a late-night admission,” Hughes said. “Ultimately, this was used successfully across a fairly broad variety of hospitalizations.”
Most patients prefer CGM
There were 135 patients who used their own CGM, and 185 inpatient encounters recorded (mean age, 58 years; 43% women; 56% white). Of the patients, 51.9% had type 2 diabetes and 27.4% had type 1 diabetes. The most commonly used CGM was the Dexcom G6 (44.3% of patients) followed by the Abbott FreeStyle Libre 2 or FreeStyle Libre 14-day sensors (43.2% of patients).
Of 1,506 validation attempts, 87.8% met the criteria to allow CGM-based insulin dosing. The mean absolute relative difference between CGM and finger-stick measures was 9.6%. When validations were placed on a Clarke grid, 99.3% fell within the grid’s two most accurate zones, A and B.
Twenty-seven nurses and 46 patients completed surveys on the policy. Of the nurses, 74% said glucose management with the new policy was effective, 67% reported it was easy to use, and 63% described it as efficient. CGM was preferred over finger sticks by 16 of 20 nurse respondents.
Of the patients who completed the survey, 63% had a positive experience with the new policy, 63% said they had positive interactions with the nursing staff about CGM use and 63% said the number of diabetes management-related interruptions was reasonable. Twenty-five of 27 patients said they preferred the CGM policy over finger-stick measurements.
Moving forward
According to Hughes, the study lacked a number of data points, including the optimal timing for validation frequency and raw CGM data. However, he said the new CGM policy is feasible and could be implemented at other institutions.
“Exploring pediatrics is a very important next step, and it’s something we’re eager to pursue,” Hughes said. “The unique dynamics in pediatrics — balancing the patient’s individual autonomy with the key roles of parents and guardians — are likely to offer new insights.”
For more information:
Michael Hughes, MD, can be reached at mhughes3@stanford.edu.
Perspective
Back to Top
The initiation of real-time continuous glucose monitoring 18 years ago is one of the most important diabetes technology advances of the current era. Like most advances in diabetes technology that are developed for outpatient care, there are often challenges with bringing this into the inpatient environment. I recall in the 1980s as a medical student and an internal medicine resident the caution and concern of using fingerstick capillary blood glucose monitoring for patients in the hospital. The standard of care on the outpatient side was urine testing, which today most would agree is perhaps a tad north of worthless for deciding insulin doses. Venous blood glucose testing in the hospital was venipuncture plasma glucose testing, which would take 1 to 3 hours for the results to return. Taking mealtime insulin and making a decision at the time was actually south of worthless, which is why many hospitals in those early days relied on urine glycosuria to decide insulin dosing.
This historical perspective is important, as CGM has become standard of care in the U.S. for insulin-requiring patients in the outpatient setting. In the very near future, we will see more admitted patients with CGM as over-the-counter CGM is launched, and we can predict many more patients (including those without diabetes) using this technology.
Currently, CGM is not approved for inpatient use by the FDA, although during the COVID-19 pandemic, CGM was allowed to enable remote monitoring of hospitalized patients. Most hospitals individualize how they allow CGM technology to be used. For example, many hospitals allow the CGM to continue (for those already using it), yet insulin dosing must be performed by conventional finger stick testing. It gets more complicated with automated insulin delivery devices where basal dose modulation is all dependent on CGM readings.
The study by Lee and colleagues is an important early step in eventually allowing us to use CGM routinely in the hospital, including those not using CGM at home. In this study, admitted patients were allowed to use their personal CGM, including 24% with automated insulin delivery systems. Accuracy by several metrics was similar to these same devices in the outpatient setting and importantly, 80% of the nursing staff preferred CGM to blood glucose monitoring. Furthermore, 63% of patients liked the CGM protocol and felt no significant interruptions to their diabetes management.
The use of routine CGM in the hospital is a matter of when, not if. Patients who use CGM appreciate not having painful finger sticks. More importantly, in theory, CGM allows the prevention of significant hypoglycemia and hyperglycemia. A major benefit of having patients more engaged in their own diabetes self-management in the hospital includes communicating with nursing staff to hopefully prevent the extremes in glycemia. Most endocrinologists have stories from their patients about dangerous diabetes management during hospitalizations, and the hope is CGM in the hospital will limit these occurrences.
It must be noted that more studies are needed, in addition to the fact no one is suggesting CGMs be used in ICUs or for IV insulin infusion decisions. Furthermore, these devices are not appropriate to use in the operating rooms. Nevertheless, the routine use of CGM in the hospital is inevitable, and like in the 1980s with blood glucose monitoring, hospitals will need to adapt to this new technology.
Collapse
Read more about
Click Here to Manage Email Alerts