Randomized trial finds continuous glucose monitoring improves TIR

Continuous glucose monitoring (CGM)-guided insulin management increased the proportion of time hospitalized patients with type 2 diabetes spent within target glucose range compared with standard point-of-care testing in a randomized pilot trial conducted in medical and surgical wards, according to a study published in Endocrine Practice.

Researchers, led by Alejandra Rivas-Montenegro, MD, of Gregorio Maranon Hospital in Madrid, Spain, conducted the trial to evaluate whether using continuous glucose monitoring (CGM) (FreeStyle Libre 2 or 3) to guide inpatient insulin adjustments improves glycemic control compared with standard point‑of‑care glucose monitoring in hospitalized adults with type 2 diabetes.

In the study, patients whose insulin adjustments incorporated CGM data achieved a mean TIR of about 78% compared with about 67% among patients managed using standard point-of-care glucose (POCG) measurements.

Study Design and Population

Researchers conducted the pilot randomized controlled trial at a tertiary hospital in Madrid, Spain, enrolling adult patients hospitalized in medical or surgical wards with type 2 diabetes requiring basal-bolus insulin therapy and an expected stay longer than 72 hours.

A total of 42 patients were initially recruited; five were excluded after early discharge, leaving 37 patients in the final analysis. Nineteen patients were randomized to CGM-guided management and 18 to standard care based on POCG testing.

All patients wore a CGM sensor (FreeStyle Libre 2 or FreeStyle Libre 3), but treatment decisions differed by group. In the intervention group, insulin adjustments incorporated sensor glucose profiles, alarms, and trend arrows. In the control group, alarms were disabled and therapeutic decisions relied solely on POCG measurements taken six times daily.

The primary outcomes included the difference in time in range (TIR) 70-180 mg/dL, hospital glycemia risk index, time below range measured as the percentage below 70 mg/dL and below 54 mg/dL, and time above range measured as the percentage above 180 mg/dL and above 250 mg/dL.

Glycemic Outcomes

CGM-guided management improved several metrics of glycemic control during hospitalization.

• Higher TIR in the CGM‑guided group (78% vs 67%)

• Lower time above range 180–250 mg/dL in the CGM group (14% vs 23%)

• More asymptomatic hypoglycemia detected by CGM (1.65 vs 0.31 events)

• No increase in clinically significant hypoglycemia, nocturnal hypoglycemia, or insulin requirements, and

• CGM accuracy was acceptable for inpatient use (overall MARD 15%; 70% zone A on DTS error grid).

The improvement in TIR corresponded to an increase of roughly 10 percentage points with CGM-guided therapy. Time below range did not differ significantly between groups, with about 3% to 4% of readings falling below 70 mg/dL in both groups.

CGM also identified more asymptomatic hypoglycemic episodes, with an average of about 1.7 events per patient compared with about 0.3 events detected in the standard monitoring group.

Gurdeep Singh, MD, FACE, of Lake Erie College of Osteopathic Medicine in Pennsylvania, and Guthrie Lourdes Hospital in Binghamton, New York, provided commentary about the implications of this study and its potential for future diabetes treatment. “These findings suggest that CGM‑guided insulin adjustments may improve inpatient glycemic quality without increasing hypoglycemia risk,” Dr. Singh said. 

Glycemic Variability, Risk Metrics and Device Performance

The overall mean glycemia risk index was about 32, corresponding to zone B, and did not differ significantly between groups. In a subgroup analysis, patients with high glycemic variability showed greater time above 250 mg/dL and higher glycemia risk index values than those with lower variability.

Across 1,105 matched glucose pairs, the CGM systems showed a mean absolute relative difference of about 15% compared with point-of-care testing.

Using the Diabetes Technology Society error grid, about 70% of readings fell within zone A and about 29% in zone B, indicating clinically acceptable accuracy for therapeutic decision making.

In a device-specific analysis, the newer FreeStyle Libre 3 demonstrated higher accuracy than FreeStyle Libre 2, with a mean absolute relative difference of about 9% and about 17%, respectively.

Radiologic procedures, including computed tomography and radiography, did not significantly affect sensor accuracy.

Patient-reported outcomes using the Glucose Monitoring Satisfaction Survey indicated high acceptance of the technology, with high scores for openness and validity, with low emotional and behavioral burden.

Strengths of the Study

According to Dr. Singh, strengths of the study include its randomized controlled design (which he said is rare in inpatient CGM research); concurrent POCG monitoring in both groups (allowing direct comparison and safety confirmation); use of both FSL2 and FSL3 (enabling real‑world evaluation of device performance); and standardized insulin protocols for both groups (reducing treatment variability).

Other strengths Dr. Singh noted include the study’s demonstration of CGM feasibility in non‑ICU inpatient settings; improvement in higher time in range (TIR),a metric increasingly linked to outcomes; detection of asymptomatic hypoglycemia (which POCG often misses); device‑specific accuracy data (including differences between FSL2 and FSL3); and patient‑reported satisfaction, showing high acceptance (validity score 4.4 of 5).

“These strengths support the growing argument that CGM can enhance inpatient diabetes management,” said Dr. Singh.

Limitations of the Study

The researchers noted several limitations, including the small sample size and pilot design. The study was conducted at a single center and did not demonstrate reductions in hypoglycemia events or hospital length of stay.

Additional factors such as fasting periods, particularly in surgical patients, were considered for insulin adjustments, but meal records were not documented. Lastly, no differences were observed in hospital stay, as it was primarily influenced by the complexity of the condition leading to hospitalization.

Dr. Singh also noted that several of the study's design limitations temper the conclusions. Specifically, “the small sample size limits statistical power and generalizability; a single‑center study in a tertiary hospital may not reflect community hospital workflows; short duration of CGM use (≥72 hours) limits assessment of longer‑term inpatient outcomes; and nursing staff were trained specifically for the study, which may not reflect typical hospital staffing or alarm fatigue.”

In addition, Dr. Singh said, “Other limitations include the pilot nature is underpowered for clinical outcomes such as LOS, infections, or mortality; MARD of 15% is higher than outpatient performance standards; high variability in baseline HbA1c and admission glucose may influence glycemic patterns; trend‑arrow–based insulin adjustments require specialized training and may not be widely implementable; and there was no standardized protocol for alarm management, which is crucial for real‑world adoption.”

Clinical Implications

According to the researchers, the findings demonstrate that CGM significantly improves glycemic control in hospitalized patients with type 2 diabetes.

“The implementation of CGM in hospitals could represent a significant advancement in diabetes management,” the researchers wrote, noting that the technology may enable “more precise and timely glucose data, enabling better-informed therapeutic decisions.”

Despite the limitations he listed, Dr. Singh noted that the study adds to the growing evidence that CGM can improve inpatient glycemic control by increasing TIR, which correlates with improved outcomes in ambulatory and emerging inpatient data; reducing hyperglycemia, particularly post‑prandial excursions; and identifying silent hypoglycemia, which is common in hospitalized patients and often missed by POCG.

He also echoed researchers' sentiments about future advancements in care while acknowledging the need for more in-depth research. “This pilot RCT provides encouraging evidence that CGM‑guided insulin therapy improves inpatient glycemic control in adults with type 2 diabetes without increasing hypoglycemia. The improvement in TIR and reduction in lower time above range, combined with acceptable accuracy and high patient satisfaction, suggest that CGM may play a meaningful role in modernizing inpatient diabetes management." he said. However, the small sample size and single‑center design necessitate larger trials before CGM can be broadly recommended as standard inpatient practice.”

Disclosures: The researchers reported no conflicts of interest.

AACE Endocrine AI is published by Conexiant under a license arrangement with the American Association of Clinical Endocrinology, Inc. (AACE^®). The ideas and opinions expressed in AACE Endocrine AI do not necessarily reflect those of Conexiant or AACE. For more information, see Policies.