Chinese clinical guidelines for continuous glucose monitoring (2018 edition)

Blood glucose monitoring is an important part of diabetes management. Continuous glucose monitoring (CGM) technology has become an effective complement to conventional blood glucose monitoring methods and has been widely applied in clinical practice. The indications for its use, the accuracy of the generated data, the interpretation of the CGM results, and the application of the results must be standardized. In December 2009, the Chinese Diabetes Society (CDS) drafted and published the first Chinese Clinical Guideline for Continuous Glucose Monitoring (2009 edition), providing a basis for the standardization of CGM in clinical application. Based on the updates of international guidelines and the increasing evidence of domestic studies, it is necessary to revise the latest CGM guidelines in China so that the recent clinical evidence can be effectively translated into clinical benefit for diabetic patients. To this end, the CDS revised the Chinese Clinical Guideline for Continuous Glucose Monitoring (2012 Edition) based on the most recent evidence from international and domestic studies.


| BACKGROUND
Blood glucose monitoring is an important part of diabetes management. The results of glucose monitoring are crucial for glycemic status assessment, prescription of the optimal treatment regimen, follow-up of patients' glucose status, and timely therapy adjustment. Selfmonitoring of blood glucose (SMBG) is the basic form of blood glucose monitoring, whereas HbA 1c is known as the gold standard for assessing long-term glycemic control. 1 However, both HbA 1c and SMBG have limitations. HbA 1c reflects the average glucose levels over the previous 2 to 3 months and, thus, does not capture important short-term aspects of glycemic control, such as the time in the target range or the severity, frequency, and duration of hyper-and hypoglycemia. In addition, there is a "delayed effect" when using it to guide therapy adjustments. Moreover, SMBG cannot capture fluctuations in blood glucose throughout the day and night because of practical limitations on the number and timing of finger sticks. Instead, continuous glucose monitoring (CGM) technology demonstrates clinical details that cannot be disclosed by conventional fasting plasma glucose and HbA 1c 2 and has been widely applied in clinical practice. More importantly, innovative technologies related to CGM, such as wearables, implants, mobile applications, and cloud technology with professional medical intervention, are emerging, which is promising in changing diabetic patients' lives dramatically. 3 To take maximum advantage of CGM technology, the indications for its use, requisite accuracy of the data that are generated, interpre-  Understanding the intraday and interday interstitial fluid glucose patterns is extremely important for studying the pathogenesis of shortterm and long-term diabetes complications. 12 (Table 1). 13 In 1999, the first CGM system was approved by the US Food and Drug Administration (FDA), 14 [15][16][17] The existing evidence has suggested that patients can achieve better hypoglycemic management under the guidance of a real-time CGM system. Moreover, the improvement in HbA 1c level was found to be positively correlated with the frequency of sensor use, which suggests that better control of blood glucose is related to the frequent use of CGM. 18,19 Retrospective CGM and real-time CGM both exhibit their own features in clinical application (Table 2).  • dramatic glycemic variability; • those patients who deliberately maintain their blood glucose at high levels due to fear of hypoglycemia; • type 2 diabetes patients whose HbA 1c is above target but are not responding to multidrug oral and/or non-insulin injectable therapies.  1. children and adolescents with type 1 diabetes whose HbA 1c less than 7%. The use of real-time CGM can help maintain good glycemic control persistently without increasing the risk of hypoglycemia.
2. children and adolescents with type 1 diabetes whose HbA 1c is greater than or equal to 7% and who are capable of using CGM daily. 1. Blood glucose concentration should be obtained using the same glucose meter and same batch of test strips.

Capillary blood glucose concentrations should be obtained at
different time points throughout the day, preferably during a period when blood glucose is relatively stable (such as before meals and before bedtime).
3. When using a CGM system that requires the input of a capillary blood glucose concentration for calibration, the glucose value should be entered immediately after the result is displayed on the meter. 4. If the user incorrectly enters the blood glucose value for calibration, one should re-enter the correct glucose value as soon as possible.

Recording diet and glucose-related events
During CGM monitoring, diet, exercise, medication, and other events should be recorded in detail.

Daily device maintenance
When wearing a CGM system, the patient should avoid exposure to any strong magnetic field and imaging examinations such as magnetic resonance imaging (MRI). In addition, conventional X-ray imaging and Computed Tomography (CT) scanning should be avoided. Some

| METHODS FOR INTERPRETING CGM GRAPHS
The following points should be noted for interpreting CGM graphs: 1. CGM data can be used to guide treatment regimens when they are accurate. Clinicians should interpret CGM results by illustrating statistical reports or charts, which facilitates optimal communication between clinicians and patients.
2. Downloading CGM data before each follow-up visit is time-saving.
In addition, it is necessary to confirm that the time on the recorder is correct. If the time is not correct, the downloaded results will be wrong, especially for postprandial blood glucose data judged based on a "meal event" icon on the report. 4. For beginners, the practical method to interpret CGM graphs and profiles is "three-step method." For 3-day CGM data, the first step is to analyse the nocturnal blood glucose; the second step is to analyse the preprandial glucose levels; and the third step is to analyse the postprandial glucose levels. In each step, hypoglycemia should be noted first, followed by hyperglycemia, and then the underlying reasons causing abnormal glucose should be identified to guide treatment adjustment. For 14-day CGM data, the first step is to analyse the time interval to achieve the target; the second step is to analyse glucose fluctuations; and the third step is to analyse the risk of hypoglycemia.  Table 3 and the Supplementary Table. These parameters are mostly used for research, and their clinical significance and role in guiding diabetes treatment are still under investigation.
Among the CGM metrics, time in range (TIR) generally refers to the time spent in a patient's target glucose range (usually 3.9-10 mmol/L).
TIR measurements add valuable information to evaluate the glycemic control and were found to be associated with the prevalence of diabetic complications in type 2 diabetes. 100 In recent years, TIRs are recommended as key metrics of glycemic control for evaluating and comparing different glucose-lowering interventions 11   When HbA 1c levels were 6.0%, 6.5%, and 7.0%, the corresponding CGM 24-hour mean glucose levels were 6.6, 7.2, and 7.8 mmol/L, respectively. 112

| Glucose management indicator
Modern CGM technology lasting for 10 days or more of CGM data is usually sufficient for an estimate of mean glucose, time in target range, and time in hyperglycemia, while 14 days or more of CGM data provide a better estimate for time in hypoglycemia and glucose variability.
Using a standard formula, a value called "estimated HbA 1c " was generated from the mean glucose, and the term was later replaced by "glucose management indicator" (GMI). Many patients and clinicians find the GMI to be a helpful educational tool in understanding the CGMgenerated glucose profiles and will facilitate optimal diabetes management and the adjustment of anti-diabetic therapy. 113   Record: measurements in total:_____; MG:_____ mmol/L; SD:_____mmol/L; CV:_____%; maximum and minimum glucose were_____mmol/L and _____mmol/L; percentage of time in target range (3.9-10 mmol/L) was_____%; Percentage of time in hyperglycemic ranges greater than or equal to 7.8 mmol/L, greater than or equal to 10 mmol/L, and greater than or equal to 13.9 mmol/L were_____%,_____% and_____%, respectively; Percentage of time in hypoglycemic ranges less than or equal to 3.9 mmol/L and less than or equal to 2.8 mmol/L were_____h_____min (_____%) and_____h_____min_____(_____%).
Reporter: Reviewer: Report Time: The lack of unified content and format of the CGM report limits the clinical interpretation and application of CGM monitoring results.
Therefore, standardization of CGM reports is extremely important. 114 Currently, CGM monitoring reports generally include three parts: (a) general items: basic information about the subject, clinical diagnosis, inspection date, medical staff signature and date; (b) CGM data; and (c) clinical implications of the results. An example of a CGM report is shown in Table 5. At present, some scholars have developed CGM management software to reduce the clinical workload and to facilitate the optimal management of CGM results. 115 Ambulatory Glucose Profile (AGP) is one of the standardized visualization tools of CGM data that are now available. 11 Key metrics include target range, glucose exposure, glycemic variability, hypoglycemia, and hyperglycemia. The visual display of the AGP pools glucose data as if the data were recorded during a 24-hour period. The result is a single curve representing average glucose values, with interquartile and interdecile ranges shaded to demonstrate glycemic variability. 116 As a standardized report form, AGP is useful in translating glucose data into more actionable information both for clinicians and patients.
In conclusion, to standardize the application of CGM technology in clinical practice, the following elements are required: a clear clinical purpose, strict indications, a standardized reporting system, and appropriate treatment adjustment(s) based on CGM results.

CONFLICT OF INTEREST
No competing financial interests exist.

AUTHOR CONTRIBUTIONS
All authors discussed and drafted the manuscript. WJ, LC and JZ wrote and revised the paper and all authors have read and approved the final manuscript.