Abstract
Objective:
Continuous glucose monitor (CGM) use is associated with improved glucose control. We describe the effect of continued and interrupted CGM use on hemoglobin A1c (HbA1c) in youth with public insurance.
Methods:
We reviewed 956 visits from 264 youth with type 1 diabetes (T1D) and public insurance. Demographic data, HbA1c and two-week CGM data were collected. Youth were classified as never user, consistent user, insurance discontinuer, and self-discontinuer. Visits were categorized as never-user visit, visit before CGM start, visit after CGM start, visit with continued CGM use, visit with initial loss of CGM, visit with continued loss of CGM, and visit where CGM is regained after loss. Multivariate regression adjusting for age, sex, race, diabetes duration, initial HbA1c, and body mass index were used to calculate adjusted mean and delta HbA1c.
Results:
Adjusted mean HbA1c was lowest for the consistent user group (HbA1c 8.6%;[95%CI 7.9,9.3]). Delta HbA1c (calculated from visit before CGM start) was lower for visit after CGM start (−0.39%;[95%CI −0.78,−0.02]) and visit with continued CGM use (−0.29%;[95%CI −0.61,0.02]), whereas it was higher for visit with initial loss of CGM (0.40%;[95%CI −0.06,0.86]), visit with continued loss of CGM (0.46%;[95%CI 0.06,0.85]), and visit where CGM is regained after loss (0.57%;[95%CI 0.06,1.10]).
Conclusions:
Youth with public insurance using CGM have improved HbA1c, but only when CGM use is uninterrupted. Interruptions in use, primarily due to gaps in insurance coverage of CGM, were associated with increased HbA1c. These data support both initial and ongoing coverage of CGM for youth with T1D and public insurance.
Keywords: diabetes technology, health policy, insurance, minority health, pediatric type 1 diabetes
1 |. INTRODUCTION
Optimal glucose control paired with improved quality of life is an important management goal for youth with type 1 diabetes (T1D) and providers who care for them.1–3 Incorporation of continuous glucose monitors (CGM) in the management of pediatric T1D has demonstrated improvements in both diabetes outcomes and quality of life.4–9 9 CGM use is associated with improved hemoglobin A1c (HbA1c) irrespective of insulin delivery mode.6 Furthermore, the newer generations of CGM are factory calibrated and do not require fingerstick blood glucose checks for accuracy. For these reasons, in 2018, the International Society for Pediatric and Adolescent Diabetes (ISPAD) encouraged the use of CGM in youth with diabetes in the chapter on Diabetes Technology.2 In 2020, the American Diabetes Association (ADA) Standards of Care state that CGM is the preferred method to monitor glucose and the benefits of CGM use are observed with adherence and ongoing use in youth with T1D using insulin.3,4
In the United States, insurance coverage determines many management choices for youth with T1D. In addition to inconsistent CGM coverage by state, public insurers have more restrictive CGM approval policies when compared to private insurers. For example, in order to receive and maintain approval for CGM coverage for children in California, public insurance policies have long required at least three fingerstick blood glucose checks per day, despite limited clinical need once on a CGM.10 Data from national and international registries demonstrate that the use of CGM has increased in the last decade.8,9 However, youth with public insurance in the United States use CGM at lower rates than those with private insurance.8,9 In addition, data have consistently demonstrated that youth who have public insurance or are of low socioeconomic status (SES) have higher mean HbA1c, higher rates of diabetic ketoacidosis, and lower diabetes related quality of life measures.11–17 We have previously demonstrated that when youth with public insurance are provided CGM, they consistently use them.11
In this study, we describe CGM use in our publicly insured youth with T1D with a particular focus on understanding the effect of interrupted and uninterrupted CGM use and its association with HbA1c changes. We hypothesize that interruption of CGM use (especially due to loss of public insurance coverage) will result in increased HbA1c.
2 |. METHODS
In this retrospective cohort study, we reviewed the records of all youth with T1D who have public insurance (Medi-Cal/Medicaid, Medicare, or California Children’s Services) and have attended Stanford Children’s Diabetes Clinic between 2010 and 2019.
For those with public insurance who have never used CGM, data were collected:
At the most recent clinic visit available in the electronic medical record (follow-up visit); and
Two years prior to the most recent visit or their first clinic visit if seen in our clinic for <2 years, whichever was the oldest (baseline visit).
HbA1c, insulin delivery method, self-monitoring blood glucose (SMBG), and CGM use were gathered for each visit. SMBG data were obtained from scanned downloads of glucometers.
For those with public insurance who have used CGM, the baseline visit was defined as the clinic visit when the CGM was started. If the CGM was started at home or in another location, the baseline visit was considered to be the clinic visit immediately before CGM initiation. HbA1c, insulin delivery method, SMBG, and CGM use were gathered for the baseline visit. CGM use at every clinic visit following the baseline visit was recorded (CGM use: yes/no). For visits where CGM use continued, HbA1c, insulin delivery mode, SMBG, time in range (TIR; 70–180 mg/dL), % hyperglycemia (≥180 mg/dL), % hypoglycemia (<70 mg/dL), and days worn over a two-week period were captured. CGM data were obtained from a download of the CGM data scanned into the medical record. For visits where CGM use was discontinued, reason for discontinuation (as documented in clinic notes), HbA1c, insulin delivery mode, and SMBG (from scanned download of the glucometer) were collected. All data were gathered via chart review from the electronic medical record. We excluded 10 youth who had public insurance and the diagnosis of T1D but did not have follow-up clinical visits at Stanford Children’s.
Youth (n = 264) and their visits (n = 956, range of 2–14 visits per youth) were categorized by CGM use as outlined below.
Youth were categorized into one of four user type groups by CGM use or discontinuation (Figure 1):
Never User: Youth who have never used CGM.
Consistent User: Youth who used CGM continuously throughout the study period without any interruptions in CGM use.
Insurance Discontinuer: Youth where the first documented loss of CGM was due to issues with insurance coverage (paperwork, supply delivery, changes to coverage policy).
Self-Discontinuer: Youth where the first documented loss of CGM was due to personal preference of the youth (not wanting to wear a device, irritation from tape, loss of interest).
FIGURE 1.
(a) Schematic of the Categorization of Youth Group Types. Legend: Of the 264 youth, 149 youth never used continuous glucose monitor (CGM), whereas 115 youth used CGM. Of the 115 youth who used CGM, 69 youth used CGM continuously throughout the study period without any interruptions in CGM use. 30 youth lost CGM due to issues with insurance coverage (paperwork, supply delivery, changes to coverage policy) and 16 youth stopped CGM due to personal preference (not wanting to wear a device, irritation from tape, loss of interest). (b) Schematic of visit type categorization. Legend: Visits for youth who used CGM were categorized into visit before CGM start (n = 115) and visit after CGM start (n = 114). The remaining 426 visits were compared to the visit prior to determine if CGM use continued without interruptions (visit with continued CGM use) or if there were interruptions to CGM use. Visit with initial Loss of CGM refers to the visit immediately following CGM discontinuation, visit with continued loss of CGM are visits where there is no CGM use at the prior or current visit, and visit where CGM is regained after loss denotes visits where there CGM use in the current visit but not at the prior visit. *1 youth lost CGM access between visit before CGM start and visit after CGM start due to supply issues and was therefore categorized as visit with initial loss of CGM
Baseline and follow-up visits (n = 956) of youth from the never user group were categorized as never user visit. For youth who used CGM, visits were further categorized into one of six groups by comparing CGM use at current visit to the visit prior (Figure 1):
Visit Before CGM Start: The visit immediately before starting CGM.
Visit After CGM Start: The visit immediately after starting CGM.
Visit With Continued CGM Use: CGM use at both the prior and current visit.
Visit With Initial Loss of CGM: The visit immediately following CGM discontinuation.
Visit With Continued Loss of CGM: No CGM use at the prior or current visit.
Visit Where CGM is Regained After Loss: CGM use in the current visit but not at the prior visit.
2.1 |. Glycemic data
The primary measure of glycemic control was HbA1c. For those who wore CGM, TIR (70–180 mg/dL), % hyperglycemia (≥180 mg/dL), % hypoglycemia (<70 mg/dL), and days of wear were collected from two-week CGM data downloads scanned into the electronic medical record through manual chart review or provider documentation.
2.2 |. Demographic data
For each youth in the study, age, sex, family’s primary language, self-reported race and ethnicity, diabetes duration, and body mass index (BMI; kg/m2) were extracted from record review of the electronic medical record.
2.3 |. Analysis
For individual level data, descriptive statistics, t-tests, and chi-square analyses were used to measure differences across youth group type (never user, consistent user, insurance discontinuer, and self-discontinuer). Multivariate linear regression controlling for age, sex, duration of diabetes, and initial HbA1c was used to calculate adjusted mean HbA1c and evaluate for significance by youth group types.18
The impact of CGM loss or gain was analyzed using visit-to-visit data for visit after CGM start, visit with continued CGM use, visit with initial loss of CGM, visit with continued loss of CGM, and visit where CGM is regained after loss. To account for the repeated measures that each youth contributes through multiple visits, a mean change in HbA1c, defined as HbA1c at most recent visit minus HbA1c at visit before CGM start, was calculated. An adjusted mean delta HbA1c was then modeled using multivariate regression least square means18 with the model adjusting for sex, race, age, diabetes duration, and BMI. In a final step, the regression analyses were repeated for individual level data and visit level data excluding youth who started CGM on or before May 2016 (analysis included 251 youth and 871 visits). CGM coverage policies changed in California in June 2016 to formalize a pathway for coverage, therefore, this final analysis was carried out in order to evaluate any effect of a formal policy on glycemic outcomes. All statistical analyses were executed with the statistical software R 3.5.
3 |. RESULTS
Participant characteristics of youth (n = 264) and visits (n = 956) are outlined in Tables 1 and 2, respectively. Consistent with our clinic’s publicly insured demographics, Hispanic Whites represented the largest race/ethnicity group and approximately 25% of our youths’ families speak Spanish as the primary language. In our cohort, 56% never used CGM (never user). Compared to those who used CGM, the never user group was older (13.5 ± 3.7 years vs 12.1 ± 4.6 years, P = .02), with shorter diabetes duration (4.4 ± 4.4 years vs 5.2 ± 4.5 years, P = .15), had fewer SMBG at baseline (5.4 ± 2.6 checks per day vs 3.5 ± 2.0 checks per day; P < .001), lower insulin pump use (12.8% vs 47.0%, P < .001), and higher HbA1c at baseline (9.7% ± 2.1% vs 8.8% ± 1.7%; P < .001). The never user and user groups were contemporaneous populations with similar time frames for the baseline visits (July 2013-May 2019 for the user group and June 2012-March 2019 for the never user group).
TABLE 1.
Participant characteristics by youth group type
| All (n = 264) | Never user (n = 149) | Usera (n = 115) | Consistent user (n = 69) | Insurance discontinuer (n = 30) | Self-discontinuer (n = 16) | |
|---|---|---|---|---|---|---|
| Ageb | 12.9 ± 4.2 | 13.5 ± 3.7 | 12.1 ± 4.6 | 12.3 ± 5.9 | 11.4 ± 4.5 | 12.5 ± 2.9 |
| Diabetes durationb | 4.8 ± 4.4 | 4.4 ± 4.4 | 5.2 ± 4.5 | 5.1 ± 5.0 | 5.2 ± 3.7 | 5.6 ± 3.6 |
| Sexc | ||||||
| Female | 128 (48.5) | 69 (46.3) | 59 (51.3) | 33 (47.8) | 17 (56.7) | 9 (56.2) |
| Male | 136 (51.5) | 80 (53.7) | 56 (48.7) | 36 (52.2) | 13 (43.3) | 7 (43.8) |
| Racec | ||||||
| Non-hispanic white | 72 (27.3) | 38 (25.5) | 34 (29.6) | 21 (30.4) | 9 (30.0) | 4 (25.0) |
| Hispanic | 129 (48.9) | 80 (53.7) | 49 (42.6) | 27 (39.1) | 16 (53.3) | 6 (37.5) |
| African American | 15 (5.7) | 11 (7.4) | 4 (3.5) | 1 (1.4) | 1 (3.3) | 2 (12.5) |
| Asian or Pacific Islander | 14 (5.3) | 6 (4.0) | 8 (7.0) | 5 (7.2) | 2 (6.7) | 1 (6.2) |
| Native American or Alaskan native | 1 (0.4) | 0 (0) | 1 (0.9) | 0 (0) | 1 (3.3) | 0 (0) |
| Other | 15 (5.7) | 5 (3.4) | 10 (8.7) | 9 (13.0) | 0 (0) | 1 (6.2) |
| Unknown | 18 (6.8) | 9 (6.0) | 9 (7.8) | 6 (8.7) | 1 (3.3) | 2 (12.5) |
| Primary languagec | ||||||
| English | 187 (70.8) | 101 (67.8) | 86 (74.8) | 49 (71.0) | 22 (73.3) | 15 (93.8) |
| Spanish | 70 (26.5) | 43 (28.9) | 27 (23.5) | 19 (27.5) | 7 (23.3) | 1 (6.2) |
| Other | 7 (2.7) | 5 (3.4) | 2 (1.7) | 1 (1.4) | 1 (3.3) | 0 (0) |
| Self-monitoring blood glucoseb | 4.4 ± 2.5 | 3.5 ± 2.0 | 5.4 ± 2.6 | 5.2 ± 2.3 | 6.0 ± 3.2 | 5.2 ± 2.4 |
| Insulin deliveryc | ||||||
| MDI | 188 (70.7%) | 127 (85.2%) | 61 (53.0%) | 38 (55.1%) | 18 (60.0%) | 5 (31.2%) |
| Insulin pump | 73 (27.4%) | 19 (12.8%) | 54 (47.0%) | 31 (44.9%) | 12 (40.0%) | 11 (68.8%) |
| Both | 1 (0.4%) | 1 (0.7%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Missing | 2 (0.8%) | 2 (1.3%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Baseline HbA1c, %b | 9.3 ± 2.0 | 9.7 ± 2.1 | 8.7 ± 1.7 | 8.7 ± 1.8 | 8.9 ± 1.6 | 8.5 ± 1.4 |
User group (n = 115) is composed of consistent users (n = 69), insurance discontinuer (n = 30), and self-discontinuer (n = 16).
Mean ± SD.
n (%).
TABLE 2.
Characteristics of visit type
| All visits (n = 956) | Never user visit (n = 301) | Visit before CGM start (n = 115) | Visit after CGM start (n = 114) | Visit with continued CGM use (n = 253) | Visit with initial loss of CGM (n = 53) | Visit with continued loss of CGM (n = 76) | Visit where CGM is regained after loss (n = 44) | |
|---|---|---|---|---|---|---|---|---|
| Age, yearsa | 13.1 ± 4.4 | 14.3 ± 3.9 | 12.2 ± 4.6 | 12.5 ± 4.6 | 12.3 ± 4.8 | 13.1 ± 4.0 | 12.6 ± 3.8 | 13.2 ± 3.4 |
| Diabetes duration, yearsa | 5.6 ± 4.2 | 5.4 ± 4.5 | 5.1 ± 4.4 | 5.5 ± 4.5 | 5.4 ± 3.9 | 6.4 ± 3.6 | 6.6 ± 3.4 | 6.6 ± 3.6 |
| Sexb | ||||||||
| Female | 476 (49.6) | 138 (45.8) | 59 (51.3) | 59 (51.8) | 132 (52.2) | 28 (52.8) | 35 (46.1) | 24 (54.5) |
| Male | 483 (50.4) | 163 (54.2) | 56 (48.7) | 55 (48.2) | 121 (47.8) | 25 (47.2) | 41 (53.9) | 20 (45.5) |
| Raceb | ||||||||
| African American | 44 (4.6) | 24 (8.0) | 2 (1.7) | 2 (1.8) | 6 (2.4) | 2 (3.8) | 4 (5.3) | 1 (2.3) |
| Asian or Pacific Islander | 58 (6.0) | 12 (4.0) | 8 (7.0) | 8 (7.0) | 17 (6.7) | 3 (5.7) | 8 (10.5) | 2 (4.5) |
| Hispanic | 462 (48.2) | 160 (53.2) | 49 (42.6) | 49 (43.0) | 121 (47.8) | 26 (49.1) | 38 (50.0) | 21 (47.7) |
| Native American or Alaskan Native | 4 (0.4) | 0 (0) | 1 (0.9) | 1 (0.9) | 0 (0) | 1 (1.9) | 1 (1.3) | 0 (0) |
| Non-Hispanic White | 271 (28.3) | 77 (25.6) | 36 (31.3) | 35 (30.7) | 76 (30.0) | 15 (28.3) | 16 (21.1) | 14 (31.8) |
| Other | 49 (5.1) | 10 (3.3) | 10 (8.7) | 10 (8.8) | 15 (5.9) | 0 (0) | 3 (3.9) | 1 (2.3) |
| Unknown | 71 (7.4) | 18 (6.0) | 9 (7.8) | 9 (7.9) | 18 (7.1) | 6 (11.3) | 6 (7.9) | 5 (11.4) |
| Primary languageb | ||||||||
| English | 696 (72.6) | 205 (68.1) | 90 (75.6) | 73 (73.0) | 187 (73.3) | 38 (74.5) | 62 (77.5) | 34 (77.3) |
| Spanish | 239 (24.9) | 86 (28.6) | 27 (22.7) | 25 (25.0) | 65 (25.5) | 10 (19.6) | 16 (20.0) | 8 (18.2) |
| Other | 24 (2.5) | 10 (3.3) | 2 (1.7) | 2 (2.0) | 3 (1.2) | 3 (5.9) | 2 (2.5) | 2 (4.5) |
| Insulin deliveryb | ||||||||
| MDI | 575 (60.1%) | 254 (84.4%) | 61 (53.0%) | 60 (52.6%) | 120 (47.4%) | 27 (50.9%) | 32 (42.1%) | 21 (47.7%) |
| Insulin pump | 375 (39.2%) | 41 (13.6%) | 54 (47.0%) | 54 (47.4%) | 133 (52.6%) | 26 (49.1%) | 44 (57.9%) | 23 (52.3%) |
| Both | 2 (0.2%) | 2 (0.7%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Missing | 4 (0.4%) | 4 (1.3%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| BMIa | 21.4 ± 6.7 | 23.1 ± 6.2 | 20.4 ± 7.4 | 20.1 ± 7.5 | 20.2 ± 6.9 | 21.8 ± 5.7 | 22.0 ± 4.9 | 22.3 ± 5.0 |
| HbA1c, %a | 9.0 ± 1.9 | 9.7 ± 2.1 | 8.7 ± 1.7 | 8.4 ± 1.4 | 8.5 ± 1.6 | 9.2 ± 1.7 | 9.5 ± 1.4 | 9.0 ± 1.7 |
Abbreviations: BMI, body mass index;CGM, continuous glucose monitor; HbA1c, hemoglobin A1c; MDI, multiple daily injections.
Mean ± SD.
n (%).
3.1 |. Youth characteristics
Of those youth who used CGM, 61% (n = 69) were in the consistent user group and had continuous use of CGM without any interruptions. Over a 14-day period, the consistent user group wore their CGM 11.6 ± 3.2 days and had an average glucose of 209 ± 43 mg/dL with 36.9% ± 17.4% TIR, 2.4% ± 4.8% hypoglycemia, and 59.7% ± 19.3% hyperglycemia.
Among youth who experienced interruptions in CGM use (n = 46), 65% of the interruptions were due to insurance-related reasons and not due to personal preference. The most commonly cited reason to stop CGM use in the self-discontinuer group was issues with the sensor (n = 10), such as accuracy of the sensor, transmitter issues, or irritation from tape. Of those in the insurance discontinuer group, 80% used a Dexcom G5 and 20% used a Dexcom G6. Of those in the self-discontinuer group, 68% used a Dexcom G5 and 42% used a Dexcom G4. Compared to the self-discontinuer group (n = 16), the insurance discontinuer group (n = 30) was younger (11.4 ± 4.5 years vs 12.5 ± 2.9 years; P = .82), more likely to be Hispanic white (53.3% vs 37.5%; P = .06), and spoke Spanish as their primary language (23.3% vs 6.2%; P = .47). When they wore CGM, the insurance discontinuer group wore CGM 11.7 ± 3.5 days in a 14-day period, with a mean glucose of 210 ± 43 mg/dL, 36.6% ± 18.5% TIR, 2.4% ± 3.1% hypoglycemia, and 60.9% ± 20.2% hyperglycemia. When they wore CGM, the self-discontinuer group wore CGM 8.8 ± 4.4 days in a 14-day period, with a mean glucose of 200 ± 45 mg/dL, 31.4% ± 5.2% TIR, 3.0% ± 3.7% hypoglycemia, and 65.5% ± 17.8% hyperglycemia.
3.2 |. Mean HbA1c by youth group type
Adjusted mean HbA1c at the most recent visit for the four youth group types (never user, consistent user, self-discontinuer, and insurance discontinuer) is presented in Figure 2. The never user group had the highest adjusted mean HbA1c (9.7% [95% CI 9.1, 10.4]), whereas the consistent user group had the lowest adjusted mean HbA1c (8.6% [95% CI 7.9, 9.3]; P < .001 compared to never user). Adjusted mean HbA1c in the insurance discontinuer group was 9.0% ([95% CI 8.2, 9.8]; P = .79 compared to consistent user) and the self-discontinuer group was 9.1% ([95% CI 8.0, 10.1]; P = .82 compared to consistent user). When compared to the never user group, the adjusted mean HbA1c was not statistically different for the insurance discontinuer group (P = .19) or the self-discontinuer group (P = .47).
FIGURE 2.
Hemoglobin A1c (HbA1c) at most recent visit for the four user groups
3.3 |. Change in HbA1c in a visit-to-visit comparison
Adjusted mean change in HbA1c in a visit-to-visit comparison for each visit type is presented in Table 3. Delta HbA1c, calculated as visit after CGM start minus visit before CGM start, was −0.31%, 95% CI [−0.66, 0.05]. Delta HbA1c for the visit with continued CGM use was delta HbA1c −0.30% (95% CI [−0.62, 0.01]; P = 1.00 compared to visit after CGM start). A statistically significant rise in HbA1c was noted between the visit before CGM start to visit types involving CGM interruptions as seen in visit with initial loss of CGM (delta HbA1c 0.39% 95% CI [−0.06, 0.84], P = .027), visit with continued loss of CGM (delta HbA1c 0.48% 95% CI [0.08, 0.87], P = .002) and visit where CGM is regained after loss (delta HbA1c 0.51% 95% CI [0.02, 1.01], P = .011). When compared to the visit with continued CGM use, higher delta HbA1c are observed in the visit with initial loss of CGM (P = .011), visit with continued loss of CGM (P < .001), and visit where CGM is regained after loss (P = .004). For reference, adjusted mean HbA1c by visit type are provided in Supplemental Table 1.
TABLE 3.
Adjusted mean change in HbA1c by visit type with model adjusting for sex, race, age, diabetes duration, and BMI
| Visit type | Delta HbA1ca [95% CI] | P-valueb | |
|---|---|---|---|
| Visit after CGM start | −0.31 | [−0.66, 0.05] | Reference group |
| Visit with continued CGM use | −0.30 | [−0.62, 0.01] | 1.00 |
| Visit with initial loss of CGM | 0.39 | [−0.06, 0.84] | .027 |
| Visit with continued loss of CGM | 0.48 | [0.08, 0.87] | .002 |
| Visit where CGM is regained after loss | 0.51 | [0.02, 1.01] | .011 |
Bold values denote statistically significant P-values which is the case for the last three visit types. Abbreviations: BMI, body mass index; CGM, continuous glucose monitor; HbA1c, hemoglobin A1c.
Delta HbA1c is calculated for each visit by subtracting every visit HbA1c from the visit before CGM start HbA1c.
P-values for visit after CGM start as the reference group.
Adjusted mean HbA1c for the four youth group types and adjusted mean change in HbA1c in the visit-to-visit comparison were not different when excluding those who started CGM before June 2016, when CGM coverage policies changed in California (Supplemental Table 2a,b).
4 |. DISCUSSION
In this real-world study of 264 youth with T1D and public insurance, we report that the initiation of CGM is associated with an improvement in HbA1c. Specifically, ongoing and uninterrupted use of CGM is associated with a sustained improvement in HbA1c in youth with public insurance. We also document an adverse impact on HbA1c with the loss of or inconsistent access to CGM. Furthermore, the reacquisition of CGM after a loss does not ameliorate the adverse impact of the loss of CGM on HbA1c. Our findings are consistent with the recommendations outlined by ISPAD and ADA: CGM use should be considered for all youth with T1D (ISPAD chapter 21.4, ADA section 13.19), benefits of CGM in youth with T1D is observed with adherence and ongoing use (ISPAD chapter 21.4.3, ADA sections 13.19 and 7.12), and those who use CGM “should have continued access across third payer parties” (ADA section 7.16).2–4 As noted in the ISPAD diabetes technology chapter, access to CGM is a critical step to ensure equal access for more advanced systems, such as hybrid closed loop systems (ISPAD chapter 21.6.4).2 Findings from this study provide evidence to support current expert opinion for future ISPAD and ADA guidelines specifically advocating for CGM for youth with public insurance.
Prior studies evaluating the effect of CGM use on youth with public insurance have failed to demonstrate improvements in HbA1c.11,19 Findings from this study suggest that ongoing and uninterrupted use of CGM is a modifying factor in the relationship between CGM and HbA1c. Therefore, these data raise important considerations when interpreting studies evaluating efficacy of CGM in youth with public insurance. First, because the consistency of CGM access has an impact on glycemic outcomes, consistency of access should be considered in the interpretation of study results. Second, CGM use has increased dramatically in youth with T1D over the past decade. However, disparity in CGM use by SES has widened in the United States and Germany in the past decade.20 Third, consistent with prior reports, youth who identify as Hispanic have lower rates of CGM use21 and we report a greater number of insurance-related interruptions to CGM use in these youth. Similarly, youth whose family’s primary language is Spanish experienced more insurance-related interruptions, although the small sample size was not powered to detect this difference, these findings may indicate that language is a barrier to CGM access.
These data demonstrate that a significant portion (40%) of youth with public insurance in our clinic have their access to CGM interrupted. Consistent with our prior report,11 publicly insured youth who had gained approval for CGM use had high wear time. It is important to note that the majority of interruptions (65%) were not due to personal preference but was rather due to insurance-related interruptions. The majority of the interruptions was due to the requirements needed to maintain uninterrupted coverage that are unique to public insurance providers in California that private insurers do not require. Public insurers in California began to offer CGM coverage in June 2016. In order to obtain CGM coverage, public insurers require ≥3 SMBG per day, insulin use, a commitment to continually wear CGM, demonstration of hypoglycemia unawareness or maintenance of hyperglycemia due to fear of hypoglycemia, reapproval paperwork 6 months after starting CGM, and yearly approval thereafter in order to receive and maintain CGM coverage.22 Re-approval of CGM required demonstration of ≥3 SMBG per day until the release of the Dexcom G6, a factor in our study period. In contrast, most private insurers simply require annual paperwork requesting a CGM for initial and continued coverage of CGM.
Similar to cost considerations reported for intermittently scanned CGM, CGM may be able to offset monthly costs of test strips thereby offering a potential cost neutral solution for broadening access while improving outcomes.23–25 For those who were self-discontinuers, improvements in adhesives, customizable alarm thresholds, and factory calibration which obviate the need for fingerstick blood glucoses with newer generation CGM, may improve adherence and use of CGM technology.26,27
There are limitations to this study including the retrospective design and a single site. There are likely limitations to generalizability because providers at Stanford Children’s have a tendency to encourage diabetes technology adoption for all youth using insulin, although diabetes technology use is increasing across the US.7 These data include youth with older generation CGM (Dexcom G4 and G5) and the Medtronic Guardian which required SMBG for CGM calibration and/or insulin dosing decisions. With newer factory calibrated devices, adherence to device use will likely show further improvement.26,27 The never user and user groups have differences at baseline, including differences in SMBG, insulin pump use, and baseline HbA1c. The most likely explanation for this observation is that California’s public insurers require three checks per day in order to cover CGM and four checks per day are required to reimburse insulin pump use. Additionally, higher HbA1c is associated with decreased technology use.9,20 Given these baseline differences between the never user and user groups and to remove any impact these baseline differences have on the study outcomes, we structured the visit to visit analysis to compare each visit type to the visit before CGM start. Despite these limitations, the results from this study raise important policy and practice considerations for publicly insured youth with T1D. These data also demonstrate clinically meaningful reduction in HbA1c that is observed in youth with sustained CGM use. These data are consistent with prior reports that CGM use has a beneficial impact on HbA1c5–9 and expands this to include youth with public insurance. Therefore, sustained and uninterrupted access to CGM in youth with T1D may be one strategy to help bridge the disparity in HbA1c in youth with public vs private insurance.
5 |. CONCLUSIONS
These data demonstrate that when provided with uninterrupted access to CGM, youth with public insurance have improvements in HbA1c with CGM use. Despite continued engagement in clinic visits, youth with interrupted access to CGM have increased HbA1c. The majority of interruptions in CGM use were due to insurance related reasons and not due to personal preference. Furthermore, we demonstrate that the reacquisition of CGM after a loss does not return the full benefit of uninterrupted CGM use. Public policy changes are needed to eliminate interruptions to CGM access for youth with T1D who have public insurance. In line with published guidelines, these data support uninterrupted and sustained access to CGM as important policy and practice measures to adopt for youth with T1D who have public insurance.
Supplementary Material
ACKNOWLEDGMENTS
Dr A. A. is supported by the Maternal Child Health Research Institute at Stanford University and is an Ernest and Amelia Gallo Endowed Postdoctoral Fellow. Dr D. M. M. has research support from the NIH, JDRF, NSF, and the Helmsley Charitable Trust and his institution has research support from Medtronic, Dexcom, Insulet, Bigfoot Biomedical, Tandem, and Roche. Dr D. M. M. has consulted for Abbott, the Helmsley Charitable Trust, Sanofi, Novo Nordisk, Eli Lilly and Insulet. He is supported by P30DK116074.
Footnotes
PEER REVIEW
The peer review history for this article is available at https://publons.com/publon/10.1111/pedi.13082.
SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of this article.
REFERENCES
- 1.Caccavale LJ, Corona R, LaRose JG, Mazzeo SE, Sova AR, Bean MK. Exploring the role of motivational interviewing in adolescent patient-provider communication about type 1 diabetes. Pediatr Diabetes. 2019;20(2):217–225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sherr JL, Tauschmann M, Battelino T, et al. ISPAD clinical practice consensus guidelines 2018: diabetes technologies. Pediatr Diabetes. 2018;19:302–325. [DOI] [PubMed] [Google Scholar]
- 3.American Diabetes Association. 13. Children and adolescents: standards of medical care in diabetes−2020. Diabetes Care. 2020;43:20. [DOI] [PubMed] [Google Scholar]
- 4.American Diabetes Association. 7. Diabetes technology Standards of Medical Care in Diabetes—2020. 2020;43:12. [DOI] [PubMed] [Google Scholar]
- 5.Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. Effectiveness of continuous glucose monitoring in a clinical care environment: evidence from the Juvenile Diabetes Research Foundation continuous glucose monitoring (JDRF-CGM) trial. Diabetes Care. 2010;33(1):17–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Miller KM, Beck RW, Foster NC, Maahs DM. HbA1c levels in type 1 diabetes from early childhood to older adults: a deeper dive into the influence of technology and socio-economic status on HbA1c in the T1D exchange clinic registry findings. Diabetes Technol Ther 2020; dia.2019.0393. 10.1089/dia.2019.0393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Miller KM, Hermann J, Foster N, et al. Longitudinal changes in continuous glucose monitoring use among individuals with type 1 diabetes: international comparison in the German and Austrian DPV and U.S. T1D exchange registries. Diabetes Care. 2020;43(1):e1–e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.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(2):66–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.DeSalvo DJ, Miller KM, Hermann JM, et al. Continuous glucose monitoring and glycemic control among youth with type 1 diabetes: international comparison from the T1D exchange and DPV initiative. Pediatr Diabetes. 2018;19(7):1271–1275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Anderson JE, Gavin JR, Kruger DF. Current eligibility requirements for CGM coverage are harmful, costly, and unjustified. Diabetes Technol Ther 2020;22(3):169–173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Prahalad P, Addala A, Buckingham BA, Wilson DM, Maahs DM. Sustained continuous glucose monitor use in low-income youth with type 1 diabetes following insurance coverage supports expansion of continuous glucose monitor coverage for all. Diabetes Technol Ther 2018;20(9):632–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nielsen NF, Gaulke A, Eriksen TM, Svensson J, Skipper N. Socioeconomic inequality in metabolic control among children with type 1 diabetes: a Nationwide longitudinal study of 4,079 Danish children. Diabetes Care. 2019;42(8):1398–1405. [DOI] [PubMed] [Google Scholar]
- 13.Hassan K, Loar R, Anderson BJ, Heptulla RA. The role of socioeconomic status, depression, quality of life, and glycemic control in type 1 diabetes mellitus. J Pediatr 2006;149(4):526–531. [DOI] [PubMed] [Google Scholar]
- 14.Prahalad P, Tanenbaum M, Hood K, Maahs DM. Diabetes technology: improving care, improving patient-reported outcomes and preventing complications in young people with type 1 diabetes. Diabet Med 2018;35(4):419–429. [DOI] [PubMed] [Google Scholar]
- 15.Cengiz E, Xing D, Wong JC, et al. Severe hypoglycemia and diabetic ketoacidosis among youth with type 1 diabetes in the T1D exchange clinic registry. Pediatr Diabetes. 2013;14(6):447–454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Govan L, Maietti E, Torsney B, et al. The effect of deprivation and HbA1c on admission to hospital for diabetic ketoacidosis in type 1 diabetes. Diabetologia. 2012;55(9):2356–2360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Auzanneau M, Lanzinger S, Bohn B, et al. Area deprivation and regional disparities in treatment and outcome quality of 29,284 pediatric patients with type 1 diabetes in Germany: a cross-sectional multicenter DPV analysis. Diabetes Care. 2018;41(12): 2517–2525. [DOI] [PubMed] [Google Scholar]
- 18.Lenth RV. Least-squares means: the R package lsmeans. J Stat Softw 2016;69(1):11107. [Google Scholar]
- 19.Sequeira PA, Montoya L, Ruelas V, et al. Continuous glucose monitoring pilot in low-income type 1 diabetes patients. Diabetes Technol Ther 2013;15(10):855–858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Addala A, Auzanneau M, Miller K, et al. A decade of disparities in diabetes technology use and HbA1c in pediatric type 1 diabetes: a trans-Atlantic comparison. Pediatr Diabetes. 2019;20(S28):P313. [Google Scholar]
- 21.Willi SM, Miller KM, DiMeglio LA, et al. Racial-ethnic disparities in management and outcomes among children with type 1 diabetes. Pediatrics. 2015;135(3):424–434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Department of Health Care Services. Continuous glucose monitoring (cgm) as a ccs/ghpp program benefit- revised with attachment [Internet]. Available from: https://www.dhcs.ca.gov/services/ccs/Documents/CCS.NL.14-0818.Continuous.Glucose.Monitoring.pdf
- 23.Fokkert MJ, Damman A, van Dijk PR, et al. Use of FreeStyle libre flash monitor register in The Netherlands (FLARE-NL1): patient experiences, satisfaction, and cost analysis. Int J Endocrinol 2019;2019: 4649303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.McQueen RB, Ellis SL, Campbell JD, Nair KV, Sullivan PW. Cost-effectiveness of continuous glucose monitoring and intensive insulin therapy for type 1 diabetes. Cost Eff Resour Alloc 2011;9(1):13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wan W, Skandari MR, Minc A, et al. Cost-effectiveness of continuous glucose monitoring for adults with type 1 diabetes compared with self-monitoring of blood glucose: the DIAMOND randomized trial. Diabetes Care. 2018;41(6):1227–1234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.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(3):484–492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wong J, Barley R, Hanes S, et al. Parental perspectives: identifying profiles of parental attitudes and barriers related to diabetes device use. Diabetes Technol Ther 2020. 10.1089/dia.2019.0492. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.