Diabetes Mellitus
Conditions
Keywords
CGM, Continuous Glucose Monitor
Brief summary
The purpose of this study is to determine if patient's own Continuous Glucose Monitor (CGMs) worn in the non-ICU hospital setting have adequate accuracy for blood glucose monitoring when compared to point-of-care (POC) capillary glucose measurement, and to determine if alerts given by CGMs worn in the non-ICU hospital would prevent episodes of hyperglycemia and hypoglycemia.
Detailed description
All patients admitted to the hospital in inpatient or observation status and having a Continuous Glucose Monitor (CGMs) attached will be screened and identified by providers and nurses. An order will be placed by providers to continue patient's home CGM use in the hospital. CGM patient agreement describing the hospital policy for use of CGMs will be provided to the patient for signature. Patients will be encouraged to continue to wear their CGMs in the hospital unless their presence interferes with patients' medical care. CGM presence will be recorded by nursing in patient chart under LDA (lines, drains, airway) section. Patients will be approached by the research staff and will be offered enrollment in the study. A research consent will be reviewed with patients and will be provided for signature. Patients will be asked to continue to use their home CGMs in the hospital. If CGM sensor will reach the end of life, patients have the option to place a new sensor and to follow manufacturer recommendations for calibration and setting. Patients are responsible to provide their own sensors, transmitters and readers during the hospitalization period. If it is determined that the current sensor will interfere with hospital medical and surgical care (MRI testing, procedure at the site), patients will be asked to remove the CGM sensor and transmitter. Patients will have the option to replace the sensor after the procedure if they so desire. Patients will continue to monitor their glucose level via CGM. For flash sensors, patients will be asked to scan their sensor at least every 8 and as needed. CGMs that require calibration will be calibrated by the patients using hospital glucometer readings. . Patients alerted by their CGM about low glucose levels or fast downward or upward trending will have to notify their nursing staff. A POC blood glucose check will be done by nursing staff to confirm the alert or the trend. Treatment decisions will be made based on POC readings. During hospitalization patients will continue to have their blood glucose checked with POC glucometers per hospital policy. Treatment of diabetes and the associated complications, dosing of insulin, change from PO to insulin treatment and back to PO medications during hospitalization or at discharge will be done by the admitting team with help from the inpatient Endocrinology Consultation Service. Decisions about insulin dosing or other diabetes treatments will be made based on hospital glucometer readings, and not based on CGM readings alone. Patients will be asked to give the research team provider access to their cloud CGM software, or to allow the team to download their CGM data prior to discharge if their reading devices do not automatically synchronize with the cloud software. CGM data will be compared with POC glucometer readings obtained in the hospital at approximately the same time. Different patient variables (demographics, comorbidities, labs and vitals, administered medications) will be collected from electronic health record and will be evaluated to determine if they interfere with CGM readings. The accuracy of CGM for glucose measurement in the hospital will be calculated.
Interventions
DEVICEContinuous Glucose Monitor (CGM)
CGM data will be downloaded prior to discharge and it will be compared with hospital glucometer readings taken at the same time.
Sponsors
Mayo Clinic
Study design
Allocation
NA
Intervention model
SINGLE_GROUP
Primary purpose
DIAGNOSTIC
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Patients with Type 1 or 2 Diabetes Mellitus, * Patients 18 years of age or older, * Patients admitted under medical or surgical services. * Patients treated with diet alone, insulin (SQ, insulin pump) or oral hypoglycemic medications. * Have their own Continuous Glucose Monitoring (CGM) monitor present on admission or able to be placed after admission.
Exclusion criteria
* Individuals under 18 years of age. * COVID-19 infection. * Infection of the skin at the CGM site requiring removal of the sensor. * Patients with altered Mental Status. * Patients unable to scan their flash CGMs at least every 8 hours. * Inability to provide written consent. * Hospitalized for less than 24 hours.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD) | From date of enrollment and until discharge from the hospital up to 30 days. | The absolute relative difference (ARD) was defined as the absolute difference between control - point of care (POC) glucose reading and the corresponding CGM reading divided by POC glucose reading and expressed as percentage. ARD was calculated as mean absolute relative difference (MARD) with standard deviation. MARD was calculated between matched pairs of POC (point of care) glucose readings and the closest CGM reading. MARD was summarized as mean (standard deviation) or median (range) for glucose readings \<70 mg/dl, 70-180 mg/dL, \>180 mg/dl, 181-250mg/dl, \>250 mg/dl respectively. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| CGM Recorded Duration of Hypoglycemic Episodes | From date of enrollment and until discharge from the hospital up to 30 days. | CGM recorded duration of hypoglycemic episodes |
| CGM Recorded Hyperglycemic Episodes | From date of enrollment and until discharge from the hospital up to 30 days. | The incidence of hyperglycemic episodes |
| CGM Recorded Duration of Hyperglycemic Episodes | From date of enrollment and until discharge from the hospital up to 30 days. | The duration of hyperglycemic episodes |
| Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | From date of enrollment and until discharge from the hospital up to 30 days. | The correlation between MARD point of care (POC) and Creatinine, Glomerular Filtration Rate (GFR), Bicarbonate, Hemoglobin (Hb), Mean Arterial Pressure (MAP), and Peripheral Oxygen Saturation (SpO2). The Spearman's rank correlation method was used to calculate the correlation coefficient. |
| CGM Recorded Hypoglycemia Episodes | From date of enrollment and until discharge from the hospital up to 30 days. | The incidence of hypoglycemic episodes |
| Hospitalization Length of Stay | From date of enrollment and until discharge from the hospital up to 30 days. | Duration of hospital stay in hours |
| Mortality Rate at 30 Days | From date of enrollment up to 60 days. | Mortality rate at 30 days post discharge from the hospital |
| Number of Re-hospitalization Events at the Same Institution Within 30 Days From the Discharge Date of the Preceding Admission. | From date of enrollment up to 60 days. | Number of re-hospitalization events that occurred at the same institution within 30 days from the discharge date of the preceding admission. |
| In Hospital Mortality | From date of enrollment and until discharge from the hospital up to 30 days. | Mortality during index hospitalization |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| CGM Patients Patients with diabetes mellitus admitted to the hospital and using a CGM were encouraged to continue to use these devices in inpatient setting. The device alarms of high or low glucose levels were communicated to the nursing staff.
Continuous Glucose Monitor (CGM): CGM data was downloaded prior to discharge and it was compared with hospital glucometer readings taken at the same time. | 142 |
| Total | 142 |
Withdrawals & dropouts
| Period | Reason | FG000 |
|---|---|---|
| Overall Study | Physician Decision | 40 |
| Overall Study | Withdrawal by Subject | 1 |
Baseline characteristics
| Characteristic | CGM Patients |
|---|---|
| Age, Continuous | 59.97 years STANDARD_DEVIATION 12.68 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 4 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 135 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 3 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 2 Participants |
| Race (NIH/OMB) Asian | 6 Participants |
| Race (NIH/OMB) Black or African American | 19 Participants |
| Race (NIH/OMB) More than one race | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 3 Participants |
| Race (NIH/OMB) White | 112 Participants |
| Region of Enrollment United States | 142 participants |
| Sex: Female, Male Female | 48 Participants |
| Sex: Female, Male Male | 94 Participants |
Adverse events
| Event type | EG000 affected / at risk |
|---|---|
| deaths Total, all-cause mortality | 2 / 142 |
| other Total, other adverse events | 0 / 142 |
| serious Total, serious adverse events | 0 / 142 |
Outcome results
Primary
Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD)
The absolute relative difference (ARD) was defined as the absolute difference between control - point of care (POC) glucose reading and the corresponding CGM reading divided by POC glucose reading and expressed as percentage. ARD was calculated as mean absolute relative difference (MARD) with standard deviation. MARD was calculated between matched pairs of POC (point of care) glucose readings and the closest CGM reading. MARD was summarized as mean (standard deviation) or median (range) for glucose readings \<70 mg/dl, 70-180 mg/dL, \>180 mg/dl, 181-250mg/dl, \>250 mg/dl respectively.
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
Population: 3316 total CGM-POC readings pairs were collected. Those 3316 POC-CGM reading pairs were further broken down into the following categories based on the POC value: CGM-POC Pairs with POC\<70 mg/dL, CGM-POC Pairs with POC between 70-180 mg/dL, CGM-POC Pairs with POC between 181-250mg/dl, CGM-POC Pairs with POC\>250 mg/dl, respectively.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| CGM Patients | Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD) | MARD for CGM-POC pairs with POC <70 mg/dL | 22.6 MARD percentage | Standard Deviation 19.5 |
| CGM Patients | Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD) | MARD for CGM-POC pairs with POC between 70-180 mg/dL | 14.6 MARD percentage | Standard Deviation 12.8 |
| CGM Patients | Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD) | MARD for CGM-POC pairs with POC between 181-250 mg/dL | 11.7 MARD percentage | Standard Deviation 10.1 |
| CGM Patients | Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD) | MARD for CGM-POC pairs with POC>250 mg/dL | 10.4 MARD percentage | Standard Deviation 8.7 |
| CGM Patients | Continuous Glucose Monitors (CGM) Accuracy - Mean Absolute Relative Difference (MARD) | MARD for all CGM-POC pairs | 13.7 MARD percentage | Standard Deviation 12.3 |
Secondary
CGM Recorded Duration of Hyperglycemic Episodes
The duration of hyperglycemic episodes
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| CGM Patients | CGM Recorded Duration of Hyperglycemic Episodes | Hyperglycemic Episodes >180 mg/dL | 2419.44 Minutes | Standard Deviation 2804.85 |
| CGM Patients | CGM Recorded Duration of Hyperglycemic Episodes | Hyperglycemic Episodes >250 mg/dL | 801.99 Minutes | Standard Deviation 1171.28 |
p-value: 0.076Wilcoxon (Mann-Whitney)
p-value: 0.038Wilcoxon (Mann-Whitney)
Secondary
CGM Recorded Duration of Hypoglycemic Episodes
CGM recorded duration of hypoglycemic episodes
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| CGM Patients | CGM Recorded Duration of Hypoglycemic Episodes | Hypoglycemic Episodes <54 mg/dL | 37.810 Minutes | Standard Deviation 191.77 |
| CGM Patients | CGM Recorded Duration of Hypoglycemic Episodes | Hypoglycemic Episodes <70 mg/dL | 162.81 Minutes | Standard Deviation 572.29 |
p-value: 0.203Wilcoxon (Mann-Whitney)
p-value: <0.001Wilcoxon (Mann-Whitney)
Secondary
CGM Recorded Hyperglycemic Episodes
The incidence of hyperglycemic episodes
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| CGM Patients | CGM Recorded Hyperglycemic Episodes | Hyperglycemic Episodes >180 mg/dL | 10.216 Hyperglycemic Episodes | Standard Deviation 12.794 |
| CGM Patients | CGM Recorded Hyperglycemic Episodes | Hyperglycemic Episodes >250 mg/dL | 5.935 Hyperglycemic Episodes | Standard Deviation 8.958 |
Secondary
CGM Recorded Hypoglycemia Episodes
The incidence of hypoglycemic episodes
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| CGM Patients | CGM Recorded Hypoglycemia Episodes | Hypoglycemic Episodes <54 mg/dL | 0.47 Hypoglycemic Episodes | Standard Deviation 1.94 |
| CGM Patients | CGM Recorded Hypoglycemia Episodes | Hypoglycemic Episodes <70 mg/dL | 1.45 Hypoglycemic Episodes | Standard Deviation 3.33 |
p-value: 0.129Wilcoxon (Mann-Whitney)
p-value: 0.008Wilcoxon (Mann-Whitney)
Secondary
Hospitalization Length of Stay
Duration of hospital stay in hours
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Value (MEDIAN) |
|---|---|---|
| CGM Patients | Hospitalization Length of Stay | 116.5 Hours |
Secondary
In Hospital Mortality
Mortality during index hospitalization
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| CGM Patients | In Hospital Mortality | 2 Participants |
Secondary
Mortality Rate at 30 Days
Mortality rate at 30 days post discharge from the hospital
Time frame: From date of enrollment up to 60 days.
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| CGM Patients | Mortality Rate at 30 Days | 0 Participants |
Secondary
Number of Re-hospitalization Events at the Same Institution Within 30 Days From the Discharge Date of the Preceding Admission.
Number of re-hospitalization events that occurred at the same institution within 30 days from the discharge date of the preceding admission.
Time frame: From date of enrollment up to 60 days.
| Arm | Measure | Value (COUNT_OF_UNITS) |
|---|---|---|
| CGM Patients | Number of Re-hospitalization Events at the Same Institution Within 30 Days From the Discharge Date of the Preceding Admission. | 60 Hospitalizations Events |
Secondary
Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters
The correlation between MARD point of care (POC) and Creatinine, Glomerular Filtration Rate (GFR), Bicarbonate, Hemoglobin (Hb), Mean Arterial Pressure (MAP), and Peripheral Oxygen Saturation (SpO2). The Spearman's rank correlation method was used to calculate the correlation coefficient.
Time frame: From date of enrollment and until discharge from the hospital up to 30 days.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| CGM Patients | Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | Creatinine | -0.08 Correlation Coefficient |
| CGM Patients | Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | 139GFR | -0.02 Correlation Coefficient |
| CGM Patients | Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | Bicarbonate | 0.024 Correlation Coefficient |
| CGM Patients | Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | Hb | -0.012 Correlation Coefficient |
| CGM Patients | Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | MAP | 0.003 Correlation Coefficient |
| CGM Patients | Variation in Mean Absolute Relative Difference (MARD) in Relation to Pharmacological and Physiological Parameters | SpO2 | 0.034 Correlation Coefficient |
p-value: <0.0001Spearman Correlation coefficients
p-value: 0.0002Spearman Correlation coefficients
p-value: 0.22Spearman Correlation coefficients
p-value: 0.48Spearman Correlation coefficients
p-value: 0.77Spearman Correlation coefficients
p-value: 0.0516Spearman Correlation coefficients