Abstract
Background
The accuracy of the FreeStyle Libre 2 system in critically ill patients needing insulin therapy remains inadequately evaluated.
Objective
To evaluate the clinical and numerical accuracy of continuous glucose monitoring (FreeStyle Libre 2) and capillary glucometry (StatStrip) compared with central laboratory glucose in critically ill patients requiring insulin therapy.
Methods
We conducted a diagnostic accuracy study, evaluating simultaneously two index tests, the FreeStyle Libre 2 and capillary glucose, compared to the central laboratory glucose (venous or arterial) as reference standard. The study included critically ill adult patients admitted to the intensive care unit (ICU) with diabetes or stress hyperglycemia, requiring insulin therapy, and ventilatory or vasopressor support. Numerical accuracy was assessed using ISO 15197:2013 criteria and Mean Absolute Relative Difference (MARD). Clinical accuracy was evaluated using Clarke and Parkes error grid analysis.
Results
A total of 157 paired measurements were collected. FreeStyle Libre 2 showed a MARD comparable to capillary glucose (10.43% vs 7.58%; p=0.14). Neither method met ISO 15197:2013 numerical accuracy criteria. In the clinical accuracy analysis, FreeStyle Libre 2 classified 100% of measurements within zones A+B on both the Clarke and Parkes grids, whereas capillary glucose achieved 97.6% and 100%, respectively.
Conclusions
FreeStyle Libre 2 showed numerical accuracy comparable to capillary blood glucose and consistently reliable clinical performance in critically ill patients needing insulin therapy. These results support its potential as an alternative to capillary glucose monitoring in the ICU. More studies involving a greater number of hypoglycaemic events are required to confirm its effectiveness in this critical range.
Keywords: diabetes mellitus, continuous glucose monitoring, critical care, glycemic control, ICU
Plain Language Summary
People admitted to intensive care units often require regular blood sugar checks, especially when they are receiving insulin. These measurements usually require finger-prick tests or blood samples. New devices, such as the FreeStyle Libre 2 sensor, can measure glucose continuously through a small sensor placed on the skin. This provides a more complete evaluation of glucose control, and reduces the need for repeated needle sticks. In this study, we evaluated the performance of FreeStyle Libre 2 sensor in critically ill adults by comparing its readings with blood sugar measurements obtained through laboratory testing and routine bedside testing (finger-prick tests). We analyzed 157 measurements from seven patients receiving intensive care. Overall, the sensor provided results that were close enough to laboratory measurements to support most clinical decisions. These results suggest that the FreeStyle Libre 2 sensor could be a useful tool for monitoring glucose levels in critically ill patients, potentially reducing the number of finger-prick tests required. However, larger studies are needed to confirm these findings.
Introduction
In critically ill patients, poor glycemic control is associated with increased mortality rates estimated to be between 6% and 89%. 1 This risk is particularly high when hyperglycaemia is present, with an increased risk of between 20% and 300% in individuals with or without diabetes. 2 Even mild hypoglycaemia has been shown to correlate with higher mortality rates, 3 which are common among patients receiving insulin 4 or undergoing strict glycaemic control strategies that can increase the risk of severe hypoglycaemia by up to 15-fold. 5
Glycaemic control in the intensive care unit (ICU) is typically managed based on measurements taken using capillary glucometry; however, this method requires frequent testing and carries a higher risk of undetected hypoglycemia. 1 Continuous glucose monitoring (CGM) has become an alternative, providing real-time glucose trends, reducing the need for repeated capillary tests, and enabling early detection of dysglycemia, benefits well documented in outpatient and general inpatient populations.2-4 In ICU settings, CGM use has been linked to lower mortality (RR 0.54, 95% CI 0.34–0.85) and reduced hypoglycemia (RR 0.35, 95% CI 0.25–0.49). 5
Despite these potential benefits, CGM use in the ICU remains controversial, and no formal guidelines currently recommend its routine use.6,7 Although multiple studies have examined CGM accuracy in the ICU, most focused on older devices, many of which are no longer available on the market.8-10 A recent study evaluating the FreeStyle Libre 2 during the first 96 hours of ICU stay reported adequate accuracy; however, participants had low clinical severity. Additionally, the study did not specifically analyze how hypo- or hypertension affected device performance. 11 As a result, it remains unclear whether the CGM systems currently on the market can provide both numerical and clinical accuracy comparable to standard glucose measurement methods in critically ill patients.
This study aimed to evaluate the clinical and numerical accuracy of CGM (FreeStyle Libre 2) and capillary blood glucose in severely ill intensive care patients with diabetes mellitus or stress-induced hyperglycaemia. Central laboratory glucose was used as the reference standard. The study was based on experience in an intensive care unit at a referral hospital in Colombia.
Methods
This diagnostic accuracy study was conducted on a consecutive series of samples collected from patients treated in the ICU of Hospital Universitario San Ignacio (Bogotá, Colombia) between September and November 2025. Participants included adults aged 18 years or older with a diagnosis of diabetes mellitus and/or stress-induced hyperglycemia (>180 mg/dL) who required initiation of insulin therapy along with invasive mechanical ventilation or vasopressor support. Exclusion criteria comprised patients with an expected ICU stay or survival of less than 48 hours and pregnant women. All eligible participants were screened, and written informed consent was obtained from patients and/or their surrogates prior to enrollment. The study protocol received approval from the Ethics Committee of Pontificia Universidad Javeriana and Hospital Universitario San Ignacio (approval number FM-CIE-1324-23, approved on January 10, 2024).
Sociodemographic data and diabetes history were gathered at enrollment. Clinical information, admission laboratory tests (hemoglobin, blood urea nitrogen, creatinine), and severity scores (APACHE II and SOFA) were obtained from medical records. Daily clinical variables, including vital signs, ventilatory support, vasopressor dose, insulin dose, and renal replacement therapy, were recorded prospectively. All data were entered into a standardized case report form created for the study.
Two index tests were evaluated: CGM using the FreeStyle Libre 2 system (Abbott Diabetes Care) and capillary glucometry with the StatStrip device (Nova Biomedical). Both were compared to the reference standard, central laboratory glucose (ARCHITECT PLUS ci4100, Abbott), obtained from either a central venous catheter or an arterial line. The FreeStyle Libre 2 sensor was placed on the posterior region of the arm following the manufacturer’s instructions; after a one-hour calibration period, the device was activated. The three tests were performed simultaneously three times daily for up to fifteen days; the interval between index tests and reference test sampling was less than one minute, and central laboratory samples were processed within five minutes. Because FreeStyle Libre 2 sensor measures interstitial rather than blood glucose, a physiological lag may occur. No lag-time correction was applied, as the study evaluated device performance under routine clinical conditions. Results from the FreeStyle Libre 2 and the laboratory were masked from the clinical team, who managed patients solely with capillary glucometry according to institutional protocols. Similarly, laboratory personnel were blinded to the CGM test results.
The accuracy of FreeStyle Libre 2 and capillary glucometry measurements was evaluated using central laboratory glucose as the reference standard, focusing on both clinical and numerical accuracy. Clinical accuracy was assessed with the Clarke and Parkes error grids, which categorize paired measurements into five zones (A–E) based on their clinical impact: Zone A, clinically accurate; Zone B, benign discrepancies; Zone C, potential for unnecessary treatment; Zone D, risk of missing hypo- or hyperglycemia; and Zone E, risk of incorrect treatment. Adequate clinical accuracy was defined as ≥95% of points in Zones A+B for Clarke and ≥99% for Parkes.12,13 Numerical accuracy was determined according to ISO 15197:2013 criteria (±15 mg/dL for glucose below 100 mg/dL, or ±15% for glucose at or above 100 mg/dL), with ≥95% compliance. 14 The Mean Absolute Relative Difference (MARD)—calculated as the absolute difference between the test and reference divided by the reference value, multiplied by 100—was also reported, with values below 15% considered acceptable. 15 Furthermore, episodes of hypoglycemia (<70 mg/dL) and hyperglycemia (>180 mg/dL) detected by each method were documented.
Sample size estimation was based on the expected differences in the proportion of measurements in Parkes Zones A+B: 90% for FreeStyle Libre 2, 16 and 98% for the StatStrip glucometer, 17 assuming a two-sided alpha of 0.05, yielding a required sample size of 145 paired measurements.
Baseline characteristics were summarized using descriptive statistics: categorical variables as absolute and relative frequencies, and continuous variables as either mean and standard deviation or median and interquartile range, depending on distribution. The proportion of ISO 15197:2013 compliance for FreeStyle Libre 2 and capillary glucometry was compared using Z-tests, both for the full sample and stratified by glucose levels ≥100 and <100 mg/dL. Similarly, the proportions of measurements within Clarke and Parkes Zones A+B were compared between methods. Differences in median MARD between FreeStyle Libre 2 and capillary glucometry were evaluated using the Mann–Whitney U test. Exploratory graphical analyses were conducted to evaluate how factors that could affect CGM accuracy impact the magnitude of differences between diagnostic test results. These factors included mean arterial pressure, haemoglobin concentration, creatinine and blood urea nitrogen levels.
A p-value <0.05 was regarded as statistically significant. Statistical analyses were conducted using Stata SE 19 (StataCorp, College Station. Texas, USA).
Results
A total of 157 simultaneous measurement sets from the three different glucose assessment methods were collected from seven patients meeting the inclusion criteria. The sociodemographic and clinical characteristics of these patients are shown in Table 1. Three patients had a pre-existing diagnosis of diabetes prior to ICU admission and were receiving outpatient treatment with SGLT2 inhibitors, DPP-4 inhibitors, and a basal-bolus insulin regimen. Of all samples, 82 (52.2%) measurements were taken while patients were on vasopressor support, 141 (89.8%) during invasive mechanical ventilation, and 20 (12.7%) during renal replacement therapy. Central laboratory glucose was measured either through a central venous catheter in 108 (68.8%) samples or an arterial line in 49 (31.2%) samples.
Table 1.
Sociodemographic and Clinical Characteristics of Critically Ill Patients Undergoing Glucose Monitoring
| Variable | N = 7 |
|---|---|
| Demographic Characteristics | |
| Sex male, n (%) | 5 (71.43) |
| Age (years), mean (SD) | 60 (14.05) |
| BMI (Kg/m2), mean (SD) | 24.48 (4.46) |
| Diabetes-Related Characteristics | |
| Diabetes mellitus, n (%) | 3 (42.86) |
| Time since diagnosis (years), mean (SD) | 20 (17.32) |
| Stress-induced hyperglycemia, n (%) | 4 (57.14) |
| Baseline Comorbid Conditions | |
| Ischemic cardiopathy, n (%) | 1 (14.29) |
| Cerebrovascular disease, n (%) | 0 |
| Diabetic nephropathy, n (%) | 2 (28.57) |
| Admission-Related Variables | |
| Admission diagnosis, n (%) | |
| Neurological diagnosis | 3 (42.86) |
| Sepsis | 2 (28.57) |
| Gastrointestinal diagnosis | 2 (28.57) |
| Type of admission, n (%) | |
| Medical | 3 (42.86) |
| Surgical | 4 (57.14) |
| Clinical Severity and Biochemical Profile | |
| SOFA Score, mean (SD) | 7.71 (2.5) |
| APACHE II Score, median (IQR) | 16 (13 – 31) |
| Hb (g/dl), mean (SD) | 11.61 (4.52) |
| Cr (mg/dl), mean (SD) | 1.71 (1.24) |
| BUN (mg/dl), mean (SD) | 39.7 (27.11) |
| Hospitalization Outcomes | |
| Hospital length of stay (days), mean (SD) | 16.29 (7.02) |
| ICU length of stay (days), median (IQR) | 8 (7 – 12) |
| Mechanical ventilation, n (%) | 7 (100) |
| Duration of mechanical ventilation (days), mean (SD) | 9.14 (6.47) |
| Vasopressor support, n (%) | 6 (85.71) |
| Duration of vasopressor therapy (days), mean (SD) | 6.5 (4.49) |
| Renal replacement therapy, n (%) | 4 (57.14) |
| Duration of renal replacement therapy (days), mean (SD) | 4.25 (0.96) |
| Death, n (%) | 5 (71.43) |
APACHE II, Acute Physiology and Chronic Health Evaluation II; BMI, body mass index; BUN, blood urea nitrogen; Cr, serum creatinine; Hb, hemoglobin; IQR, interquartile range; SD, standard deviation; SOFA, Sequential Organ Failure Assessment.
The median glucose values for the three methods were as follows: central laboratory glucose: 148 mg/dL (IQR 117–189), FreeStyle Libre 2: 130 mg/dL (IQR 105–172), and capillary glucose: 144 mg/dL (IQR 120–178). Based on the reference standard, 2 measurements (1.3%) were in the hypoglycemia range, 108 (68.8%) in the normoglycemia range, and 47 (29.9%) in the hyperglycemia range. The agreement between these categories in the index tests and those obtained with the reference test is shown in Table 2. Of the two measurements in the hypoglycaemia range according to the reference standard, the FreeStyle Libre 2 identified one and the capillary meter did not detect any. The first measurement was 62 mg/dL according to the reference standard, compared with 91 mg/dL by capillary glucose measurement and 70 mg/dL by FreeStyle Libre 2. The second measurement was 68 mg/dL according to the reference standard, compared with 81 mg/dL by capillary glucose measurement and 62 mg/dL by FreeStyle Libre 2.
Table 2.
Cross-Classification of Glucose Categories Using Serum Glucose as the Reference Method Compared With Capillary Glucometry and FreeStyle Libre 2 in Critically Ill Patients
| | Capillary glucometry | FreeStyle libre 2 | |||||
|---|---|---|---|---|---|---|---|
| < 70 (n = 0) | 70 – 180 (n = 110) | > 180 (n = 47) | < 70 (n = 2) | 70 – 180 (n = 108) | > 180 (n = 47) | ||
| Central Laboratory Glucose | < 70 (n=2) | 0 | 2 (1.8) | 0 | 1 (50) | 1 (0.9) | 0 |
| 70 – 180 (n = 108) | 0 | 108 (98.2) | 0 | 1 (50) | 107 (99.1) | 0 | |
| > 180 (n = 47) | 0 | 0 | 47 (100) | 0 | 0 | 47 (100) | |
Regarding numerical accuracy, the FreeStyle Libre 2 showed a MARD of 10.43% (IQR 5.69–16.8), while capillary glucose had a MARD of 7.58% (IQR 4.07–14.29, p = 0.14). According to ISO 15197:2013 criteria, the FreeStyle Libre 2 correctly classified 102 (64.9%) measurements, compared to 122 (77.7%) for capillary glucose (p = 0.0125). Among measurements with glucose ≥100 mg/dL, numerical accuracy was lower for FreeStyle Libre 2 (65.1%) compared to capillary glucose (80.1%) (p = 0.0056). For glucose measurements <100 mg/dL, FreeStyle Libre 2 correctly classified 31 (64.5%) and capillary glucose 16 (56.3%) (p = 0.5805).
Clinical accuracy results assessed using Clarke and Parkes error grids are shown in Figure 1 and Figure 2. In the Clarke grid, 131 measurements (83.4%) from the FreeStyle Libre 2 fell in Zone A, and 26 (16.6%) in Zone B, with 100% in Zones A+B. Capillary glucose measurements yielded 131 (83.4%) in Zone A and 23 (14.7%) in Zone B, for a total of 98.1% in Zones A+B, with no statistically significant difference (p = 0.1573). In the Parkes grid, FreeStyle Libre 2 had 135 measurements (85.9%) in Zone A and 22 (14%) in Zone B, again with 100% in Zones A+B. Capillary glucose measurements included 146 (92.9%) in Zone A and 11 (7%) in Zone B, also reaching 100% in Zones A+B. Notably, no device recorded any measurement in Zone E in either grid. No severe hypoglycemia events were observed during the study. Of the two measurements <70 mg/dL according with the reference standard, the FreeStyle Libre 2 correctly identified one event, whereas the capillary glucose detected none.
Figure 1.
Clarke error grid assessment of capillary glucose (Red) and FreeStyle libre 2 (Blue) Measurements against central laboratory glucose in critically Ill patients
Figure 2.
Parkes error grid assessment of capillary glucose (Red) and FreeStyle libre 2 (Blue) Measurements against central laboratory glucose in critically Ill patients
Figure 3 illustrates the relationship between mean arterial pressure and the glucose difference relative to the reference standard for both index methods. No systematic pattern of variation in glucose difference with blood pressure was observed, and the dispersion of measurements was similar across the full range of mean arterial pressures for both devices.
Figure 3.
Glucose differences [Reference – Index] across levels of mean arterial pressure in critically Ill patients. Red: [Reference glucose – Capillary glucose]. Blue: [Reference glucose – FreeStyle libre 2 glucose]. Reference test: Central laboratory glucose
Figure 4 shows the association between hemoglobin, creatinine, and blood urea nitrogen (BUN) levels and the glucose difference relative to the reference standard for both index methods. FreeStyle Libre 2 didn’t show a clear systematic bias across these variables.
Figure 4.
Glucose differences [Reference – Index] across levels of hemoglobin, creatinine and blood urea nitrogen (BUN) in critically Ill patients. Red: [Reference glucose – Capillary glucose]. Blue: [Reference glucose – FreeStyle libre 2 glucose]. Reference test: Central laboratory glucose
Discussion
This study compared the numerical and clinical accuracy of the FreeStyle Libre 2 with capillary and central laboratory glucose measurements in critically ill patients with diabetes mellitus or stress-induced hyperglycaemia requiring insulin therapy. Our results suggest that the FreeStyle Libre 2 is numerically accurate to a similar degree as capillary glucose. Both devices produced a proportion of measurements under the recommended values. In contrast, FreeStyle Libre 2 clinical accuracy remained consistently high, with all measurements within the safe zones of the Clarke and Parkes error grids.
Regarding numerical accuracy, the FreeStyle Libre 2 shows slightly lower performance than capillary glucose; however, neither method met the ISO 15197:2013 criteria. This aligns with the systematic review by Dávila-Ruales et al., 18 which reported that none of the ambulatory CGM systems fully adhered to the requirements of ISO 15197:2013, with compliance rates ranging from 64.7% to 91.1%. Similarly, studies assessing subcutaneous CGM devices in intensive care settings have shown wide variability in ISO 15197:2013 compliance, ranging from 7% to 89.3%, without meeting the required threshold. 8 In terms of MARD, the FreeStyle Libre 2 shows acceptable values, with no statistically significant differences compared to capillary glucose. This performance is consistent with a recent meta-analysis of subcutaneous CGM in intensive care units, which reported a pooled MARD of 13% (IQR 11–14.7), 8 and a recent diagnostic accuracy study of FreeStyle Libre 2 in critically ill patients that found a MARD of 12 ± 2%. 11
For clinical accuracy, the FreeStyle Libre 2 demonstrated excellent performance, with all measurements falling within Clarke and Parkes zones A or B. In contrast, capillary glucose produced some values in zones C and D. This degree of clinical concordance mirrors findings from a recent meta-analysis reporting 98.7% (IQR 98–99.7%) of CGM measurements within zones A+B in critically ill patients, 8 and is consistent with the results of Chiumello et al, who evaluated FreeStyle Libre 2 in this population. 11 Taken together, these results suggest that despite limitations in numerical accuracy, the clinical performance of the FreeStyle Libre 2 remains consistent and sufficiently safe to support therapeutic decision-making in critically ill patients.
Regarding hypoglycemia detection, only two events occurred in our study. The FreeStyle Libre 2 identified one event, while capillary glucose detected none. Although these numbers are too small to draw definitive conclusions, prior evidence in ambulatory populations shows good diagnostic performance for hypoglycemia detection (sensitivity 85.7%, specificity 95.3%). 18 In contrast, data from intensive care units are more varied, with reported sensitivities for subcutaneous CGM ranging from 17% to 100%, depending on device type. 8 Overall, these findings suggest that while CGM may provide valuable information, its accuracy in detecting hypoglycemia in critically ill patients is inconsistent and should be interpreted with caution. Larger studies with more hypoglycemic events are needed to better understand device performance in this critical range.
Conditions such as anemia and renal disease have been reported to affect the accuracy of glucose measurements. 19 However, in our study, no increased variability was observed in FreeStyle Libre 2 measurements under these conditions.
This study has several strengths. These include a consecutive sampling protocol that reduced methodological bias; a direct comparison of FreeStyle Libre 2 with capillary glucose, the method routinely used in intensive care units, against a solid reference standard (central laboratory glucose); and the inclusion of patients with high illness severity, such as individuals needing mechanical ventilation and vasopressor support, a group often underrepresented in previous CGM research. Using a contemporary CGM device also increases the clinical relevance of our findings. The main limitation of this study is the low incidence of hypoglycemia, which restricts the ability to draw conclusions in this glycemic range. Importantly, this limitation does not affect the main conclusion: neither method met ISO 15197:2013 standards, but both showed acceptable MARD values and satisfactory clinical performance, with more consistent results for the FreeStyle Libre 2.
Conclusions
Overall, our findings suggest that the FreeStyle Libre 2 provides adequate clinical accuracy and numerical performance comparable to capillary glucose in critically ill patients receiving insulin therapy. In this context, continuous glucose monitoring may be considered a viable alternative to capillary glucose for the management of normoglycemia and hyperglycemia in the intensive care unit. However, additional studies with a larger number of hypoglycemic events are required to confirm device performance within this critical range.
Acknowledgements
The authors thank the nursing staff of Hospital Universitario San Ignacio, Bogotá, Colombia, and the members of the Department of Internal Medicine at Pontificia Universidad Javeriana, Bogotá, Colombia, for their support and collaboration during the development of the present research study.
Footnotes
Author Contributions: Miguel Aguilar-Schotborgh: Methodology, Investigation, Formal analysis, Writing - Original Draft. Oscar Muñoz-Velandia: Methodology, Formal analysis, Writing - Review & Editing. Diana Cristina Henao-Carrillo: Methodology, Writing - Review & Editing. Luis Triana Moreno: Methodology, Writing - Review & Editing. Jhon Moreno Mogollon: Resources, Data Curation. Ana Maria Gómez: Conceptualization, Methodology, Writing - Review & Editing, Supervision.
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
ORCID iDs
Miguel Aguilar-Schotborgh https://orcid.org/0000-0002-0610-4877
Oscar Muñoz-Velandia https://orcid.org/0000-0001-5401-0018
Diana Cristina Henao-Carrillo https://orcid.org/0000-0002-1353-148X
Luis Triana Moreno https://orcid.org/0000-0002-2609-3132
Jhon Moreno Mogollon https://orcid.org/0009-0004-2783-9719
Ana Maria Gómez https://orcid.org/0000-0002-8907-3470
Ethical Considerations
Approved by the Ethics Committee of Pontificia Universidad Javeriana/Hospital Universitario San Ignacio (FM-CIE-1324-23, approved on January 10, 2024).
Consent to Participate
Written informed consent was obtained from all participants or their legal surrogates prior to enrollment.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.*
References
- 1.Egi M, Bellomo R, Stachowski E, et al. Hypoglycemia and Outcome in Critically Ill Patients. Mayo Clin Proc. 2010;85:217-224. doi: 10.4065/MCP.2009.0394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gomez AM, Umpierrez GE. Continuous glucose monitoring in insulin-Treated patients in non-ICU settings. J Diabetes Sci Technol. 2014;8:930-936. doi: 10.1177/1932296814546025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gómez AM, Umpierrez GE, Muñoz OM, et al. Continuous Glucose Monitoring Versus Capillary Point-of-Care Testing for Inpatient Glycemic Control in Type 2 Diabetes Patients Hospitalized in the General Ward and Treated With a Basal Bolus Insulin Regimen. J Diabetes Sci Technol. 2015;10:325-329. doi: 10.1177/1932296815602905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Price C, Ditton G, Russell GB, Aloi J. Reliability of Inpatient CGM: Comparison to Standard of Care. J Diabetes Sci Technol. 2023;17:329-335. doi: 10.1177/19322968211062168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Yao Y, Zhao Y-H, Zheng W-H, Huang H-B. Subcutaneous continuous glucose monitoring in critically ill patients during insulin therapy: a meta-analysis. Am J Transl Res. 2022;14:4757-4767. [PMC free article] [PubMed] [Google Scholar]
- 6.Honarmand K, Sirimaturos M, Hirshberg EL, et al. Society of Critical Care Medicine Guidelines on Glycemic Control for Critically Ill Children and Adults 2024. Crit Care Med. 2024;52:E161-E181. doi: 10.1097/CCM.0000000000006174. [DOI] [PubMed] [Google Scholar]
- 7.Adapp C, ElSayed NA, McCoy RG, et al. 16. Diabetes Care in the Hospital: Standards of Care in Diabetes—2025. Diabetes Care. 2025;48:S321-S334. doi: 10.2337/DC25-S016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nielsen CG, Grigonyte-Daraskeviciene M, Olsen MT, et al. Accuracy of continuous glucose monitoring systems in intensive care unit patients: a scoping review. Intensive Care Medicine. 2024;50(12 202450):2005–2018. doi: 10.1007/S00134-024-07663-6. [DOI] [PubMed] [Google Scholar]
- 9.Sadhu AR, Serrano IA, Xu J, et al. Continuous Glucose Monitoring in Critically Ill Patients With COVID-19: Results of an Emergent Pilot Study. J Diabetes Sci Technol. 2020;14:1065-1073. doi: 10.1177/1932296820964264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Finn E, Schlichting L, Grau L, Douglas IS, Pereira RI. Real-World Accuracy of CGM in Inpatient Critical and Noncritical Care Settings at a Safety-Net Hospital. Diabetes Care. 2023;46:1825-1830. doi: 10.2337/DC23-0089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chiumello D, Passeri M, Coppola S, et al. Continuous interstitial glucose monitoring in diabetic and non-diabetic critically ill patients is simple and accurate: comparison with venous, arterial and capillary glucose measurements. Acta Diabetol. 2025;62:1173-1181. doi: 10.1007/S00592-025-02531-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Clarke WL, Cox D, Gonder-Frederick LA, Carter W, Pohl SL. Evaluating clinical accuracy of systems for self-monitoring of blood glucose. Diabetes Care. 1987;10:622-628. doi: 10.2337/DIACARE.10.5.622. [DOI] [PubMed] [Google Scholar]
- 13.Parkes JL, Slatin SL, Pardo S, Ginsberg BH. A new consensus error grid to evaluate the clinical significance of inaccuracies in the measurement of blood glucose. Diabetes Care. 2000;23:1143-1148. doi: 10.2337/DIACARE.23.8.1143. [DOI] [PubMed] [Google Scholar]
- 14.Jendrike N, Baumstark A, Kamecke U, Haug C, Freckmann G. Evaluation of a Blood Glucose Monitoring System’s Measurement Accuracy. J Diabetes Sci Technol. 2017;11:1275-1276. doi: 10.1177/1932296817727550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Freckmann G, Mende J, Pleus S, et al. Mean Absolute Relative Difference of Blood Glucose Monitoring Systems and Relationship to ISO 15197. J Diabetes Sci Technol. 2022;16:1089-1095. doi: 10.1177/19322968211001402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Alva S, Bailey T, Brazg R, et al. Accuracy of a 14-Day Factory-Calibrated Continuous Glucose Monitoring System With Advanced Algorithm in Pediatric and Adult Population With Diabetes. J Diabetes Sci Technol. 2022;16:70-77. doi: 10.1177/1932296820958754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chan PCR, Rozmanc M, Seiden-Long I, Kwan J. Evaluation of a point-of-care glucose meter for general use in complex tertiary care facilities. Clin Biochem. 2009;42:1104-1112. doi: 10.1016/J.CLINBIOCHEM.2009.03.023. [DOI] [PubMed] [Google Scholar]
- 18.Dávila-Ruales V, Gilón LF, Gómez AM, Muñoz OM, Serrano MN, Henao DC. Evaluating the precision and reliability of real-time continuous glucose monitoring systems in ambulatory settings: a systematic review. Ther Adv Endocrinol Metab. 2024;15:20420188241304459. doi: 10.1177/20420188241304459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gómez Medina A, González CA, Muñoz OM, et al. HbA1c overestimates the glucose management indicator: a pilot study in patients with diabetes, chronic kidney disease not on dialysis, and anemia using isCGM. Ther Adv Endocrinol Metab. 2024;15:20420188241252546. doi: 10.1177/20420188241252546. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.*