DIABETES TECHNOLOGY: STANDARDS OFMEDICALCAREINDIABETESD2021 - AMERICAN DIABETES ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Diabetes Care Volume 44, Supplement 1, January 2021 S85
7. Diabetes Technology: Standards American Diabetes Association
of Medical Care in Diabetesd2021
Diabetes Care 2021;44(Suppl. 1):S85–S99 | https://doi.org/10.2337/dc21-S007
The American Diabetes Association (ADA) “Standards of Medical Care in Diabetes”
includes the ADA’s current clinical practice recommendations and is intended to
provide the components of diabetes care, general treatment goals and guidelines,
and tools to evaluate quality of care. Members of the ADA Professional Practice
Committee, a multidisciplinary expert committee (https://doi.org/10.2337/dc21-
7. DIABETES TECHNOLOGY
SPPC), are responsible for updating the Standards of Care annually, or more
frequently as warranted. For a detailed description of ADA standards, statements,
and reports, as well as the evidence-grading system for ADA’s clinical practice
recommendations, please refer to the Standards of Care Introduction (https://doi
.org/10.2337/dc21-SINT). Readers who wish to comment on the Standards of Care
are invited to do so at professional.diabetes.org/SOC.
Diabetes technology is the term used to describe the hardware, devices, and software
that people with diabetes use to help manage their condition, from lifestyle to blood
glucose levels. Historically, diabetes technology has been divided into two main
categories: insulin administered by syringe, pen, or pump, and blood glucose
monitoring as assessed by meter or continuous glucose monitor. More recently,
diabetes technology has expanded to include hybrid devices that both monitor
glucose and deliver insulin, some automatically, as well as software that serves as a
medical device, providing diabetes self-management support. Diabetes technology,
when coupled with education and follow-up, can improve the lives and health of
people with diabetes; however, the complexity and rapid change of the diabetes
technology landscape can also be a barrier to patient and provider implementation.
Recommendation
7.1 Use of technology should be individualized based on a patient’s needs, desires,
skill level, and availability of devices. E
Technology is rapidly changing, but there is no “one-size-fits-all” approach to
technology use in people with diabetes. Insurance coverage can lag behind device
availability, patient interest in devices and willingness to change can vary, and
providers may have trouble keeping up with newly released technology. Not-for-profit
websites can help providers and patients make decisions as to the initial choice of
devices. Other sources, including health care providers and device manufacturers, can Suggested citation: American Diabetes Associa-
help people troubleshoot when difficulties arise. tion. 7. Diabetes technology: Standards of Medical
Care in Diabetesd2021. Diabetes Care 2021;
44(Suppl. 1):S85–S99
SELF-MONITORING OF BLOOD GLUCOSE © 2020 by the American Diabetes Association.
Readers may use this article as long as the work is
Recommendations properly cited, the use is educational and not for
7.2 People who are on insulin using self-monitoring of blood glucose should be profit, and the work is not altered. More infor-
encouraged to test when appropriate based on their insulin regimen. This may mation is available at https://www.diabetesjournals
.org/content/license.S86 Diabetes Technology Diabetes Care Volume 44, Supplement 1, January 2021
therapy of patients taking insulin. In recent Counterfeit Strips
include testing when fasting, prior Patients should be advised against pur-
years, continuous glucose monitoring (CGM)
to meals and snacks, at bedtime, chasing or reselling preowned or second-
has emerged as a method for the assessment
prior to exercise, when low blood hand test strips, as these may give incorrect
of glucose levels (discussed below). Glucose
glucose is suspected, after treat- results. Only unopened and unexpired
monitoring allows patients to evaluate their
ing low blood glucose until they vials of glucose test strips should be used
individual response to therapy and assess
are normoglycemic, and prior to ensure SMBG accuracy.
whether glycemic targets are being safely
to and while performing critical
achieved. Integrating results into diabetes
tasks such as driving. B
management can be a useful tool for guiding Optimizing SMBG Monitor Use
7.3 Providers should be aware of the
medical nutrition therapy and physical ac- SMBG accuracy is dependent on the in-
differences in accuracy among glu-
tivity, preventing hypoglycemia, or adjusting strument and user, so it is important to
cose metersdonly U.S. Food and
medications (particularly prandial insulin evaluate each patient’s monitoring tech-
Drug Administration–approved me-
doses). The patient’s specific needs and nique, both initially and at regular intervals
ters with proven accuracy should
goals should dictate SMBG frequency and thereafter. Optimal use of SMBG requires
be used, with unexpired strips,
timing or the consideration of CGM use. proper review and interpretation of the
purchased from a pharmacy or
data, by both the patient and the provider,
licensed distributor. E
Meter Standards to ensure that data are used in an effective
7.4 When prescribed as part of a
Glucose meters meeting U.S. Food and and timely manner. In patients with type 1
diabetes self-management edu-
Drug Administration (FDA) guidance for diabetes, there is a correlation between
cation and support program, self-
meter accuracy provide the most reliable greater SMBG frequency and lower A1C
monitoring of blood glucose may
data for diabetes management. There (7). Among patients who check their blood
help to guide treatment decisions
are several current standards for accu- glucose at least once daily, many report
and/or self-management for pa-
racy of blood glucose monitors, but the taking no action when results are high or
tients taking less frequent insulin
two most used are those of the Inter- low (8). Patients should be taught how to
injections. B
national Organization for Standardization use SMBG data to adjust food intake,
7.5 Although self-monitoring of blood
(ISO) (ISO 15197:2013) and the FDA. The exercise, or pharmacologic therapy to
glucose in patients on noninsulin
current ISO and FDA standards are com- achieve specific goals. Some meters now
therapies has not consistently
pared in Table 7.1. In Europe, currently provide advice to the user in real time,
shown clinically significant reduc-
marketed monitors must meet current ISO when monitoring glucose levels (9), while
tions in A1C, it may be helpful
standards. In the U.S., currently marketed others can be used as a part of integrated
when altering diet, physical ac-
monitors must meet the standard under health platforms (10).
tivity, and/or medications (par-
which they were approved, which may not The ongoing need for and frequency of
ticularly medications that can
be the current standard. Moreover, the SMBG should be reevaluated at each rou-
cause hypoglycemia) in conjunc-
monitoring of current accuracy is left to the tine visit to avoid overuse, particularly if
tion with a treatment adjustment
manufacturer and not routinely checked SMBG is not being used effectively for
program. E
by an independent source. self-management (8,11,12).
7.6 When prescribing self-monitoring
Patients assume their glucose monitor
of blood glucose, ensure that
is accurate because it is FDA cleared, but Patients on Intensive Insulin Regimens
patients receive ongoing instruc-
often that is not the case. There is sub- SMBG is especially important for insulin-
tion and regular evaluation of
stantial variation in the accuracy of widely treated patients to monitor for and pre-
technique, results, and their abil-
used blood glucose monitoring systems vent hypoglycemia and hyperglycemia.
ity to use data, including upload-
(2,3). The Diabetes Technology Society Most patients using intensive insulin regi-
ing/sharing data (if applicable),
Blood Glucose Monitoring System Sur- mens (multiple daily injections or insulin
from self-monitoring of blood glu-
veillance Program provides information pump therapy) should be encouraged to
cose devices to adjust therapy. E
on the performance of devices used for assess glucose levels using SMBG (and/or
7.7 Health care providers should be
SMBG (https://diabetestechnology.org/ CGM) prior to meals and snacks, at bed-
aware of medications and other
surveillance). In one analysis, only 6 of the time, occasionally postprandially, prior to
factors, such as high-dose vita-
top 18 glucose meters met the accuracy exercise, when they suspect low blood
min C and hypoxemia, that can
standard (4). glucose, after treating low blood glucose
interfere with glucose meter ac-
There are single-meter studies in which until they are normoglycemic, and prior
curacy and provide clinical man-
benefits have been found with individual to and while performing critical tasks
agement as indicated. E
meter systems, but few that compare such as driving. For many patients using
meters in a head-to-head manner. Cer- SMBG, this will require checking up to
Major clinical trials of insulin-treated pa- tain meter system characteristics, such as 6–10 times daily, although individual
tients have included self-monitoring of the use of lancing devices that are less needs may vary. A database study of
blood glucose (SMBG) as part of multi- painful (5) and the ability to reapply blood almost 27,000 children and adolescents
factorial interventions to demonstrate to a strip with an insufficient initial sample, with type 1 diabetes showed that, after
the benefit of intensive glycemic con- may also be beneficial to patients (6) and adjustment for multiple confounders, in-
trol on diabetes complications (1). SMBG may make SMBG less burdensome for creased daily frequency of SMBG was
is thus an integral component of effective patients to perform. significantly associated with lower A1Ccare.diabetesjournals.org Diabetes Technology S87
Table 7.1—Comparison of ISO 15197:2013 and FDA blood glucose meter accuracy standards
Setting FDA (206,207) ISO 15197:2013 (208)
Home use 95% within 15% for all BG in the usable BG range† 95% within 15% for BG $100 mg/dL
99% within 20% for all BG in the usable BG range† 95% within 15 mg/dL for BG ,100 mg/dL
99% in A or B region of consensus error grid‡
Hospital use 95% within 12% for BG $75 mg/dL
95% within 12 mg/dL for BG ,75 mg/dL
98% within 15% for BG $75 mg/dL
98% within 15 mg/dL for BG ,75 mg/dL
BG, blood glucose; FDA, U.S. Food and Drug Administration; ISO, International Organization for Standardization. To convert mg/dL to mmol/L,
see http://endmemo.com/medical/unitconvert/Glucose.php. †The range of blood glucose values for which the meter has been proven accurate
and will provide readings (other than low, high, or error). ‡Values outside of the “clinically acceptable” A and B regions are considered “outlier”
readings and may be dangerous to use for therapeutic decisions (209).
(20.2% per additional check per day) and SMBG can reduce A1C by 0.25–0.3% reading and a message indicating that the
with fewer acute complications (13). at 6 months (21–23), but the effect was value may be incorrect.
attenuated at 12 months in one analysis (21). Interfering Substances. There are a few
Patients Using Basal Insulin and/or Oral
Agents Reductions in A1C were greater (20.3%) in physiologic and pharmacologic factors
The evidence is insufficient regarding trialswherestructuredSMBGdatawereused that interfere with glucose readings. Most
when to prescribe SMBG and how often to adjust medications, but A1C was not interfere only with glucose oxidase sys-
monitoring is needed for insulin-treated changedsignificantlywithoutsuchstructured tems (25). They are listed in Table 7.2.
patients who do not use intensive insulin diabetes therapy adjustment (23). A key
regimens, such as those with type 2 di- consideration is that performing SMBG alone Table 7.2—Interfering substances for
does not lower blood glucose levels. To be glucose readings
abetes using basal insulin with or without
useful, the information must be integrated Glucose oxidase monitors
oral agents. However, for patients using
into clinical and self-management plans. Uric acid
basal insulin, assessing fasting glucose Galactose
with SMBG to inform dose adjustments Glucose Meter Inaccuracy Xylose
to achieve blood glucose targets results Although many meters function well Acetaminophen
in lower A1C (14,15). under a variety of circumstances, providers L-DOPA
In people with type 2 diabetes not and people with diabetes need to be aware Ascorbic acid
using insulin, routine glucose monitoring of factors that can impair meter accuracy. A Glucose dehydrogenase monitors
may be of limited additional clinical Icodextrin (used in peritoneal dialysis)
meter reading that seems discordant with
benefit. By itself, even when combined clinical reality needs to be retested or
with education, it has showed limited tested in a laboratory. Providers in inten- CONTINUOUS GLUCOSE
improvement in outcomes (16–19). How- sive care unit settings need to be partic- MONITORING DEVICES
ever, for some individuals, glucose mon- ularly aware of the potential for abnormal
itoring can provide insight into the See Table 7.3 for definitions of types of
meter readings, and laboratory-based val-
impact of diet, physical activity, and CGM devices.
ues should be used if there is any doubt.
medication management on glucose Some meters give error messages if meter Recommendations
levels. Glucose monitoring may also be readings are likely to be false (24). 7.8 When prescribing continuous glu-
useful in assessing hypoglycemia, glu- Oxygen. Currently available glucose
cose monitoring (CGM) devices,
cose levels during intercurrent illness, monitors utilize an enzymatic reaction robust diabetes education, train-
or discrepancies between measured linked to an electrochemical reaction, ei- ing, and support are required for
A1C and glucose levels when there is ther glucose oxidase or glucose dehydro- optimal CGM device implementa-
concern an A1C result may not be reliable genase (25). Glucose oxidase monitors tion and ongoing use. People using
in specific individuals. It may be useful are sensitive to the oxygen available and CGM devices need to have the
when coupled with a treatment adjust- should only be used with capillary blood in ability to perform self-monitoring
ment program. In a year-long study of patients with normal oxygen saturation. of blood glucose in order to
insulin-naive patients with suboptimal ini- Higher oxygen tensions (i.e., arterial blood calibrate their monitor and/or
tial glycemic stability, a group trained in or oxygen therapy) may result in false low verify readings if discordant from
structured SMBG (a paper tool was used at glucose readings, and low oxygen tensions
their symptoms. B
least quarterly to collect and interpret (i.e., high altitude, hypoxia, or venous blood
7.9 When used properly, real-time
seven-point SMBG profiles taken on 3 con- readings) may lead to false high glucose
continuous glucose monitors in
secutive days) reduced their A1C by 0.3% readings. Glucose dehydrogenase–based
conjunction with multiple daily
more than the control group (20). A trial of monitors are not sensitive to oxygen.
injections and continuous subcu-
once-daily SMBG that included en- Temperature. Because the reaction is sen-
taneous insulin infusion A and
hanced patient feedback through mes- sitive to temperature, all monitors have
other forms of insulin therapy C
saging found no clinically or statistically an acceptable temperature range (25).
are a useful tool to lower and/or
significant change in A1C at 1 year (19). Most will show an error if the temperature
maintain A1C levels and/or reduce
Meta-analyses have suggested that is unacceptable, but a few will provide aS88 Diabetes Technology Diabetes Care Volume 44, Supplement 1, January 2021
CGM measures interstitial glucose (which added benefit. This device (FreeStyle
hypoglycemia in adults and youth
correlates well with plasma glucose, al- Libre 2) and one rtCGM (Dexcom G6)
with diabetes.
though at times can lag if glucose levels are have both been designated as integrated
7.10 When used properly, intermit-
rising or falling rapidly). There are two continuous glucose monitoring (iCGM)
tentlyscannedcontinuousglucose
basic types of CGM devices: those that devices (https://www.accessdata.fda.gov/
monitors in conjunction with mul-
are owned by the user, unblinded, and scripts/cdrh/cfdocs/cfpcd/classification
tiple daily injections and continu-
intended for frequent/continuous use .cfm?id5682). This is a higher standard,
ous subcutaneous insulin infusion
(real-time [rt]CGM and intermittently set by the FDA, so these devices can be
B and other forms of insulin ther-
scanned [is]CGM) and those that are reliably integrated with other digitally con-
apy C can be useful and may lower owned and applied in/by the clinic, which nected devices, including automated in-
A1C levels and/or reduce hypo- provide data that is blinded or unblinded sulin dosing systems.
glycemia in adults and youth with for a discrete period of time (professional Some real-time systems require cali-
diabetes to replace self-monitor- CGM). Table 7.3 provides the definitions bration by the user, which varies in fre-
ing of blood glucose. for the types of CGM devices. For devices quencydependingonthedevice.Additionally,
7.11 In patients on multiple daily that provide patients unblinded data, for some CGM systems, the FDA suggests
injections and continuous subcu- most of the published randomized con- SMBG for making treatment decisions.
taneous insulin infusion, real-time trolled trials (RCTs) have been performed Devices that require SMBG confirmation
continuous glucose monitoring using rtCGM devices that have alarms are called “adjunctive,” while those that
(CGM) devices should be used as and alerts. The RCT results have largely do not are called “nonadjunctive.” An
close to daily as possible for been positive, in terms of reducing either RCT of 226 adults suggested that a CGM
maximal benefit. A Intermittently A1C levels and/or episodes of hypogly- device could be used safely and effec-
scanned CGM devices should be cemia, as long as participants regularly tively without regular confirmatory SMBG
scanned frequently, at a minimum wear the devices (26–29). These devices in patients with well-controlled type 1
once every 8 h. provide glucose readings continuously diabetes at low risk of severe hypoglyce-
7.12 When used as an adjunct to pre- to a smartphone or reader that can be mia (33). Two CGM devices are approved
and postprandial self-monitor- viewed by the patient and/or a care- by the FDA for making treatment deci-
ing of blood glucose, continu- giver. It is difficult to determine how sions without SMBG calibration or con-
ous glucose monitoring can help much the need to swipe a device to firmation (34,35). For patients with
to achieve A1C targets in diabetes obtain a result, combined with a lack of type 1 diabetes using rtCGM, an impor-
and pregnancy. B alarms and alerts, matters in terms of tant predictor of A1C lowering for all age-
7.13 Use of professional continuous outcomes, although results from these groups was frequency of sensor use (26). In
glucose monitoring (CGM) and/or devices (isCGM) have not shown con- this study, overall use was highest in those
intermittent real-time or intermit- sistent improvements in glycemic out- aged $25 years (who had the most im-
tentlyscannedCGMcanbehelpful comes (30). However, data from longitudinal provement in A1C) and lower in younger
in identifying and correcting pat- trials (without a control group for com- age-groups.
terns of hyper- and hypoglycemia parison) show improvement in A1C levels The abundance of data provided by
and improving A1C levels in peo- (31). There is one small study in patients CGM offers opportunities to analyze
ple with diabetes on noninsulin as at risk for hypoglycemia that compared patient data more granularly than was
well as basal insulin regimens. C rtCGM with isCGM (32). The study showed previously possible, providing additional
7.14 Skin reactions, either due to irri- improvement in time spent in hypoglyce- information to aid in achieving glycemic
tation or allergy, should be as- mia with rtCGM compared with isCGM. targets. A variety of metrics have been
The newest version of the isCGM system proposed (27) and are discussed in Sec-
sessed and addressed to aid in
has an optional alert for a high or low tion 6 “Glycemic Targets” (https://doi
successful use of devices. E
glucose value (without the capacity for .org/10.2337/dc21 -S006). CGM is es-
7.15 People who have been using
providing predictive alerts), but it still sential for creating the ambulatory glu-
continuous glucose monitors
requires that the device be swiped to cose profile (AGP) and providing data on
should have continued access
reveal the glucose level and trend arrows, time in range, percentage of time spent
across third-party payers. E
and RCT data are lacking in terms of above and below range, and variability
Table 7.3—Continuous glucose monitoring (CGM) devices
Type of CGM Description
Real-time CGM (rtCGM) CGM systems that measure and display glucose levels continuously
Intermittently scanned CGM (isCGM) CGM systems that measure glucose levels continuously but only display glucose values when swiped
by a reader or a smartphone
Professional CGM CGM devices that are placed on the patient in the provider’s office (or with remote instruction) and worn
for a discrete period of time (generally 7–14 days). Data may be blinded or visible to the person
wearing the device. The data are used to assess glycemic patterns and trends. These devices are not
fully owned by the patientdthey are a clinic-based device, as opposed to the patient-owned rtCGM/
isCGM devices.care.diabetesjournals.org Diabetes Technology S89
(36). Access to CGM devices should be benefit of rtCGM in patients on MDI, in glycemic control following 6 months of
considered from the outset of the di- there were significant reductions in A1C: rtCGM use (60). However, observational
agnosis of diabetes that requires insulin 20.6% in one (28,43) and 20.43% in the feasibility studies of toddlers demonstrated
management (37,38). This allows for other (29). No reduction in A1C was seen a high degree of parental satisfaction and
close tracking of glucose levels with in a small study performed in under- sustained use of the devices despite the
adjustments of insulin dosing and life- served, less well-educated adults with inability to change the degree of glycemic
style modifications and removes the type 1 diabetes (44). In the adult subset control attained (63).
burden of frequent SMBG monitoring. of the JDRF CGM study, there was a Registry data have also shown an
Interruption of access to CGM is asso- significant reduction in A1C of 20.53% association between rtCGM use and
ciated with a worsening of outcomes (55) in patients who were primarily trea- lower A1C levels (55,64), even when
(39); therefore, it is important for indi- ted with insulin pump therapy. Better limiting assessment of rtCGM use to
viduals on CGM to have consistent access adherence in wearing the rtCGM device participants on injection therapy (64).
to the devices. resulted in a greater likelihood of an Impact on HypoglycemiadChildren
improvement in glycemic control (26,45). There are no studies solely including
Education and Training
Primary Outcome: HypoglycemiadAdults pediatric patients that assess rates of
In general, no device used in diabetes
In studies in adults where reduction in hypoglycemia as the primary outcome.
management works optimally without
education, training, and follow-up. De- episodes of hypoglycemia was the pri- Some of the studies where pediatric and
mary end point, significant reductions adult patients were combined together
vice companies offer online tutorials and
were seen in individuals with type 1 did show potential reductions in hypo-
training videos as well as written material
diabetes on MDI or CSII (46–48). In glycemia (16,65,66).
on their use. Patients vary in terms of
one study in patients who were at higher
comfort level with technology, and some
risk for episodes of hypoglycemia (48),
prefer in-person training and support. Real-time CGM Use in Type 2 Diabetes
there was a reduction in rates of all levels
Programs that involve training and Studies in people with type 2 diabetes are
of hypoglycemia (see Section 6 “Glycemic
support have been shown to improve heterogeneous in design: in two, partic-
Targets,” https://doi.org/10.2337/dc21-
outcomes in both adults and children ipants were using basal insulin with oral
S006, for hypoglycemia definitions). rtCGM
using isCGM (40–42). Individuals using agents or oral agents alone (67,68); in
may be particularly useful in insulin-
CGM should also be trained on how to one, individuals were on MDI alone (69).
treated patients with hypoglycemia
use SMBG, for use with devices that re-
unawareness and/or frequent hypogly- The findings in studies with MDI alone
quire calibration, for testing if CGM values (69) and in two studies in people using
cemic episodes, although studies have
seem incongruent with the patient’s sense oral agents with or without insulin
not been powered to show consistent
of their glucose levels, and if the CGM (67,68) showed significant reductions
reductions in severe (level 3) hypogly-
device fails or is not available. in A1C levels. The Multiple Daily Injec-
cemia (26,49,50).
tions and Continuous Glucose Monitor-
Real-time CGM Device Use in Adults Impact on Glycemic ControldChildren ing in Diabetes (DIAMOND) study in
and Children With Diabetes When data from adult and pediatric people with type 2 diabetes on MDI
Data exist to support the use of real-time participants are analyzed together, showed a reduction in A1C but no re-
CGM in adults and children, both those rtCGM use in RCTs has been associated duction in hypoglycemia (69). Studies in
on multiple daily injections (MDI) and with reduction in A1C levels (49–51). Yet, individuals with type 2 diabetes on oral
those on continuous subcutaneous in- in the JDRF CGM trial, when youth were agents with or without insulin did not
sulin infusion (CSII). This is true in studies analyzed by age-group (8- to 14-year-olds show reductions in rates of hypoglycemia
both in people with type 1 diabetes and and 15- to 24-year-olds), no change in (67,68).
those with type 2 diabetes, although data A1C was seen, likely due to poor rtCGM
in individuals with type 2 diabetes is adherence (26). Indeed, in a secondary Intermittently Scanned CGM Device
primarily in adults. analysis of that RCT’s data in both pedi- Use in Adults and Children With
In terms of RCTs in people with type 1 atric cohorts, those who used the sensor Diabetes
diabetes, there are four studies in adults $6 days/week had an improvement in their The original isCGM device (to which the
with A1C as the primary outcome glycemic control (56). One critical com- majority of the published data applies)
(28,29,43–45), three studies in adults ponent to success with CGM is near- did not provide alarms and alerts but is an
with hypoglycemia as the primary out- daily wearing of the device (49,55, option used by many patients. There are
come (46–48), four studies in adults 57–59). One RCT showed no improve- relatively few RCT data proving benefit
and children with A1C as the primary ment in glycemic outcomes in children aged in people with diabetes, but there are
outcome (26,49–51), and three studies 4–10 years of age, regardless of how often it multiple longitudinal and observational
in adults and children with hypoglyce- was worn (60). studies. One RCT, designed to show a
mia as a primary outcome (52–54). Though data from small observational reduction in episodes of hypoglycemia in
Primary Outcome: A1C ReductiondAdults studies demonstrate that rtCGM can be patients with type 1 diabetes at higher
In general, A1C reduction was shown in worn by patients ,8 years old and the use risk for hypoglycemia, showed a signif-
studies where the baseline A1C was of rtCGM provides insight to glycemic pat- icant benefit in terms of time spent in a
higher. In two larger studies in adults terns (61,62), an RCT in children aged 4–9 hypoglycemic range (P , 0.0001) (46).
with type 1 diabetes that assessed the years did not demonstrate improvements Another RCT, assessing the ability ofS90 Diabetes Technology Diabetes Care Volume 44, Supplement 1, January 2021
isCGM to prevent episodes of recurrent, program. Another review showed some identify patterns of hypo- and hypergly-
severe hypoglycemia, showed no benefit benefits in terms of A1C reduction as well cemia (93). Professional CGM can be
(70). In one RCT of isCGM in people with as improvement in quality of life (84). A helpful to evaluate patients when either
type 2 diabetes on a variety of insulin review that included studies conducted rtCGM or isCGM is not available to the
regimens and with an initial A1C of using a variety of trial designs, including patient or the patient prefers a blinded
;8.8%, no reduction in A1C was seen; prospective and retrospective cohort stud- analysis or a shorter experience with
however, the time spent in a hypogly- ies, showed overall a reduction in A1C unblinded data. It can be particularly
cemic range was reduced by 43% (71). (20.26%) in people with type 1 and useful to evaluate periods of hypoglyce-
In a study of isCGM in individuals with type 2 diabetes, but there was no differ- mia in patients on agents that can cause
type 2 diabetes on MDI, the A1C was ence in time in range or hypoglycemic hypoglycemia in order to make medica-
reduced by 0.82% in the intervention episodes (83). tion dose adjustments. It can also be
group and 0.33% in the control group Other benefits are discussed in a re- useful to evaluate patients for periods of
(P 5 0.005) with no change in rates of view (82) that supported the use of isCGM hyperglycemia.
hypoglycemia (72). Multiple observa- as a more affordable alternative to rtCGM There are some data showing benefit
tional studies have shown benefit in systems for individuals with diabetes who of intermittent use of CGM (rtCGM
terms of A1C reduction, reductions in are on intensive insulin therapy. In many or isCGM) in individuals with type 2
hypoglycemia, and/or improvements in cases, isCGM is the preferred alternative diabetes on noninsulin and/or basal
quality of life in both children and adults compared with SMBG (85,86). It can also insulin therapies (68,94). In these RCTs,
(31,41,73–78). An observational study improve adherence to monitoring in patients patients with type 2 diabetes not on
from Belgium showed no improvements who are in extremely poor control (87). intensive insulin regimens used CGM
in A1C or quality of life after a year of intermittently compared with patients
isCGM use, with a reduction in episodes Real-time CGM Device Use in randomized to SMBG. Both early (68) and
of severe hypoglycemia and time absent Pregnancy late improvements in A1C were found
from work compared with patient recall One well-designed RCT showed a reduc- (68,94).
of events during the 6 months prior to tion in A1C levels in adult women with Use of professional or intermittent CGM
starting CGM (79). type 1 diabetes on MDI or CSII who were should always be coupled with analysis
There are several published reviews of pregnant using CGM in addition to stan- and interpretation for the patient,
data available on isCGM (80–83). The dard care, including optimization of pre- along with education as needed to
Norwegian Institute of Public Health and postprandial glucose targets (88). It adjust medication and change lifestyle
conducted an assessment of isCGM demonstrated the value of CGM in behaviors.
clinical effectiveness, cost-effectiveness, pregnancy complicated by type 1 di-
and safety for individuals with type 1 abetes by showing a mild improvement Side Effects of CGM Devices
and type 2 diabetes, based on data avail- in A1C without an increase in hypogly- Contact dermatitis (both irritant and
able to January 2017 (80). The authors cemia as well as reductions in large-for- allergic) has been reported with all
concluded that, although there were gestational-age births, length of stay, devices that attach to the skin
few quality data available at the time and neonatal hypoglycemia (88). An (95–97). In some cases this has been
of the report, isCGM may increase treat- observational cohort study that evalu- linked to the presence of isobornyl
ment satisfaction, increase time in range, ated the glycemic variables reported acrylate, which is a skin sensitizer and
and reduce frequency of nocturnal hy- using CGM found that lower mean can cause an additional spreading allergic
poglycemia, without differences in A1C glucose, lower standard deviation, reaction (98–100). Patch testing can be
or quality of life or serious adverse and a higher percentage of time in done to identify the cause of the contact
events. The Canadian Agency for Drugs target range were associated with dermatitis in some cases (101). Identify-
and Technologies in Health reviewed lower risk of large-for-gestational-age ing and eliminating tape allergens is
existing data on isCGM performance births and other adverse neonatal out- important to ensure comfortable use
and accuracy, hypoglycemia, effect on comes (89). Use of the CGM-reported of devices and enhance patient adher-
A1C, and patient satisfaction and quality mean glucose is superior to use of ence (102–105). In some instances, use of
of life and concluded that the system estimated A1C, glucose management an implanted sensor can help avoid skin
could replace SMBG, particularly in pa- indicator, and other calculations to es- reactions in those who are sensitive to
tients who require frequent monitoring timate A1C given the changes to A1C that tape (106,107).
(81). A 2020 systematic review of RCTs occur in pregnancy (90). Two studies
assessing efficacy and patient satisfac- employing intermittent use of rtCGM INSULIN DELIVERY
tion with isCGM revealed improvements showed no difference in neonatal out- Insulin Syringes and Pens
in A1C levels in some subgroups of comes in women with type 1 diabetes
Recommendations
patients (e.g., those with type 2 diabe- (91) or gestational diabetes mellitus (92).
7.16 For people with diabetes who re-
tes) but concluded that additional ben-
quire insulin, insulin syringes or
efit in terms of time in range, glycemic Use of Professional and Intermittent
insulin pens may be used for
variability, and hypoglycemia was un- CGM
insulin delivery with consider-
clear (30). Benefit was enhanced in Professional CGM devices, which provide
ation of patient preference, in-
individuals with type 1 diabetes when retrospective data, either blinded or un-
sulin type and dosing regimen,
combined with a structured education blinded, for analysis, can be used tocare.diabetesjournals.org Diabetes Technology S91
settings with appropriate storage and
cost, and self-management ca- 7.21 Insulin pump therapy may be con-
cleansing (113).
pabilities. B sidered as an option for adults and
Insulin pens offer added convenience
7.17 Insulin pens or insulin injection aids youth with type 2 diabetes and
by combining the vial and syringe into
may be considered for patients otherformsofdiabeteswhoareon
a single device. Insulin pens, allowing
with dexterity issues or vision multiple daily injections who are
push-button injections, come as dispos-
impairment to facilitate the ad- able to safely manage the device. B
able pens with prefilled cartridges or re-
ministration of accurate insulin 7.22 Individuals with diabetes who have
usable insulin pens with replaceable insulin
doses. C been successfully using contin-
cartridges. Pens vary with respect to dosing
7.18 Smartpensmaybeusefulforsome uous subcutaneous insulin infu-
increment and minimal dose, which can
patients to help with dose capture sion should have continued access
range from half-unit doses to 2-unit dose
and dosing recommendations. E across third-party payers. E
increments. U-500 pens come in 5-unit dose
7.19 U.S. Food and Drug Administration–
increments. Some reusable pens include a
approved insulin dose calcula- CSII, or insulin pumps, have been avail-
memory function, which can recall dose
tors/decision support systems able in the U.S. for over 40 years. These
amounts and timing. “Smart” pens that can
may be helpful for titrating insulin devices deliver rapid-acting insulin through-
be programmed to calculate insulin doses
doses. E out the day to help manage blood glucose
and provide downloadable data reports are
levels. Most insulin pumps use tubing to
also available. These pens are useful to assist
Injecting insulin with a syringe or pen is deliver insulin through a cannula, while a
patient insulin dosing in real time as well as
the insulin delivery method used by most few attach directly to the skin, without
for allowing clinicians to retrospectively re-
people with diabetes (108,109), although tubing.
view the insulin doses that were given and
inhaled insulin is also available. Others Most studies comparing MDI with CSII
make insulin dose adjustments (114).
use insulin pumps or automated insulin have been relatively small and of short
Needle thickness (gauge) and length is
delivery devices (see sections on those duration. However, a recent systematic
another consideration. Needle gauges
topics below). For patients with diabetes review and meta-analysis concluded that
range from 22 to 33, with higher gauge
who use insulin, insulin syringes and pens pump therapy has modest advantages
indicating a thinner needle. A thicker
are both able to deliver insulin safely and
needle can give a dose of insulin more for lowering A1C (20.30% [95% CI 20.58
effectively for the achievement of glyce- quickly, while a thinner needle may cause to 20.02]) and for reducing severe hy-
mic targets. When choosing among de- less pain. Needle length ranges from 4 to poglycemia rates in children and adults
livery systems, patient preferences, cost, 12.7 mm, with some evidence suggesting (120). There is no consensus to guide
insulin type and dosing regimen, and self- shorter needles may lower the risk of choosing which form of insulin adminis-
management capabilities should be con- intramuscular injection.When reused, nee- tration is best for a given patient, and
sidered. It is important to note that while dles may be duller and thus injection more research to guide this decision-making is
many insulin types are available for pur- painful. Proper insulin injection technique needed (121). Thus, the choice of MDI or
chase as either pens or vials, others may is a requisite for obtaining the full benefits an insulin pump is often based upon the
only be available in one form or the other ofinsulintherapy.Concerns with technique individual characteristics of the patient
and there may be significant cost differ- and use of the proper technique are out- and which is most likely to benefit them.
ences between pens and vials (see Table lined in Section 9 “Pharmacologic Ap- Newer systems, such as sensor-augmented
9.3 for a list of insulin product costs with proaches to Glycemic Treatment” (https:// pumps and automatic insulin delivery
dosage forms). Insulin pens may allow doi.org/10.2337/dc21-S009). systems, are discussed elsewhere in this
people with vision impairment or dex- Bolus calculators have been developed section.
terity issues to dose insulin accurately to aid in dosing decisions (115–119). These Adoption of pump therapy in the U.S.
(110–112), while insulin injection aids are are subject to FDA approval to ensure shows geographical variations, which may
also available to help with these issues. safety in terms of dosing recommenda- be related to provider preference or center
(For a helpful list of injection aids, see tions. People who are interested in using characteristics (122,123) and socioeco-
http://main.diabetes.org/dforg/pdfs/ these systems should be encouraged to nomic status, as pump therapy is more
2018/2018-cg-injection-aids.pdf.) In- use those that are FDA approved. Pro- common in individuals of higher socio-
haled insulin can be useful in people vider input and education can be helpful economic status as reflected by race/
who have an aversion to injections. for setting the initial dosing calculations ethnicity, private health insurance, fam-
The most common syringe sizes are with ongoing follow-up for adjustments ily income, and education (123,124).
1 mL, 0.5 mL, and 0.3 mL, allowing doses of as needed. Given the additional barriers to optimal
up to 100 units, 50 units, and 30 units of diabetes care observed in disadvantaged
U-100 insulin, respectively. In a few parts Insulin Pumps groups (125), addressing the differences
of the world, insulin syringes still have Recommendations
in access to insulin pumps and other
U-80 and U-40 markings for older insu- 7.20 Insulin pump therapy may be con- diabetes technology may contribute to
lin concentrations and veterinary insulin, sidered as an option for all adults fewer health disparities.
and U-500 syringes are available for the and youth with type 1 diabetes Pump therapy can be successfully started
use of U-500 insulin. Syringes are gen- who are able to safely manage at the time of diagnosis (126,127). Practical
erally used once but may be reused by the device. A aspects of pump therapy initiation in-
the same individual in resource-limited clude assessment of patient and familyS92 Diabetes Technology Diabetes Care Volume 44, Supplement 1, January 2021
readiness (although there is no consen- Diabetes Control and Complications Trial therapy (130). Another pump option in
sus on which factors to consider in adults (DCCT) (144), data suggest that CSII may people with type 2 diabetes is a dispos-
[128] or pediatric patients), selection of reduce the rates of severe hypoglycemia able patchlike device, which provides a
pump type and initial pump settings, compared with MDI (143,145–147). continuous, subcutaneous infusion of
patient/family education of potential There is also evidence that CSII may rapid-acting insulin (basal) as well as
pump complications (e.g., diabetic ke- reduce DKA risk (143,148) and diabetes 2-unit increments of bolus insulin at
toacidosis [DKA] with infusion set failure), complications, in particular, retinopathy the press of a button (153,155,158).
transition from MDI, and introduction of and peripheral neuropathy in youth, Use of an insulin pump as a means for
advanced pump settings (e.g., temporary compared with MDI (65). Finally, treat- insulin delivery is an individual choice for
basal rates, extended/square/dual wave ment satisfaction and quality-of-life mea- people with diabetes and should be
bolus). sures improved on CSII compared with considered an option in patients who
Older individuals with type 1 diabetes MDI (149,150). Therefore, CSII can be are capable of safely using the device.
benefit from ongoing insulin pump ther- used safely and effectively in youth with
apy. There are no data to suggest that type 1 diabetes to assist with achieving
measurement of C-peptide levels or anti- targeted glycemic control while reduc- Combined Insulin Pump and Sensor
bodies predicts success with insulin pump ing the risk of hypoglycemia and DKA, Systems
therapy (129,130). Additionally, frequency improving quality of life, and prevent- Recommendations
of follow-up does not influence outcomes. ing long-term complications. Based on 7.23 Sensor-augmented pump therapy
Access to insulin pump therapy should be patient–provider shared decision-making, with automatic low glucose sus-
allowed/continued in older adults as it is in insulin pumps may be considered in all pend may be considered for
younger people. pediatric patients with type 1 diabetes. adults and youth with diabetes
Complications of the pump can be In particular, pump therapy may be to prevent/mitigate episodes of
caused by issues with infusion sets (dis- the preferred mode of insulin delivery hypoglycemia. B
lodgement, occlusion), which place pa- for children under 7 years of age (66). 7.24 Automated insulin delivery sys-
tients at risk for ketosis and DKA and thus Because of a paucity of data in adoles- tems may be considered in youth
must be recognized and managed early cents and youth with type 2 diabetes, and adults with type 1 diabetes
(131); lipohypertrophy or, less frequently, there is insufficient evidence to make to improve glycemic control. A
lipoatrophy (132,133); and pump site recommendations. 7.25 Individual patients may be using
infection (134). Discontinuation of pump Common barriers to pump therapy systems not approved by the U.S.
therapy is relatively uncommon today; adoption in children and adolescents are Food and Drug Administration,
the frequency has decreased over the concerns regarding the physical interfer- such as do-it-yourself closed-loop
past few decades, and its causes have ence of the device, discomfort with the systems and others; providers
changed (134,135). Current reasons for idea of having a device on the body, cannot prescribe these systems
attrition are problems with cost, wear- therapeutic effectiveness, and financial but should provide safety infor-
ability, dislike for the pump, suboptimal burden (141,151). mation/troubleshooting/backup
glycemic control, or mood disorders (e.g., advice for the individual devices
anxiety or depression) (136). Insulin Pumps in Patients With Type 2 to enhance patient safety. E
and Other Types of Diabetes
Insulin Pumps in Youth Traditional insulin pumps can be consid- Sensor-Augmented Pumps
The safety of insulin pumps in youth has ered for the treatment of people with Sensor-augmented pumps that suspend
been established for over 15 years (137). type 2 diabetes who are on MDI as well as insulin when glucose is low or predicted
Studying the effectiveness of CSII in low- those who have other types of diabetes to go low within the next 30 min have
ering A1C has been challenging because resulting in insulin deficiency, for in- been approved by the FDA. The Auto-
of the potential selection bias of obser- stance, those who have had a pancrea- mation to Simulate Pancreatic Insulin
vational studies. Participants on CSII may tectomy and/or individuals with cystic Response (ASPIRE) trial of 247 patients
have a higher socioeconomic status that fibrosis (152–156). Similar to data on with type 1 diabetes and documented
may facilitate better glycemic control insulin pump use in people with type 1 nocturnal hypoglycemia showed that
(138) versus MDI. In addition, the fast diabetes, reductions in A1C levels are not sensor-augmented insulin pump therapy
pace of development of new insulins and consistently seen in individuals with with a low glucose suspend function sig-
technologies quickly renders compari- type 2 diabetes when compared with nificantly reduced nocturnal hypoglyce-
sons obsolete. However, RCTs compar- MDI, although they have been in some mia over 3 months without increasing
ing CSII and MDI with insulin analogs studies (154,157). Use of insulin pumps in A1C levels (51). In a different sensor-
demonstrate a modest improvement in insulin-requiring patients with any type augmented pump, predictive low glucose
A1C in participants on CSII (139,140). Ob- of diabetes may improve patient satis- suspend reduced time spent with glucose
servational studies, registry data, and faction and simplify therapy (130,152). ,70 mg/dL from 3.6% at baseline to 2.6%
meta-analysis have also suggested an im- For patients judged tobeclinicallyinsulin (3.2% with sensor-augmented pump
provement of glycemic control in partic- deficient who are treated with an in- therapy without predictive low glucose
ipants on CSII (141–143). Although tensive insulin regimen, the presence or suspend) without rebound hyperglyce-
hypoglycemia was a major adverse ef- absence of measurable C-peptide levels mia during a 6-week randomized cross-
fect of intensified insulin regimen in the does not correlate with response to over trial (159). These devices may offercare.diabetesjournals.org Diabetes Technology S93
the opportunity to reduce hypoglycemia carbohydrate ratios, correction doses, (191,192). Others assist in improving di-
for those with a history of nocturnal hy- and insulin activity. Therefore, these set- abetes outcomes by remotely monitoring
poglycemia. Additional studies have been tings can be evaluated and changed based patient clinical data (for instance, wireless
performed, in adults and children, showing on the patient’s insulin requirements. monitoring of glucose levels, weight, or
the benefits of this technology (160–162). blood pressure) and providing feedback
Digital Health Technology
Automated Insulin Delivery Systems
and coaching (193–198). There are text
Automated insulin delivery systems in- Recommendation messaging approaches that tie into a va-
crease and decrease insulin delivery 7.26 Systems that combine technology riety of different types of lifestyle and
based on sensor-derived glucose level and online coaching can be ben- treatment programs, which vary in terms
eficial in treating prediabetes and of their effectiveness (199,200). For many
to begin to approximate physiologic in-
diabetes for some individuals. B of these interventions, there are limited
sulin delivery. These systems consist of
RCT data and long-term follow-up is lack-
three components: an insulin pump, a
Increasingly, people are turning to the ing. But for an individual patient, opting
continuous glucose sensor, and an algo-
internet for advice, coaching, connec- into one of these programs can be helpful
rithm that determines insulin delivery.
tion, and health care. Diabetes, in part and, for many, is an attractive option.
With these systems, insulin delivery can
because it is both common and numeric,
not only be suspended but also increased Inpatient Care
lends itself to the development of apps
or decreased based on sensor glucose
and online programs. The FDA approves Recommendation
values. While eventually insulin delivery
and monitors clinically validated, digital, 7.27 Patients using diabetes devices
in closed-loop systems may be truly
usually online, health technologies in- should be allowed to use them
automated, currently meals must be an-
tended to treat a medical or psycholog- in an inpatient setting when
nounced. A so-called hybrid approach,
ical condition; these are known as digital proper supervision is available. E
hybrid closed-loop, has been adopted
therapeutics or “digiceuticals” (188). Other
in first-generation closed-loop systems
applications, such as those that assist in Patients who are comfortable using their
and requires users to bolus for meals
displaying or storing data, encourage a diabetes devices, such as insulin pumps
and snacks. Multiple studies, using a variety healthy lifestyle or provide limited clinical and sensors, should be given the chance to
of systems with varying algorithms, pump, data support. Therefore, it is possible to use them in an inpatient setting if they are
and sensors, have been performed in find apps that have been fully reviewed competent to do so (201,202). Patients
adults and children (163–173). Evidence and approved and others designed and who are familiar with treating their own
suggests such systems may reduce A1C promoted by people with relatively little glucose levels can often adjust insulin
levels and improve time in range (174– skill or knowledge in the clinical treat- doses more knowledgably than inpatient
178). They may lower the risk of exercise- ment of diabetes. staff who do not personally know the
related hypoglycemia (179) and may have An area of particular importance is patient or their management style. How-
psychosocial benefits (180–183). Use of that of online privacy and security. There ever, this should occur based on the
these systems depends on patient prefer- are established cloud-based data collec- hospital’s policies for diabetes manage-
ence and selection of patients (and/or tion programs, such as Tidepool, Glooko, ment, and there should be supervision to
caregivers) who are capable of safely and and others, that have been developed be sure that the individual can adjust their
effectively using the devices. with appropriate data security features insulin doses in a hospitalized setting
Some people with type 1 diabetes and are compliant with the U.S. Health where factors such as infection, certain
have been using “do-it-yourself” (DIY) Insurance Portability and Accountability medications, immobility, changes in diet,
systems that combine a pump and an Act of 1996. These programs can be and other factors can impact insulin sen-
rtCGM with a controller and an algorithm useful for monitoring patients, both by sitivity and the response to insulin.
designed to automate insulin delivery the patients themselves as well as their With the advent of the coronavirus
(184–187). These systems are not ap- health care team (189). Consumers should disease 2019 pandemic, the FDA has
proved by the FDA, although there are read the policy regarding data privacy and allowed CGM use in the hospital for
efforts underway to obtain regulatory sharing before entering data into an ap- patient monitoring (203). This approach
approval for them. The information on plication and learn how they can control has been employed to reduce the use of
how to set up and manage these systems the way their data will be used (some personal protective equipment and more
is freely available on the internet, and programs offer the ability to share more or closely monitor patients, so that medical
there are internet groups where people less information, such as being part of a personnel do not have to go into a patient
inform each other as to how to set up and registry or data repository or not). room solely for the purpose of measuring a
use them. Although these systems cannot There are many online programs that glucose level. Studies are underway to
be prescribed by providers, it is important offer lifestyle counseling to aid with weight assess the effectiveness of this approach,
to keep patients safe if they are using loss and increase physical activity (190). which may ultimately lead to the routine
these methods for automated insulin de- Many of these include a health coach and use of CGM for monitoring hospitalized
livery. Part of this entails making sure can create small groups of similar patients patients (204,205).
people have a “backup plan” in case of in social networks. There are programs that
pump failure. Additionally, in most DIY aim to treat prediabetes and prevent pro- The Future
systems, insulin doses are adjusted based gression to diabetes, often following the The pace of development in diabetes
on the pump settings for basal rates, model of the Diabetes Prevention Program technology is extremely rapid. NewS94 Diabetes Technology Diabetes Care Volume 44, Supplement 1, January 2021
approaches and tools are available each blood glucose meter that provides personalized monitoring of blood glucose on glucose control in
year. It is hard for research to keep up guidance, insight, and encouragement. J Diabe- patients with non-insulin-treated type 2 diabe-
tes Sci Technol 2020;14:318–323 tes: a meta-analysis of randomized controlled
with these advances because by the trials. J Diabetes Sci Technol 2018;12:183–189
10. Shaw RJ, Yang Q, Barnes A, et al. Self-
time a study is completed, newer ver- monitoring diabetes with multiple mobile health 24. Sai S, Urata M, Ogawa I. Evaluation of
sions of the devices are already on the devices. J Am Med Inform Assoc 2020;27:667– linearity and interference effect on SMBG and
market. The most important component 676 POCT devices, showing drastic high values, low
in all of these systems is the patient. 11. Gellad WF, Zhao X, Thorpe CT, Mor MK, Good values, or error messages. J Diabetes Sci Technol
CB, Fine MJ. Dual use of Department of Veterans 2019;13:734–743
Technology selection must be appropri- Affairs and Medicare benefits and use of test 25. Ginsberg BH. Factors affecting blood glucose
ate for the individual. Simply having a strips in veterans with type 2 diabetes mellitus. monitoring: sources of errors in measurement.
device or application does not change JAMA Intern Med 2015;175:26–34 J Diabetes Sci Technol 2009;3:903–913
outcomes unless the human being en- 12. Endocrine Society and Choosing Wisely. Five 26. Juvenile Diabetes Research Foundation Con-
things physicians and patients should question. tinuous Glucose Monitoring Study Group; Tamborlane
gages with it to create positive health
Accessed 1 November 2020. Available from http:// WV, Beck RW, Bode BW, et al. Continuous
benefits. This underscores the need for www.choosingwisely.org/societies/endocrine- glucose monitoring and intensive treatment of
the health care team to assist the society/ type 1 diabetes. N Engl J Med 2008;359:1464–
patient in device/program selection and 13. Ziegler R, Heidtmann B, Hilgard D, Hofer S, 1476
to support its use through ongoing ed- Rosenbauer J, Holl R; DPV-Wiss-Initiative. Fre- 27. Danne T, Nimri R, Battelino T, et al. Interna-
quency of SMBG correlates with HbA1c and tional consensus on use of continuous glucose
ucation and training. Expectations must
acute complications in children and adolescents monitoring. Diabetes Care 2017;40:1631–1640
be tempered by realitydwe do not yet with type 1 diabetes. Pediatr Diabetes 2011;12: 28. Beck RW, Riddlesworth T, Ruedy K, et al.;
have technology that completely elimi- 11–17 DIAMOND Study Group. Effect of continuous
nates the self-care tasks necessary for 14. Rosenstock J, Davies M, Home PD, Larsen J, glucose monitoring on glycemic control in adults
treating diabetes, but the tools described Koenen C, Schernthaner G. A randomised, with type 1 diabetes using insulin injections: the
52-week, treat-to-target trial comparing insulin DIAMOND randomized clinical trial. JAMA 2017;
in this section can make it easier to
detemir with insulin glargine when administered 317:371–378
manage. as add-on to glucose-lowering drugs in insulin- 29. Lind M, Polonsky W, Hirsch IB, et al. Con-
naive people with type 2 diabetes. Diabetologia tinuous glucose monitoring vs conventional ther-
References 2008;51:408–416 apy for glycemic control in adults with type 1
1. Diabetes Control and Complications Trial Re- 15. Garber AJ. Treat-to-target trials: uses, in- diabetes treated with multiple daily insulin in-
search Group; Nathan DM, Genuth S, Lachin J, terpretation and review of concepts. Diabetes jections: the GOLD randomized clinical trial.
et al. The effect of intensive treatment of diabetes Obes Metab 2014;16:193–205 JAMA 2017;317:379–387
on the development and progression of long- 16. Farmer A, Wade A, Goyder E, et al. Impact of 30. Cowart K, Updike W, Bullers K. Systematic
term complications in insulin-dependent diabe- self monitoring of blood glucose in the manage- review of randomized controlled trials evaluating
tes mellitus. N Engl J Med 1993;329:977–986 ment of patients with non-insulin treated di- glycemic efficacy and patient satisfaction of in-
2. King F, Ahn D, Hsiao V, Porco T, Klonoff DC. A abetes: open parallel group randomised trial. termittent-scanned continuous glucose monitor-
review of blood glucose monitor accuracy. BMJ 2007;335:132 ing in patients with diabetes. Diabetes Technol
Diabetes Technol Ther 2018;20:843–856 17. O’Kane MJ, Bunting B, Copeland M, Coates Ther 2020;22:337–345
3. Brazg RL, Klaff LJ, Parkin CG. Performance VE; ESMON study group. Efficacy of self moni- 31. Evans M, Welsh Z, Ells S, Seibold A. The
variability of seven commonly used self-moni- toring of blood glucose in patients with newly impact of flash glucose monitoring on glycaemic
toring of blood glucose systems: clinical consid- diagnosed type 2 diabetes (ESMON study): rand- control as measured by HbA1c: a meta-analysis of
erations for patients and providers. J Diabetes omised controlled trial. BMJ 2008;336:1174–1177 clinical trials and real-world observational stud-
Sci Technol 2013;7:144–152 18. Simon J, Gray A, Clarke P, Wade A, Neil A, ies. Diabetes Ther 2020;11:83–95
4. Klonoff DC, Parkes JL, Kovatchev BP, et al. Farmer A; Diabetes Glycaemic Education and 32. Reddy M, Jugnee N, El Laboudi A, Spanudakis
Investigation of the accuracy of 18 marketed Monitoring Trial Group. Cost effectiveness of self E, Anantharaja S, Oliver N. A randomized con-
blood glucose monitors. Diabetes Care 2018;41: monitoring of blood glucose in patients with non- trolled pilot study of continuous glucose mon-
1681–1688 insulin treated type 2 diabetes: economic eval- itoring and flash glucose monitoring in people
5. Grady M, Lamps G, Shemain A, Cameron H, uation of data from the DiGEM trial. BMJ 2008; with type 1 diabetes and impaired awareness of
Murray L. Clinical evaluation of a new, lower pain, 336:1177–1180 hypoglycaemia. Diabet Med 2018;35:483–490
one touch lancing device for people with di- 19. Young LA, Buse JB, Weaver MA, et al.; 33. Aleppo G, Ruedy KJ, Riddlesworth TD, et al.;
abetes: virtually pain-free testing and improved Monitor Trial Group. Glucose self-monitoring REPLACE-BG Study Group. REPLACE-BG: a ran-
comfort compared to current lancing systems. in non-insulin-treated patients with type 2 di- domized trial comparing continuous glucose
J Diabetes Sci Technol. 17 July 2019 [Epub ahead abetes in primary care settings: a randomized monitoring with and without routine blood
of print]. DOI: 10.1177/1932296819856665 trial. JAMA Intern Med 2017;177:920–929 glucose monitoring in adults with well-controlled
6. Harrison B, Brown D. Accuracy of a blood 20. Polonsky WH, Fisher L, Schikman CH, et al. type 1 diabetes. Diabetes Care 2017;40:538–545
glucose monitoring system that recognizes in- Structured self-monitoring of blood glucose sig- 34. U.S. Food and Drug Administration. FDA news
sufficient sample blood volume and allows ap- nificantly reduces A1C levels in poorly controlled, release: FDA expands indication for continuous
plication of more blood to the same test strip. noninsulin-treated type 2 diabetes: results from glucose monitoring system, first to replace fin-
Expert Rev Med Devices 2020;17:75–82 the Structured Testing Program study. Diabetes gerstick testing for diabetes treatment decisions.
7. Miller KM, Beck RW, Bergenstal RM, et al.; Care 2011;34:262–267 2016. Accessed 1 November 2020. Available from
T1D Exchange Clinic Network. Evidence of a 21. Malanda UL, Welschen LMC, Riphagen II, https://www.fda.gov/newsevents/newsroom/
strong association between frequency of self- Dekker JM, Nijpels G, Bot SDM. Self-monitoring pressannouncements/ucm534056.htm
monitoring of blood glucose and hemoglobin A1c of blood glucose in patients with type 2 diabetes 35. U.S. Food and Drug Administration. FDA
levels in T1D exchange clinic registry participants. mellitus who are not using insulin. Cochrane news release:. FDA approves first continuous
Diabetes Care 2013;36:2009–2014 Database Syst Rev 2012;1:CD005060 glucose monitoring system for adults not requi-
8. Grant RW, Huang ES, Wexler DJ, et al. Patients 22. Willett LR. ACP Journal Club. Meta-analysis: ring blood sample calibration. 2017. Accessed 1 Nov-
who self-monitor blood glucose and their unused self-monitoring in non-insulin-treated type 2 di- ember 2020. Available from https://www.fda.gov/
testing results. Am J Manag Care 2015;21:e119– abetes improved HbA1c by 0.25%. Ann Intern NewsEvents/Newsroom/PressAnnouncements/
e129 Med 2012;156:JC6–JC12 ucm577890.htm
9. Katz LB, Stewart L, Guthrie B, Cameron H. 23. Mannucci E, Antenore A, Giorgino F, Scavini 36. Battelino T, Danne T, Bergenstal RM, et al.
Patient satisfaction with a new, high accuracy M. Effects of structured versus unstructured self- Clinical targets for continuous glucose monitoringYou can also read
