Role of Continuous Glucose Monitoring in Insulin-Requiring Patients with Diabetes
2018; Mary Ann Liebert, Inc.; Volume: 20; Issue: S2 Linguagem: Inglês
10.1089/dia.2018.0100
ISSN1557-8593
Autores Tópico(s)Diabetes and associated disorders
ResumoDiabetes Technology & TherapeuticsVol. 20, No. S2 Original ArticlesOpen AccessRole of Continuous Glucose Monitoring in Insulin-Requiring Patients with DiabetesJohn B. WelshJohn B. WelshPublished Online:1 Jun 2018https://doi.org/10.1089/dia.2018.0100AboutSectionsPDF/EPUB Permissions & CitationsPermissionsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail IntroductionThe alluringly simple premise of continuous glucose monitoring (CGM) is that real-time glucose concentration data can be leveraged into better in-the-moment diabetes management decisions that culminate in better outcomes. Retrospective reviews of the same data can lead to pattern discoveries that motivate changes to diet and exercise routines or to the amount and timing of insulin with respect to meals.1Because of insulin's narrow therapeutic index, CGM is particularly appropriate for people with type 1 diabetes (T1D) and for people with type 2 diabetes (T2D) who require it. CGM alarms are potentially lifesaving for people at risk for nocturnal hypoglycemia and any time for people with impaired awareness of hypoglycemia, and the ability of CGM data to be distributed in real time is critical for those involved in the support of others with diabetes. Insulin-induced hypoglycemia continues to limit therapy intensification efforts in diabetes and contributes to many population-based A1C goals not being met. With real-time CGM, the association between lower A1C values and higher rates of hypoglycemia is weakened and may even be abolished.2,3 The information and insights provided by CGM are allowing many people with insulin-requiring diabetes to set and reach more aggressive glycemic goals.Recent Advances in CGMIn the decade since publication of the JDRF CGM study,4 CGM accuracy has improved5 and its use has increased dramatically among youth and young adults with T1D,6 but is still low. As of early 2017, ∼24% of patients in the T1D Exchange Clinic Registry were using CGM.7 Broadening of inclusion criteria and improvements in CGM performance and usability have and will continue to fuel increases in patient adoption. A real-time CGM system (Dexcom G5® Mobile) can be used for insulin dosing decisions, reducing the need for self-monitored blood glucose (SMBG) testing; CGM systems that do not require user calibrations will reduce this need even further. Training materials have improved to the point where self-training on the device is an effective alternative to one-on-one training for most patients,8 allowing clinicians more time to focus on diabetes management. Introducer needles are smaller and hidden, and sensor applicators have improved to make the insertion process simpler.9 Current-generation sensors last longer and resist acetaminophen interference.10 Some instances of skin irritation attributed to the adhesive pads have been addressed by manufacturing process changes,11 and long-term implantable CGM sensors, first available outside the United States, now provide an option for patients concerned with frequent device insertions and removals. Thus, many of the barriers for wider adoption of CGM have been addressed.Evidence and ImplicationsA large and growing body of evidence supports the key messages of CGM and its use as a first-line option for insulin-requiring diabetes. Table 1 summarizes the real-world effectiveness of CGM and flash glucose monitoring in several large industry-sponsored observational studies, as well as a small, early study designed to gather normative data12 and a study of patients in Belgium13 who received funding for various CGM systems from the Belgian healthcare authority. Table 2 summarizes several major clinical trials designed to establish the efficacy of CGM and flash glucose monitoring in controlled settings.Table 1. Observational Studies of Continuous Glucose Monitoring and Flash Glucose MonitoringFirst AuthorPopulationDesignGoal(s)Device(s)Key outcome(s)Bode12T1D or T2D in the United States; blinded CGM fo collect normative dataConsecutive enrollment (n = 101)Establishment of normative glycemic characteristicsFreeStyle Navigator implantable CGMTimes in euglycemia, hypoglycemia, and hyperglycemia were 63%, 8%, and 29%Charleer13T1D/CSII in Belgium starting CGM between September 2014 and December 2016Patients (n = 515) selected by clinicians based on potential benefitExamine changes in glycemic control, acute admissions, and QoL with CGMEnlite (72%); G4 Platinum (24%); FreeStyle Navigator II (2%)CGM associated with lower A1C, fewer hypoglycemia- or DKA-related hospital admissions and admission days; fewer lost workdays; improved QoLDunn21Users of readers worldwide observed between September 2014 and May 2016Convenience sample of data from 50,831 readers that had been uploaded to a remote databaseExamine associations between screen views and glycemic controlAbbott FreeStyle LibreIncreased screen view frequency associated with lower mean glucose, more TIR, less time in hypo- and hyperglycemiaBattelino54T1D or T2D in Western Europe, Canada, or Israel observed between September 2011 and October 2013Convenience sample of data from 10,510 people that had been uploaded to CareLinkaExamine association between proportion of time that CGM system was in use and glycemic controlMedtronic CSII and CGMIncreased CGM wear time associated with lower mean glucose, less glycemic variability, and fewer hypoglycemic episodesParker20CGM users in the United States observed in August 2017Convenience sample of data from 50,000 users to view CGM dataExamine associations between CGM screen views and glycemic controlDexcom G5 MobileIncreased CGM screen views associated with greater TIR and less time in hyperglycemiaParker43Claritya account holders in the United States observed in August 2017Convenience sample of data from 50,000 Clarity account holdersExamine associations between retrospective CGM data views and glycemic controlDexcom G5 MobileFrequent logins to Clarity were associated with greater TIR and less time in hyperglycemiaParker36Shareb account holders in the United States observed in May 2017Convenience sample of data from 4511 children ages 2–10Examine associations between having one or more real-time followers and glycemic controlDexcom Share and FollowbIncreased number of followers was associated with more sensor utilization and less time in hypoglycemiaaCareLink and Clarity are web-based platforms for data storage and visualization from Medtronic and Dexcom, respectively.bShare and Follow are apps from Dexcom for real-time distribution and remote monitoring of CGM data.CGM, continuous glucose monitoring; CSII, continuous subcutaneous insulin infusion; DKA, diabetic ketoacidosis; QoL, quality of life; T1D, type 1 diabetes; T2D, type 2 diabetes; TIR, time in range.Table 2. Major Clinical Trials of Continuous Glucose Monitoring and Flash Glucose Monitoring in Adults with DiabetesStudyPopulationDesignGoal(s)Device(s)Key outcome(s)Diamond T114NCT02282397T1D in the United States;A1C 7.5%–9.9%Randomized 2:1 to CGM (n = 105) or usual care (n = 53) for 24 weeksA1C reductionDexcom G4Between-group difference of 0.6% points in favor of CGM, P < 0.001. Significant reduction in hypoglycemia in the intervention group.Diamond T216NCT02282397T2D in the United States;A1C 7.5%–9.9%Randomized 1:1 to CGM (n = 79) or usual care (n = 79) for 24 weeksA1C reductionDexcom G4Between-group difference of 0.3% points in favor of CGM, P = 0.022Gold15NCT02092051T1D in Sweden;A1C ≥7.5%Crossover CGM vs. usual care; Randomized 1:1 to 26 weeks of CGM before (n = 82) or after (n = 79) 26 weeks of usual careA1C reductionDexcom G4Between-group difference of 0.43% points in favor of CGM, P < 0.001. Significant reduction in hypoglycemia in the intervention group.I Hart CGM55NCT03028220T1D in the United Kingdom;Gold score ≥4 or recent severe hypoglycemiaRandomized 1:1 to CGM (n = 20) or flash glucose monitoring (n = 20) for 8 weeksHypoglycemia reduction, CGM vs. flash glucose monitoringDexcom G5, Abbott FreeStyle LibreCGM reduces hypoglycemia more effectively than flash glucose monitoringHypoDE34NCT02671968T1D in Germany;History of impaired hypoglycemia awareness or severe hypoglycemia in past yearRandomized 1:1 to CGM (n = 75) or usual care (n = 74) for 26 weeksHypoglycemia reduction in high-risk individualsDexcom G5Incidence of hypoglycemic events fell by 72% for CGM group, P < 0.0001Comisair56(no NCT number)T1D/MDI or CSII in Czechia;A1C 7.0%–10%Nonrandomized: CGM (n = 27) or SMBG (n = 38) for 52 weeksA1C and hypoglycemia reductionDexcom G4, Medtronic EnliteComparable reductions in A1C and hypoglycemia in CGM/MDI and CGM/CSII groupsIN CONTROL57NCT01787903T1D/MDI in the Netherlands;Impaired hypo awareness (Gold score ≥4)Randomized crossover: CGM then SMBG (n = 26) or SMBG then CGM (n = 26)Hypoglycemia reduction in high-risk individualsMedtronic EnlitePeriods of CGM use associated with more TIR, less time in hypo- and hyperglycemia, fewer severe hypoglycemic events.IMPACT58NCT02232698T1D/MDI in Sweden, Austria, Germany, Spain, and the NetherlandsA1C ≤7.5%Randomized 1:1 to flash glucose monitoring (n = 120) or usual care (n = 121)Hypoglycemia reductionAbbott FreeStyle LibreFlash glucose monitoring associated with significant reduction in hypoglycemiaREPLACE59,60NCT02082184T2D/MDI or CSII in France, Germany, and the United KingdomA1C 7.5%–12%Randomized 2:1 to flash glucose monitoring (n = 149) or usual care (n = 75)A1C reductionAbbott FreeStyle LibreFlash glucose monitoring associated with significant reduction in hypoglycemia but no difference in A1C changeMDI, multiple daily injections; SMBG, self-monitored blood glucose.In 2017, results of the Diamond14 and Gold15 randomized clinical trials were published simultaneously. Both studies were designed to compare CGM to usual care in people with T1D treated with multiple daily injections (MDI), and the Diamond study had a separately powered arm investigating CGM in the context of MDI-treated T2D.16 Both studies enrolled subjects with above-target A1C values (Diamond, 7.5%–9.9%; Gold, ≥7.5%), and both were designed to be relatively "low touch" with respect to encounter frequency.In both studies, patients and clinicians were provided with a one-sheet set of general and patient-specific guidelines that had been shown earlier17 to improve glucose control in MDI-treated patients with either T1D or T2D. General guidelines were to wear the CGM as much as possible; to look at the display device frequently; to maintain reasonable expectations; to approach alerts and alarms as helpful rather than punitive; to share the CGM data with others; and to learn from experience. Personalized guidelines involved establishing and using individual glucose targets; having a plan for preventing or responding to low glucose; knowing the plan for using CGM trends to improve prandial insulin dosing (remembering that an SMBG test is needed); and responding to high glucose levels between meals without stacking insulin. Adherence to these guidelines was not monitored in Diamond or Gold; however, they were felt to be understandable and easy to implement.17Wearing and checking the systemIn randomized controlled trials such as the JDRF CGM study4 and the STAR 3 study,18 sporadic wear of CGM was associated with sporadic A1C reductions, and wear times less than 6 days/week were associated with no A1C benefit in a separate study of youth with T1D.19 However, wearing the system is not enough; the data must be viewed and acted upon.In examining data from 50,000 anonymized real-time CGM users, Parker et al.20 found a monotonic relationship between screen view frequency and time in range (TIR): the cohort of patients with the least-frequent screen views averaged 8%) may find that the default high limit needs to be increased to perhaps 250 mg/dL.26 The recommended low alert level of 70 mg/dL generally gives the patient time to take action to prevent a clinical hypoglycemic episode and provides some margin for inaccurate readings,27 but a more conservative setting such as 80 mg/dL has been advocated to provide patients with more time to act.28 Hypoglycemia alerts <80 mg/dL and hyperglycemia alerts 6 days/week) CGM users33 highlighted three key behaviors related to this goal. First, patients reduced or eliminated correction and mealtime bolus insulin doses in response to downward trending arrows. Second, patients responded to impending hypoglycemia with carbohydrates well before becoming symptomatic. Third and most consequentially, most survey respondents reported being awoken by low-glucose alerts at least once a week, enabling them to minimize or avoid nocturnal hypoglycemia.Reductions in hypoglycemia are an important outcome in clinical trials. In the Diamond study,14 the median number of minutes per day with sensor glucose values <50 mg/dL fell by more than half (from 13 to 6) in the CGM group and rose from 18 to 20 in the control group (P < 0.001). More recently, the HypoDE study34 aimed to quantify the effect of CGM usage among hypoglycemia-prone individuals (evidenced by a history of impaired hypoglycemia awareness or severe hypoglycemia during the previous year) with T1D treated with MDI. In this study, the mean number of hypoglycemic events per 28 days among participants in the CGM group was reduced from 10.8 to 3.5; reductions among control participants were negligible (from 14.4 to 13.7). The relative contribution of daytime and nocturnal alerts, trend arrows, or unprompted screen views to these favorable reductions is a topic for further study. The availability of alerts and alarms in real time but not intermittently scanned CGM systems should play a role in selecting the CGM type that is most appropriate for each patient.35Real-time sharingSharing of CGM data allows outsourcing of some short-term diabetes management decisions and provides patients with more opportunities for collaborative pattern discovery and shared decision-making. As an example, Dexcom's "Share" and "Follow" mobile applications were designed to allow CGM data from a single "Sharer" to be distributed in real time to up to five "Followers" running the corresponding apps. Sharing is especially useful for patients who are unable to fully manage their own disease, such as young children; following is especially useful for caregivers who can intervene in a timely manner, such as parents of young children. In an exploratory analysis of Sharers ages 2–10, those with no paired Followers had 16% of their sensor glucose values in the <70 mg/dL range, whereas children with five Followers had only 8.7% of their values in this range.36 A more modest 14% reduction in the percentage of sensor glucose values <70 mg/dL was found in a similar analysis of adults 65 years and older with at least one Follower.37 Among adolescents, the effects of sharing CGM data are more variable and depend on the context of family dynamics: at the extremes, the technology can be viewed either as a trusted best friend contributing to independent decision-making or as an unwelcome parental spy.38Patients are often advised to fold CGM reviews into their daily routines and build narratives around their glucose levels and activities. In an analysis of adults in the JDRF CGM trial,4 Ritholz et al.39 characterized "Responders" as those with baseline A1Cs ≥7% and A1C reductions ≥0.5% points or those whose time <70 mg/dL decreased while maintaining A1Cs <7%. In either case, Responders gave credit to retrospective pattern analysis of their CGM data and to "significant other" or spousal involvement for the effective use of CGM.Retrospective reviewingWhile CGM data gathered in real time can guide in-the-moment management decisions, it can also inform longer term decisions related to patterns identified through retrospective data analysis. Software for CGM data visualization ranges from general purpose (Excel, Tableau) to manufacturer-specific web-based options such as CareLink® (Medtronic), Clarity® (Dexcom), and LibreView™ (Abbott). The ambulatory glucose profile40 is a device-agnostic, single-page report that is designed for rapid, easy interpretation of SMBG and/or CGM data. Its standardized reports have recently been shown to save time in the context of clinic visits, translating into more time for shared decision-making and education.41Data from Dexcom Clarity suite of mobile and web-based software for retrospectively sharing and analyzing CGM data confirm the benefits of these behaviors: in a recent analysis of 50,000 users of the Dexcom G5 Mobile app, patients who logged in four or more times to Clarity during August 2017 had significantly more TIR, lower mean sensor glucose values, and less time in hyperglycemia than patients who did not log in during this interval.42 A recent update allows Clarity users to opt in to weekly push notifications and/or e-mail messages that include summary statistics, patterns, and comparisons to the week before. Opting in to this service lowers the effort required to obtain the reports and reframes weekly retrospective reviews as the default behavior.SMBG testing and quality of lifeOn December 20, 2016, the U.S. Food and Drug Administration (FDA) approved labeling for the Dexcom G5 Mobile CGM System for making diabetes treatment decisions without the need for a confirmatory blood glucose measurement.43 This decision was at odds with a minority viewpoint suggesting that CGM data might suffice for some diabetes treatment decisions (such as carbohydrate ingestion, exercise, meal timing, and sleep), but were not accurate enough for dosing insulin.44 Three weeks later, the System was reclassified by the Centers for Medicare and Medicaid Services as durable medical equipment able to replace BGMs.45Given the shortcomings of SMBG testing,46 these decisions were welcomed by the community and were seen as legitimizing the reported behavior of many participants in the T1D Exchange Registry with respect to "dosing off the CGM."47 The quantity and quality of CGM data—hundreds of point estimates per day, trends, and alarms—far exceed that provided by home BGMs. The safety of CGM-based dosing decisions was confirmed in the REPLACE-BG study.48With the FDA decision allowing nonadjunctive use, the required number of SMBG tests decreased for most CGM users in the United States, but periodic SMBG tests are still required to calibrate most CGM systems. Recent advances in sensors and algorithms have resulted in CGM systems that are calibrated during the manufacturing process and are stable enough so as to not require calibrations during use, further reducing the need for SMBG testing. SMBG results will remain useful when CGM data are unavailable (e.g., during sensor warm-ups) or when there are mismatches between the CGM data and patient's expectations or symptoms. Users of the FreeStyle Libre flash glucose monitoring system (Abbott) are also advised to perform an SMBG test when experiencing potential hypoglycemia or hyperglycemia; when the system's glucose values are changing by more than 2 mg/(dL·min); when the system's display does not include a current glucose number or trend arrow; or to confirm system-reported hypoglycemia or impending hypoglycemia.Reductions in hypoglycemia, fewer SMBG tests, and fruitful engagement of others in diabetes management decisions may all contribute to improved quality-of-life (QoL) outcomes for people using CGM. In the Diamond study, participants completed several questionnaires related to QoL.49 Compared with subjects in the control group, those in the CGM group had greater increases in hypoglycemic confidence and greater decreases in diabetes distress. A separate observational cohort study of 515 adults with T1D on continuous subcutaneous insulin infusion (CSII) therapy (the RESCUE trial13) was prompted by the Belgian health authority's decision to reimburse real-time CGM systems for 25% of CSII users; specialized centers across Belgium were allowed to decide which patients would receive reimbursement for their systems. After at least a 12-month use of real-time CGM, subjects had improved glycemic control, lower risk of acute diabetes-related complications, and higher QoL attributed to less fear of hypoglycemia. CGM use was also associated with reduced rates of hypoglycemia-related hospitalizations and less work absenteeism, which could offset some of the costs of this technology.Prospects for the Near TermThe next few years for CGM systems will hold challenges and opportunities. Because CGM is a young technology and still undergoing rapid and sometimes abrupt changes, interpretation of longitudinal data is challenging because early conclusions were often reached with now-obsolete hardware and software. Compared with recommendations based on SMBG test results, recommendations for lifestyle and insulin dosage changes based on the direction and degree of CGM trend data are more nuanced and variable.50 Although some systems no longer require a dedicated receiver and some sensors are implantable, all real-time systems require a display device and a skin-adherent patch. Although the amount of numerical data can be intimidating to some patients, beneficial A1C reductions in the Diamond study did not depend on educational attainment, numeracy scores, or age. Clinical inertia in primary care settings,51 out-of-pocket costs, and the perceived need for extensive specialized training remain barriers to wider adoption of CGM.On March 27, 2018, the FDA created a new category of "integrated continuous glucose monitoring (iCGM)" systems. The Dexcom G6® CGM System was the first device in this category and is subject to special controls for class II that provide assurances of its safety and effectiveness. Importantly, the new classification will allow for more rapid innovation, as some changes that would have required FDA premarket approval as a class III device may now be reviewed by FDA in less time. Certain changes may even be implemented without prior FDA review once all quality system and regulatory documentation requirements are fulfilled. The G6 system is indicated to replace fingerstick blood glucose testing for diabetes treatment decisions and to autonomously communicate with digitally connected devices, including automated insulin delivery systems.The low penetration of CGM among MDI-using patients implies a large number of people who have yet to try it. The latest generation of systems promises improved accuracy, durability, ease of use, and comfort. Importantly, CGM data are easily integrated with mobile health apps that provide users with information from other sources related to insulin delivery, food, and activity. Thanks to the interoperability of the G6 System, CGM data may soon be incorporated into smart insulin pen systems that track insulin on board and provide dosing recommendations. CGM is also an emerging area of investigation for reduction of hypoglycemia and severe hypoglycemia in hospitalized patients,52 which are associated with poor outcomes. Finally and as discussed elsewhere in this supplement, CGM is an essential component of artificial pancreas systems.SummaryCGM is a representative case study in patient engagement and participatory medicine. It illustrates how technological developments and patient advocacy can combine to produce a breakthrough option for millions of people who depend on insulin. It enables patients to make better-informed decisions regarding diet, exercise, and insulin dose adjustments, improves the quality of glycemic control, reduces exposure to hyper- and hypoglycemia, and facilitates attainment of lower target A1C and average glucose levels. CGM use allows for fewer SMBG tests and improves the QoL, regardless of insulin delivery method. When shared with other individuals, the information provided by CGM systems contributes to active and collaborative decision-making. Thanks to recent improvements in algorithms and devices and to the large and growing evidence base in its favor, CGM adoption is increasing. Until a "perfect" risk-free insulin53 or a viable strategy for pancreatic beta-cell replacement is found, CGM will remain an essential technology and should be viewed as a first-line option and the standard of care in treating type 1 and insulin-requiring type 2 diabetes.Author Disclosure StatementJ.B.W. is an employee of and has a financial interest in Dexcom, Inc. J.B.W. thanks Drs. Andrew Parker, David Price, Sarah Puhr, and Tomas Walker for discussions on the topic of CGM.References1 Lee V, Thurston T, Thurston C: A comparison of discovered regularities in blood glucose readings across two data collection approaches used with a type 1 diabetic youth. Methods Inf Med 2017;56:e84–e91. Crossref, Medline, Google Scholar2 Gimenez M, Tannen AJ, Reddy M, et al.: Revisiting the relationships between measures of glycemic control and hypoglycemia in continuous glucose monitoring data sets. Diabetes Care 2018;41:326–332. Crossref, Medline, Google Scholar3 van Beers CAJ, Caris MG, DeVries JH, Serne EH: The relation between HbA1c and hypoglycemia revisited; a secondary analysis from an intervention trial in patients with type 1 diabetes and impaired awareness of hypoglycemia. J Diabetes Complications 2018;32:100–103. Crossref, Medline, Google Scholar4 Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group, Tamborlane WV, Beck RW, et al.: Continuous glucose monitoring and intensive treatment of type 1 diabetes. N Engl J Med 2008;359:1464–1476. Crossref, Medline, Google Scholar5 Facchinetti A: Continuous glucose monitoring sensors: past, present and future algorithmic challenges. Sensors (Basel) 2016;16:E2093. Crossref, Medline, Google Scholar6 Miller K, Foster N, DeSalvo D, et al.: Continuous glucose monitoring (CGM) use in type 1 diabetes: an update from the T1D exchange clinic registry. Pediatric Diabetes 2016;17(S24):49. Medline, Google Scholar7 Foster NC, Miller KM, DiMeglio LA, et al.: Marked increases in CGM use has not prevented increases in HbA1c levels in participants in the T1D Exchange (T1DX) Clini
Referência(s)