Manufacturers continue to improve performance and usability of continuous glucose monitoring (CGM) systems. As CGM becomes a standard of care, especially for people on insulin therapy, it is important to routinely gauge how satisfied people with diabetes are with this technology. This article describes survey feedback from a large cohort of people with diabetes using older and current CGM systems and highlights areas of current satisfaction, concern, and future system improvement.

Key Points
  • Ninety percent of survey respondents agreed that the majority of continuous glucose monitoring (CGM) sensors were accurate. However, only 79 and 78%, respectively, were satisfied with sensor performance on the first and last day of wear.

  • Forty-two percent agreed that accuracy varies from sensor to sensor, with 54% experiencing skin reactions or irritation using sensors.

  • Thirty-five percent were concerned about the impact of over-the-counter or prescription medications (e.g., cold and flu remedies or pain relief products) on sensor accuracy.

  • Thirty-six percent agreed that inaccurate CGM alarms or alerts negatively affected daily life, and 34% agreed that they negatively affected diabetes management.

A systematic literature review and a meta-analysis of real-world observational studies confirm that usage of continuous glucose monitoring (CGM) is associated with improved glycemic measures (1,2). Such improvements are welcome, especially given that glycemic metrics in the general U.S. adult population declined between 1999 and 2018, as evidenced by a recent National Health and Nutrition Examination Survey analysis (3). Even in the more highly technology-supported T1D Exchange population, only a minority of people with diabetes met American Diabetes Association–recommended A1C targets despite CGM usage increasing from 7 to 30% and insulin pump usage increasing from 57 to 63% in this registry from 2010 to 2018. Encouragingly, subanalysis of registry data demonstrated that A1C levels were lower in CGM users than in nonusers of CGM (4).

Despite the many advantages of CGM, there remain areas for improvement. From the standpoint of accuracy, there are notable deficiencies in the performance of the Dexcom G6 and FreeStyle Libre 2 CGM systems on day 1 of sensor application, with mean absolute relative difference (MARD) values of 18.5 and 13.2%, respectively (5).

Most studies of CGM performance focus on overall accuracy; yet, sensor-to-sensor accuracy is often highly variable (even in well-controlled regulatory studies) and suboptimal for insulin dosing decisions. For example, the Dexcom G7, which has a nonadjunctive indication, has an overall MARD of 8.2% in adults and 8.1% in children/adolescents, yet only 71.4 and 73.8% of sensors, respectively, achieved a MARD <10%—a recognized criteria for safe insulin dosing (6,7). Data on sensor-to-sensor variation for the FreeStyle Libre 2 and 3 systems are more elusive. The original FreeStyle Libre regulatory study indicated that ∼58% of sensors achieved a MARD ≤10% (8). People with diabetes may also experience variations in CGM sensor accuracy when glucose is changing rapidly (9) or at different times throughout the same day, triggered by routine activities of daily life (10) such as exercise (11), which may lead some people to question their sensor readings and require them to confirm readings using blood glucose monitoring (BGM). Additionally, high rates of sensor failure may lead to patient frustration and an inability to source sufficient sensors to ensure that CGM continues uninterrupted.

Clinicians also perceive barriers to endorsing CGM; 61% thought there were too many alarms, and 46% thought patients would not understand what to do with the information or how to use the features. Interestingly, people with diabetes were less inclined to list such barriers (12). In support of this concern on the part of clinicians, a study in 222 people with diabetes found dramatic variations in insulin correction-dose behavior in response to CGM rate of change indicators, with up to 24% of respondents stating they would omit making any adjustment in response to these insight features (13). In a study of 249 people with type 1 diabetes who discontinued CGM, reasons cited included too many alarms (32.1%), it was not accurate (30.1%), or they simply did not trust it (18.1%) (14). Lack of trust is given further weight by an analysis showing low concordance between paired reference BGM and FreeStyle Libre readings, in which only two BGM readings substantiated 238 “LO” (low glucose) Libre readings <40 mg/dL. Less than half of BGM readings confirmed 102 Libre readings between 40 and 70 mg/dL, with the authors recommending strong caution when interpreting sensor readings (15).

Skin reactions to CGM and insulin pumps remain a concern. A recent study found that 60% of people with type 1 diabetes reported skin complications, 22% of whom discontinued using their device (16). This finding was reinforced by an online survey of dermatological complications in 139 children and adolescents who reported mild (33%), moderate (20%), or severe (4%) skin reactions to their CGM sensor (17). Importantly, over- the-counter (OTC) cold and flu remedies, nutritional supplements, and polypharmacy may also affect the reliability and accuracy of CGM readings. An increasing number of interferents are being identified, to date including acetaminophen, salicylic acid, galactose, xylose, ascorbic acid, hydroxyurea, mannitol, tetracycline, ethanol, red wine, lisinopril, albuterol, and atenolol (1826). An elegant in vitro methodology for investigating the impact of single agents or cocktails of potential interferent substances was developed and subsequently validated using the Dexcom G6 and FreeStyle Libre 2 systems (27). Our current in-depth survey of people with diabetes from the T1D Exchange sought to probe their level of acceptance of and satisfaction with current CGM systems. We also ascertained to what extent people with diabetes were concerned about the performance of their CGM system and their confidence in terms of relying on it to make diabetes management decisions.

This was an observational, noninterventional, convenience sample focus group and online survey study. Survey topics were informed based on feedback from three live focus groups with a total of 24 people with type 1 diabetes (20 current and 4 former CGM users) in three separate live sessions conducted via the Zoom digital platform. Key themes and experiences using current CGM systems were recorded by T1D Exchange staff. The online survey of 605 people with diabetes was developed to further investigate the themes identified in the focus groups and from information in existing CGM literature.

Observational studies are submitted to an institutional review board (IRB) for approval or waivers sought whenever required by local law. The WCG IRB determined exempt status for the CGM survey protocol, associated surveys, and recruitment materials on 24 February 2021.

Survey data were collected using an online questionnaire hosted on Alchemer. The type of data collected in this study involved the experience of, satisfaction with, and reasons for discontinuing CGM among individuals with diabetes (Supplementary Material—Survey). Demographic and self-reported clinical and socioeconomic information was assessed in the survey for data analysis.

Recruitment criteria included age ≥18 years, diagnosed with type 1 or type 2 diabetes for >1 year, current or former users of CGM (used for at least 1 year), current residence in the United States, fluency in written English, agreement to provide informed consent, and not currently pregnant. Participants were recruited via advertisement e-mails to the T1D Exchange online community, T1D Exchange registry, and other diabetes online communities. People with type 2 diabetes who participated in the CGM survey were recruited from the Dynata platform, a third-party platform selected by the T1D Exchange. The advertisement e-mail contained a link to the electronic consent form for each participant’s review and electronic signature. The CGM survey opened to participants on 3 June and closed on 25 June 2021.

Descriptive statistical analyses were conducted on all survey questions. Categorical variables such as sex and race were summarized by frequencies and percentages. Continuous variables such as age and A1C were summarized by means and SDs or medians. Open-ended free-text questions were summarized by the study investigators into themes. Key questions related to the experience of and satisfaction with CGM use were further analyzed by subgroups such as by the use of common CGM models and by diabetes types. Responses, on a five-point Likert scale, were condensed into three categories for subgroup analyses, with “agree” and “strongly agree” combined into one “agree” category, and “disagree” and “strongly disagree” combined into one “disagree” category.

Characteristics of CGM Survey Participants

Of the 605 patients surveyed, 83% were people with type 1 diabetes, and 17% were people with type 2 diabetes, with a median age of 39 years (38 years for those with type 1 diabetes and 43 years for those with type 2 diabetes). The median duration of diabetes was 14 years longer among people with type 1 diabetes (20 years) than those with type 2 diabetes. Self-reported mean A1C values were significantly higher in people with type 2 diabetes (8.7%) than for those with type 1 diabetes (7.0%).

All respondents were treated with insulin, with more people with type 1 diabetes than with type 2 diabetes using insulin pumps (81.3 vs. 34.7%). A far higher proportion of people with type 2 diabetes than with type 1 diabetes were injecting insulin using a pen (45.5 vs. 21.8%) or syringe (17.8 vs. 10.7%). More than one-third of respondents with type 2 diabetes were taking noninsulin diabetes medications in addition to insulin, compared with only 5.5% of those with type 1 diabetes.

Ninety-three percent of all respondents (n = 563) currently used CGM, whereas 42 (6.9%) no longer used CGM. The proportion of past users was higher among respondents with type 2 diabetes (19%) than those with type 1 diabetes (5%). Of current CGM users, 100% had been using CGM for at least 1 year, with >30% using CGM for ≥5 years. By far the most used CGM system among all participants was the Dexcom G6, representing 60.7% (n = 342) of all current CGM users. The Dexcom G6 system was used by 69% of people with type 1 diabetes but only 12% of those with type 2 diabetes. By contrast, the older Dexcom G4 and G5 systems were used by 32% (n= 26) of people with type 2 diabetes, potentially reflecting a slower movement to the newer G6 system, whereas only 2% (n = 9) of people with type 1 diabetes used a G4 or G5 system. Somewhat surprisingly, Freestyle Libre systems were used by only 10% of people with type 1 diabetes, whereas use was higher in those with type 2 diabetes, representing 35% of current CGM users (n = 28). A variety of Medtronic systems were used, with a notable finding being the relatively high usage of the latest hybrid closed-loop automated insulin delivery (AID) systems (MiniMed 670G and 770G) by 15% (n = 71) of people with type 1 diabetes. A variety of Medtronic systems were also used by people with type 2 diabetes, reaching 15% (n = 13) overall, compared with 20% overall (n = 95) in people with type 1 diabetes. A further disparity between respondents with type 1 diabetes and those with type 2 diabetes was discovered regarding medication burden, although the sample size for those with type 2 diabetes was far smaller. Nonetheless, 85% of respondents with type 2 diabetes were taking at least three medications per day, compared with 61% of those with type 1 diabetes, and 43% of individuals with type 2 diabetes were taking at least six medications per day, compared with only 25% of those with type 1 diabetes (Figure 1).

FIGURE 1

Characteristics of CGM survey respondents.

FIGURE 1

Characteristics of CGM survey respondents.

Close modal

Perceptions of Skin Reactions and Pain Using CGM

More than half of all people with diabetes experienced skin reactions or irritation when using CGM sensors. A higher incidence was observed in people using the older Dexcom G4 and G5 systems (69%) compared with the more recent G6 system (50%), although data for the G4 and G5 systems were insufficient to draw firm conclusions. Fewer people with diabetes were using FreeStyle Libre systems, and for them, the percentage of people experiencing skin issues ranged from 37 to 48% with the Libre and Libre 2, respectively. A relatively high percentage of people with diabetes using the MiniMed 670G or 770G AID systems experienced skin issues (64%). There are reports of skin reactions from infusion sets (28). Given that 78% of our respondents were using insulin pumps, it is conceivable that some of the feedback about skin issues may have been conflated with negativity arising from the pump infusion set rather than the sensor per se. Nominally at least, the MiniMed 670G and 770G sensors were the most painful to apply (71%), followed by the Dexcom G4 and G5 (69%), the Dexcom G6 (54%), and the FreeStyle Libre (42%), with the least painful system being the FreeStyle Libre 2 system (24%) (Figure 2).

FIGURE 2

Pain and skin issues associated with CGM systems.

FIGURE 2

Pain and skin issues associated with CGM systems.

Close modal

Selected CGM Feedback: A Focus on Sensor Accuracy

Views on sensor accuracy were mixed. Although 90% of respondents agreed that the majority of sensors were accurate, fewer were satisfied with sensor performance on the first (79%) or last (78%) day. More concerning was the observation that 42% suspected variations in accuracy from sensor to sensor and that 15% felt that <60% of sensors could be described as accurate. Given this feedback, perhaps it is unsurprising that 32% continued to perform BGM more than six times per week with a blood glucose meter. Furthermore, 59% had previously interrupted use of their CGM for at least 1 month (Figure 3).

FIGURE 3

Selected CGM survey responses (all subjects).

FIGURE 3

Selected CGM survey responses (all subjects).

Close modal

Selected CGM Feedback: People With Type 1 Diabetes Versus Those With Type 2 Diabetes and Responses About Specific CGM Systems

Respondents with type 1 diabetes overwhelming chose the Dexcom G6 system for glucose monitoring compared with those with type 2 diabetes (69 vs. 12%), whereas more respondents with type 2 diabetes used a FreeStyle Libre system compared with those with type 1 diabetes (35 vs. 10%). On two related topics, individuals with type 2 diabetes were far more inclined to be concerned than those with type 1 diabetes about the impact of poor sensors on their confidence dosing insulin (50 vs. 21%) or their ability to make diabetes management decisions (52 vs. 19%). Users of the older Dexcom systems expressed more concern regarding the perceived impact of medications on sensor accuracy (86% for G4/G5 users vs. 24% for G6 users) and more concern that inaccurate alarms negatively affected diabetes management (74% for G4/G5 vs. 34% for G6) (Figure 4).

FIGURE 4

Selected CGM survey responses (people with type 1 diabetes [T1s] versus those with type 2 diabetes [T2s] and responses from users of specific CGM systems).

FIGURE 4

Selected CGM survey responses (people with type 1 diabetes [T1s] versus those with type 2 diabetes [T2s] and responses from users of specific CGM systems).

Close modal

Reasons Respondents Temporarily Stopped Using CGM

Three hundred and four (304) current CGM users who had previously interrupted use of their CGM were included in this analysis. Inaccurate sensor readings were cited as the main reason for temporarily stopping (n = 108 respondents), followed by sensors falling off prematurely (n = 94), cost of sensors (n = 93), skin irritation (n = 85), or simply taking a break from being attached to a device (n = 84). Alarm fatigue and false alarms are also a recognized barrier to CGM adherence (12,14), and this obstacle was confirmed by many respondents to the survey (Figure 5).

FIGURE 5

Reasons respondents stopped using CGM.

FIGURE 5

Reasons respondents stopped using CGM.

Close modal

Reasons CGM Users Performed BGM

Four hundred and ninety-five (495) CGM users who routinely performed BGM were included.

Not trusting sensor readings was selected as the top reason for performing BGM, followed by a desire or need to calibrate the CGM system. People using CGM are advised by manufacturers to perform confirmatory BGM for various reasons, and one such reason is when CGM users feel that symptoms they are experiencing do not match sensor readings. In fact, “When I feel symptoms” was the fourth highest respondent choice. A number of other clinical scenarios, including treating a low glucose, dosing insulin, and testing before going to sleep, were identified as important situations in which respondents felt that the extra reassurance of performing BGM alongside CGM is warranted (Figure 6).

FIGURE 6

Reasons CGM users also performed BGM.

FIGURE 6

Reasons CGM users also performed BGM.

Close modal

Substances or Situations That May Affect Sensor Accuracy

All survey participants (n = 605) responded to this question and could select one or all applicable substances or situations from a list of 11 options, as well as a free text “other” option in which they could specify further items. Surprisingly, dehydration was the top choice (47%), despite a lack of published literature on a direct impact of dehydration on sensor accuracy. This was closely followed by concerns regarding the impact of pain relief medications (43%), cold and flu remedies (32%), coffee (24%), red wine (15%), or vitamin C tablets (14%). Fewer respondents had concerns about the impact of prescribed medications for concomitant health conditions on the accuracy of their sensor readings, including blood pressure (11%), respiratory (8%), or lipid-lowering (6%) medications (Figure 7).

FIGURE 7

Substances or situations that may affect sensor accuracy. T1D, type 1 diabetes; T2D, type 2 diabetes.

FIGURE 7

Substances or situations that may affect sensor accuracy. T1D, type 1 diabetes; T2D, type 2 diabetes.

Close modal

Factors Affecting Respondents’ Confidence Using CGM

Despite advancements in CGM technology, significant numbers of respondents felt that CGM systems often perform suboptimally in some scenarios. More than 25% of users felt that inaccurate sensors often or very often affected their confidence dosing insulin, with a further 12% undecided about CGM performance in this regard. CGM manufacturers have publicized a number of known interferents, which may explain why >66% of respondents were aware of substances that affect accuracy. In fact, only 41% of respondents reported being “unconcerned” about how OTC or prescription medications might compromise sensor readings. Additionally, inaccurate (or false) alarms or alerts negatively affected the daily life (36%) and diabetes management (34%) of people with diabetes (Figure 8).

FIGURE 8

Factors affecting confidence when using CGM.

FIGURE 8

Factors affecting confidence when using CGM.

Close modal

The current survey provides evidence that a majority of people with diabetes are satisfied with many aspects of the usability or performance of their CGM systems. However, significant numbers of survey respondents expressed a lack of trust in CGM performance under certain conditions, had specific concerns, or expressed dissatisfaction. In addition, many respondents were neutral when appraising the value of their CGM system. Typically, patient feedback on CGM is gathered during controlled, short-term clinical studies, providing so-called “voice of the customer” feedback based on selected questions crafted by the manufacturer that tend to focus on statements that could deliver supportive product claims. By contrast, our wide-ranging survey (containing 56 questions) was developed after listening to live conversations regarding CGM within three groups of people with diabetes conducted via the Zoom platform.

To strengthen our findings per topic, we recruited more than 600 people with diabetes with significant real-world and long-term experience using one or more CGM systems, with >80% using CGM for ≥2 years and >30% using CGM for ≥5 years. Regarding our survey sample, we must be cognizant that people with diabetes in the T1D Exchange registry are typically more avid users of the latest diabetes technology devices, are more likely to be non-Hispanic White, and possibly attend more specialized endocrinology clinics in the United States. This fact was underpinned by the fact that self-reported A1C was 7.0% on average in people with type 1 diabetes in our survey.

This context has implications for how we interpret and translate our data to people with type 1 diabetes who are not treated in the top U.S.-based endocrinology clinics. Arguably, we would anticipate that respondents in our survey received the best ongoing training and support on how to maximize the value of their CGM for diabetes management, including advice on problem-solving and guidance on common factors that are known to compromise CGM performance. To this end, the respondents with type 1 diabetes in our survey might be expected to be more proficient using CGM and, therefore, less inclined to give negative feedback across the board than one might obtain in a survey conducted with people with diabetes managed in the wider community.

It is notable that our smaller sample of people with type 2 diabetes recruited via the Dynata database, a third-party platform accessing a more diverse population, had considerably poorer glycemic control, with an average A1C of 8.7%. Broadly speaking, in our survey, people with type 2 diabetes were less trusting and more concerned about the performance of their CGM systems than those with type 1 diabetes. For example, 58% of respondents with type 2 diabetes felt that accuracy varies from sensor to sensor, compared with 38% of those with type 1 diabetes, and 50% of people with type 2 diabetes agreed that poor sensor accuracy affected their confidence dosing insulin compared with 21% of those with type 1 diabetes. Similar trends were observed regarding concerns about the impact of medications on sensor accuracy and the negative impact of inaccurate alerts or alarms on daily life and diabetes management.

It is plausible that this discrepancy could be explained by the higher proportion of people with type 2 diabetes using older Dexcom G4 or G5 systems, equating to 32% of the respondents, compared with only 2% of respondents with type 1 diabetes who were using a Dexcom G4 or G5. The faster migration to the Dexcom G6 system may in part represent a stronger “push factor” to transition to newer systems among people with type 1 diabetes attending more advanced diabetes clinics, although it also likely reflects the more complex clinical needs of people with type 1 diabetes. This complexity of care is reinforced by the far higher usage of insulin pumps in people with type 1 diabetes compared with those with type 2 diabetes (81 vs. 35%). It is equally true that more people with type 2 diabetes than with type 1 diabetes used FreeStyle Libre systems (35 vs. 10%). Technology and chemistry differences aside, the differences in performance (e.g., accuracy or reliability) or usability are marginal when comparing FreeStyle Libre systems to Dexcom systems (29) and, in our view, unlikely to fully explain why people with type 2 diabetes reacted more negatively to CGM on certain questions. Furthermore, the small number of respondents with type 1 diabetes who were using FreeStyle Libre systems makes speculation and comparison with Dexcom systems less reliable.

More than 50% of respondents experienced skin reactions, and the majority felt pain when applying a sensor to the skin. The older Dexcom G4 and G5 systems performed less well than the more recent G6 system. The better performance of the G6 is welcome and comes despite the fact that G6 sensor remains resident on the skin for 3 days longer than the G4 or G5 sensors. It is understood that manufacturing process changes, including the elimination of ethyl cyanoacrylate, minimized the skin reactivity of the G6 sensor compared with adhesive configurations used for the G4 and G5 sensors (30). Despite reports that adhesive ingredients used in the FreeStyle Libre systems contain agents that provoke skin reactions (31,32), we observed a lower reported incidence of skin issues despite the Libre sensors having a wear time of 14 days. Unfortunately, because of constraints on survey length, we did not collect feedback on the onset or severity of skin issues.

Our survey focused on several aspects of sensor accuracy and how this fundamental attribute affects patient behavior, decision-making, and trust in CGM systems. Respondents reported suboptimal performance on the first or last day of an individual sensor and also reported encountering underperforming sensors that were inaccurate over their full wear time. A recent article (33) advocates for this issue to be presented more transparently in future studies. This issue is pertinent given that even data for the Dexcom G7 show that >25% of sensors did not achieve a MARD of <10%, a recognized goal for safe insulin dosing (6,7). Our data showed that people were more satisfied with the accuracy of the Dexcom G6 system than the older G4 and G5 systems. Dissatisfaction with sensor accuracy was cited as the top reason why people with diabetes temporarily stopped using CGM, closely followed by sensors falling off prematurely. Notably, more than one-third of respondents regularly performed BGM, with a lack of trust in CGM accuracy the top reason prompting BGM. Poor accuracy compromises the reliability of threshold alarms or alerts, and this issue manifested itself throughout the survey, including as a reason why people with diabetes stopped using CGM, and negatively affected daily life and diabetes management for more than one-third of respondents.

We performed a subanalysis (Supplementary Material—Data) to understand how respondents’ age might have influenced sensor satisfaction rates. We found that respondents >55 years of age were more satisfied with sensor performance than those <39 or 40–54 years of age in terms of enabling dosing or management decisions. Those in the older group were also less concerned about the impact of sensor accuracy when taking their medications. This same trend and finding persisted when we analyzed the data for either people with type 1 diabetes or people with type 2 diabetes who were >55 years of age.

A further subanalysis (Supplementary Materials—Data) evaluated the impact of diabetes duration on sensor satisfaction rates. Unsurprisingly, given the data on satisfaction levels in younger versus older people, we found that people who had been living with diabetes longer were more satisfied with sensor performance. A clear distinction was evident, with those who had a diabetes duration ≤10 years being far less satisfied and more concerned with sensor performance than those with diabetes ≥10 years.

The medication burden was significant for all respondents, with 75% taking more than three medications per day and 40% of those with type 2 diabetes and 25% of those with type 1 diabetes taking more than six medications per day. Given this fact, it is not surprising that twice as many respondents with type 2 diabetes as those with type 1 diabetes were concerned about how medications might affect sensor accuracy.

OTC and prescription medications were of particular concern to all respondents, with pain relief, cold and flu remedies, and vitamin C tablets specifically identified as potential interferents. Manufacturers have widely publicized the impact of taking a number of substances (e.g., acetaminophen, salicylic acid, and ascorbic acid) on sensor accuracy; therefore, CGM users are rightly cautious about decision-making based on CGM when taking these substances.

Given the publicity and abundant literature on substances that interfere with CGM accuracy (2227), it was surprising that dehydration was selected by respondents (especially those with type 1 diabetes) as having the most impact on accuracy. Published evidence on this topic is limited; in fact, we have been unable to find a specific study describing how hydration levels affect CGM accuracy. However, the safety information on Abbott Diabetes Care’s FreeStyle Libre website says, “Severe dehydration and excessive water loss may cause inaccurate sensor glucose readings. If you believe you are suffering from dehydration, consult your healthcare professional immediately,” although no published references or data on file are cited (19). Common causes of severe dehydration such as fever, vomiting, and/or diarrhea may be accompanied by people taking various OTC products and prescription medications. Thus, it is possible that, under these circumstances, sensor readings may be less reliable than normal.

A recent meta-analysis (11) found that CGM accuracy is negatively affected by exercise. The analysis also pointed out that the vast majority of exercise regimens mandated in clinical studies are relatively light and of short duration and, therefore, unlikely to elicit significant dehydration or rate of change in glucose. By contrast, another review (34) postulated that longer-duration exercise and the potential for dehydration could decrease the glucose supply within the interstitium, resulting in glucose levels that actually are lower when measured in the interstitial compartment compared with either capillary or venous measurements. An early study examined the differential effects of fasting and significant short-term dehydration (4.1% of body weight) on hyperglycemia induced by withdrawal of insulin in people with type 1 diabetes. Under such conditions, dehydration elicited hyperglycemia, and the mechanism appeared to be related to increased glucose production in the dehydrated state (35). A further limitation of our survey, therefore, is that we did not provide the option of exercise when we asked about substances or situations respondents felt might affect CGM accuracy, leading to the possibility that dehydration was selected as a proxy for exercise.

CGM is a game-changing technology and has evolved in the past decade to overcome many technical and usability obstacles. Our survey suggests that there remain areas for further improvement, especially concerning sensor-to-sensor accuracy, performance across sensors’ full wear time, and continuing issues with inaccurate alerts and alarms. Mistrust in CGM performance was more common than expected, often leading to users confirming readings with BGM, affecting insulin dosing and diabetes management decisions, and raising concerns about taking common OTC or prescription medications. Skin issues remain a problem for many users. Some older systems fared less well, which is understandable. In general, people with type 2 diabetes were less satisfied than those with type 1 diabetes with their CGM system’s performance.

Acknowledgments

The authors thank INL Agency for creating the infographics, tables, and charts. They also thank Dr. Linda M. Gaudiani, an employee of LifeScan at the time of writing, for her thoughts on the initial manuscript.

Funding

Funding for this study and preparation of the manuscript were provided by LifeScan.

Duality of Interest

E.H. and M.G. are current employees of LifeScan. H.N. and K.C. are current employees of the T1D Exchange. J.B. and J.L. are past employees of the T1D Exchange. No other potential conflicts of interest relevant to this article were reported.

Author Contributions

E.H. and M.G. co-created the survey with the T1D Exchange, reviewed and interpreted the survey data, and wrote the manuscript. J.B. and K.C. executed the survey, provided datasets, and reviewed the final manuscript. H.N. and J.L. analyzed and interpreted the survey data and reviewed the manuscript. E.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

This article contains supplementary material online at https://doi.org/10.2337/figshare.23866626.

1.
Aggarwal
A
,
Pathak
S
,
Goyal
R
.
Clinical and economic outcomes of continuous glucose monitoring system (CGMS) in patients with diabetes mellitus: a systematic literature review
.
Diabetes Res Clin Pract
2022
;
186
:
109825
2.
Evans
M
,
Welsh
Z
,
Seibold
A
.
Reductions in HbA1c with flash glucose monitoring are sustained for up to 24 months: a meta-analysis of 75 real-world observational studies
.
Diabetes Ther
2022
;
13
:
1175
1185
3.
Fang
M
,
Wang
D
,
Coresh
J
,
Selvin
E
.
Trends in diabetes treatment and control in U.S. adults, 1999–2018
.
N Engl J Med
2021
;
384
:
2219
2228
4.
Foster
NC
,
Beck
RW
,
Miller
KM
, et al
.
State of type 1 diabetes management and outcomes from the T1D Exchange in 2016–2018
.
Diabetes Technol Ther
2019
;
21
:
66
72
5.
Denham
D
.
A head-to-head comparison study of the first-day performance of two factory-calibrated CGM systems
.
J Diabetes Sci Technol
2020
;
14
:
493
495
6.
Garg
SK
,
Kipnes
M
,
Castorino
K
, et al
.
Accuracy and safety of Dexcom G7 continuous glucose monitoring in adults with diabetes
.
Diabetes Technol Ther
2022
;
24
:
373
380
7.
Laffel
LM
,
Bailey
TS
,
Christiansen
MP
,
Reid
JL
,
Beck
SE
.
Accuracy of a seventh-generation continuous glucose monitoring system in children and adolescents with type 1 diabetes
.
J Diabetes Sci Technol
2023
;
17
:
962
967
8.
Bailey
T
,
Bode
BW
,
Christiansen
MP
,
Klaff
LJ
,
Alva
S
.
The performance and usability of a factory-calibrated flash glucose monitoring system
.
Diabetes Technol Ther
2015
;
17
:
787
794
9.
Pleus
S
,
Schoemaker
M
,
Morgenstern
K
, et al
.
Rate-of-change dependence of the performance of two CGM systems during induced glucose swings
.
J Diabetes Sci Technol
2015
;
9
:
801
807
10.
Pleus
S
,
Stuhr
A
,
Link
M
,
Haug
C
,
Freckmann
G
.
Variation of mean absolute relative differences of continuous glucose monitoring systems throughout the day
.
J Diabetes Sci Technol
2022
;
16
:
649
658
11.
Muñoz Fabra
E
,
Díez
J-L
,
Bondia
J
,
Laguna Sanz
AJ
.
A comprehensive review of continuous glucose monitoring accuracy during exercise periods
.
Sensors (Basel)
2021
;
21
:
479
12.
Tanenbaum
ML
,
Adams
RN
,
Hanes
SJ
, et al
.
Optimal use of diabetes devices: clinician perspectives on barriers and adherence to device use
.
J Diabetes Sci Technol
2017
;
11
:
484
492
13.
Pettus
J
,
Price
DA
,
Edelman
SV
.
How patients with type 1 diabetes translate continuous glucose monitoring data into diabetes management decisions
.
Endocr Pract
2015
;
21
:
613
620
14.
Tanenbaum
ML
,
Hanes
SJ
,
Miller
KM
,
Naranjo
D
,
Bensen
R
,
Hood
KK
.
Diabetes device use in adults with type 1 diabetes: barriers to uptake and potential intervention targets
.
Diabetes Care
2017
;
40
:
181
187
15.
Wen
X
,
Ou
T
,
Zeng
N
, et al
.
Clinical interpretation of the flash continuous glucose monitoring data out of range in type 2 diabetes: a real-world study
.
Diabetes Res Clin Pract
2021
;
178
:
108989
16.
Rigo
RS
,
Levin
LE
,
Belsito
DV
,
Garzon
MC
,
Gandica
R
,
Williams
KM
.
Cutaneous reactions to continuous glucose monitoring and continuous subcutaneous insulin infusion devices in type 1 diabetes mellitus
.
J Diabetes Sci Technol
2021
;
15
:
786
791
17.
Lombardo
F
,
Passanisi
S
,
Tinti
D
,
Messina
MF
,
Salzano
G
,
Rabbone
I
.
High frequency of dermatological complications in children and adolescents with type 1 diabetes: a web-based survey
.
J Diabetes Sci Technol
2021
;
15
:
1377
1381
18.
Denham
D
.
Effect of repeated doses of acetaminophen on a continuous glucose monitoring system with permselective membrane
.
J Diabetes Sci Technol
2021
;
15
:
517
518
19.
Abbott Diabetes Care
.
Full indication and important safety information: Freestyle Libre 14 Day
.
Available from https://www.freestyle.abbott/us-en/safety-information.html. Accessed November 2022
20.
Bailey
TS
,
Alva
S
.
Landscape of continuous glucose monitoring (CGM) and integrated CGM: accuracy considerations
.
Diabetes Technol Ther
2021
;
23
(
Suppl. 3
):
S5
S11
21.
Galindo
RJ
,
Umpierrez
GE
,
Rushakoff
RJ
, et al
.
Continuous glucose monitors and automated insulin dosing systems in the hospital consensus guideline
.
J Diabetes Sci Technol
2020
;
14
:
1035
1064
22.
Basu
A
,
Slama
MQ
,
Nicholson
WT
, et al
.
Continuous glucose monitor interference with commonly prescribed medications: a pilot study
.
J Diabetes Sci Technol
2017
;
11
:
936
941
23.
Moscardó
V
,
Garcia
A
,
Bondia
J
,
Diaz
J
,
Ramos-Prol
A
,
Rossetti
P
.
Effect of ethanol consumption on the accuracy of a glucose oxidase-based subcutaneous glucose sensor in subjects with type 1 diabetes
.
Sensors (Basel)
2022
;
22
:
3101
24.
Boehm
R
,
Donovan
J
,
Sheth
D
,
Durfor
A
,
Roberts
J
,
Isayeva
I
.
In vitro sugar interference testing with amperometric glucose oxidase sensors
.
J Diabetes Sci Technol
2019
;
13
:
82
95
25.
Tellez
SE
,
Hornung
LN
,
Courter
JD
, et al
.
Inaccurate glucose sensor values after hydroxyurea administration
.
Diabetes Technol Ther
2021
;
23
:
443
451
26.
Heinemann
L
.
Interferences with CGM systems: practical relevance?
J Diabetes Sci Technol
2022
;
16
:
271
274
27.
Pfützner
A
,
Jensch
H
,
Cardinal
C
,
Srikanthamoorthy
G
,
Riehn
E
,
Thomé
N
.
Laboratory protocol and pilot results for dynamic interference testing of continuous glucose monitoring sensors
.
J Diabetes Sci Technol
.
Online ahead of print on 13 May 2022 (doi: https://doi.org/10.1177/19322968221095573)
28.
Lombardo
F
,
Passanisi
S
,
Caminiti
L
, et al
.
High prevalence of skin reactions among pediatric patients with type 1 diabetes using new technologies: the alarming role of colophonium
.
Diabetes Technol Ther
2020
;
22
:
53
56
29.
Lee
I
,
Probst
D
,
Klonoff
D
,
Sode
K
.
Continuous glucose monitoring systems: current status and future perspectives of the flagship technologies in biosensor research
.
Biosens Bioelectron
2021
;
181
:
113054
30.
Gisin
V
,
Chan
A
,
Welsh
JB
.
Manufacturing process changes and reduced skin irritations of an adhesive patch used for continuous glucose monitoring devices
.
J Diabetes Sci Technol
2018
;
12
:
725
726
31.
Mowitz
M
,
Herman
A
,
Baeck
M
, et al
.
N,N-dimethylacrylamide: a new sensitizer in the FreeStyle Libre glucose sensor
.
Contact Dermat
2019
;
81
:
27
31
32.
Herman
A
,
Mowitz
M
,
Aerts
O
, et al
.
Unexpected positive patch test reactions to sesquiterpene lactones in patients sensitized to the glucose sensor FreeStyle Libre
.
Contact Dermat
2019
;
81
:
354
367
33.
Eichenlaub
M
,
Stephan
P
,
Waldenmaier
D
, et al
.
Continuous glucose deviation interval and variability analysis (CG-DIVA): a novel approach for the statistical accuracy assessment of continuous glucose monitoring systems
.
J Diabetes Sci Technol
.
Online ahead of print on 3 November 2022 (doi: https://doi.org/10.1177/19322968221134639)
34.
Moser
O
,
Yardley
JE
,
Bracken
RM
.
Interstitial glucose and physical exercise in type 1 diabetes: integrative physiology, technology, and the gap in-between
.
Nutrients
2018
;
10
:
93
35.
Burge
MR
,
Garcia
N
,
Qualls
CR
,
Schade
DS
.
Differential effects of fasting and dehydration in the pathogenesis of diabetic ketoacidosis
.
Metabolism
2001
;
50
:
171
177
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.