Correspondence to Dr Rozalina G McCoy; [email protected]
WHAT IS ALREADY KNOWN ON THIS TOPIC
Racial and ethnic disparities in utilization of diabetes technology have been consistently documented. Few studies have explored these disparities on a national scale or among youth covered by commercial health insurance.
WHAT THIS STUDY ADDS
This study highlights the importance of race/ethnicity and socioeconomic status (e.g., income) as factors that influence utilization of diabetes technology. Racial disparities in use are lessened at higher income levels. Black youth remain at the greatest risk of exclusion from these essential technologies.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Targeted interventions at both individual and systemic levels are critical to addressing inequities in utilization of diabetes technology.
Introduction
Over the past two decades, there has been a significant rise in the incidence and prevalence of type 1 diabetes (T1D) in both the USA and globally.1 2 While it continues to be most prevalent among non-Hispanic white individuals, this uptake in incidence is evident across all ethnicities and races, including Hispanic and non-Hispanic black individuals.1–3 Since the Diabetes Control and Complication Trial study4 intensive insulin therapy has been the standard of care to optimize glycemic management and achieve individualized glycemic targets in children and adolescents with T1D to prevent its long-term complications. The advent and increasingly widespread use of diabetes technology have made it possible to continuously deliver insulin and monitor glucose levels in real-time, allowing for more precise and safe diabetes management and attainment of individualized glycemic goals.5 6
Continuous glucose monitors (CGMs), insulin pumps, and hybrid closed-loop systems have significantly improved glycemic management, resulting in a better quality of life for both youth and their parents and reducing their fear of hypoglycemia.7 Due to these positive impacts on clinical and non-clinical outcomes, professional societies’ guidelines recommend that insulin pumps and CGMs be considered a standard of care for all youth with T1D.6 8 9
Recognizing the changing demographics of T1D and the pervasive disparities in diabetes-related outcomes among youth from racial and ethnic minoritized communities,10 it is crucial to identify potential disparities in technology adoption across different races and ethnicities. Inequalities in the utilization of diabetes technology among youth have been demonstrated at local and regional levels, with an emphasis on the influence of racial and socioeconomic factors.11 12 In this study, we evaluate the impacts of race/ethnicity and socioeconomic status (SES), as measured by the annual household income, on the utilization of diabetes technology nationwide by examining the fill rates of insulin pumps and CGMs among commercially insured children and adolescents with T1D.
Methods
Study design
This is a retrospective cross-sectional analysis of medical and pharmacy administrative claims data from OptumLabs Data Warehouse (OLDW), a deidentified claims database of commercially insured beneficiaries representing a diverse mixture of ages, racial and ethnic groups, income levels, and geographical regions across the USA.13
Study population
The study population consisted of youth (age <18 years) with T1D who first entered the study cohort between January 1, 2010 and March 31, 2020 upon first meeting Healthcare Effectiveness Data and Information Set (HEDIS) claims-based criteria for diabetes over a 12-month period, specifically: (1) any inpatient admission with an International Classification of Diseases (ICD)-9/ICD-10 diagnosis of diabetes; (2) two outpatient Evaluation and Management visits on different calendar days that include an ICD-9/ICD-10 diabetes diagnosis, or (3) a fill for at least one glucose-lowering medication other than metformin. We then required individuals to have 12 months of enrollment following this HEDIS date to be used as the observation period for determining diabetes type (to exclude youth without T1D, as previously described14 15), ensure age <18 by end of this observation period, and ascertain covariates and outcomes (online supplemental figure 1). Thus, the observation period used for this cross-sectional analysis spanned January 1, 2011 to March 31, 2021.
Utilization of pumps and CGMs
Using medical and pharmacy claims, we identified fills for insulin pumps and CGMs15 (online supplemental table 1) during the observation period.
Covariates
Patient age, sex, race/ethnicity, and annual household income were identified from OLDW enrollment files as of the end of the observation period (ie, HEDIS date +12 months). Race is classified in OLDW as a single variable with categories of white, black, Asian, Hispanic, or other/unknown. Individuals of other or unknown race/ethnicity were included in the analysis as a separate category but were excluded from the final report. Clinical information ascertained during the observation period included diabetes type, emergency department (ED) visit or hospitalization for the primary diagnosis of hyperglycemic crisis (ie, diabetic ketoacidosis, hyperglycemic hyperosmolar state),16 and clinical encounter with an endocrinologist, clinical encounter with a certified diabetes care and education specialist (CDCES).
Statistical analysis
Patient characteristics were summarized as means (SDs) and counts (percentages). Significance was determined by using either a standard t-test or χ2 test appropriately to means and frequencies. Insulin pump and CGM fill rates were reported overall and by patient race/ethnicity and annual household income. Stratifications of race/ethnicity and annual household income were also analyzed for a more detailed descriptive analysis. The Cochren-Armitage trend test was used to analyze the utilization of insulin pumps and CGM by income level while stratifying by race/ethnicity. Finally, we examined the association between patient race/ethnicity, annual household income, and odds of insulin pump or CGM fills using logistic regression, controlling for age, sex, US region, and prior ED visits/hospitalizations for hyperglycemic crises. Data management and analyses were conducted using SAS Enterprise Guide V.7.1 (SAS Institute).
Results
The study cohort was composed of 13,246 youth with T1D, with mean age 12.3 years (SD 3.7) and 52.4% were male. Among the study sample, 76.5% were white, 6.7% were black, 7.3% were Hispanic, and 2.1% were Asian (table 1). The annual household income was <US$40,000 in 8.1% of youth and ≥US$200,000 in 18.3% of youth. Overall, 4093 youth (30.9%) had a fill for an insulin pump and 4785 (36.1%) had a fill for a CGM. Children under 3 years of age were more likely to have an insulin pump than teens aged 15–17 years (50.6% vs 26.0%; p<0.001). During this time, 18.7% had one or more ED visit or hospitalization for a hyperglycemic crisis.
Table 1Clinical and demographic characteristics of the study cohort
| Overall cohort (N=13,246) | Had a pump (N=4093) | No pump (N=9153) | P value | Had a CGM (N=4785) | No CGM (N=8461) | P value | |
| Age, years, mean (SD) | 12.3 (3.7) | 11.7 (3.8) | 12.5 (3.6) | <0.0001 | 11.8 (3.8) | 12.6 (3.6) | <0.001 |
| Age category, years, N (%) | <0.0001 | <0.001 | |||||
| 1–3 | 257 (1.9%) | 130 (3.2%) | 127 (1.4%) | 145 (3.0%) | 112 (1.3%) | ||
| 4–5 | 544 (4.1%) | 188 (4.6%) | 356 (3.9%) | 246 (5.1%) | 298 (3.5%) | ||
| 6–11 | 4022 (30.4%) | 1400 (34.2%) | 2622 (28.6%) | 1613 (33.7%) | 2409 (28.5%) | ||
| 12–14 | 3962 (29.9%) | 1216 (29.7%) | 2746 (30.0%) | 1376 (28.8%) | 2586 (30.6%) | ||
| 15–17 | 4461 (33.7%) | 1159 (28.3%) | 3302 (36.1%) | 1405 (29.4%) | 3056 (36.1%) | ||
| Sex, N (%) | 0.0043 | 0.79 | |||||
| Female | 6311 (47.6%) | 2026 (49.5%) | 4285 (46.8%) | 2287 (47.8%) | 4024 (47.6%) | ||
| Male | 6935 (52.4%) | 2067 (50.5%) | 4868 (53.2%) | 2498 (52.2%) | 4437 (52.4%) | ||
| US Region, N (%) | <0.0001 | <0.001 | |||||
| Midwest | 4004 (30.2%) | 1257 (30.7%) | 2747 (30.0%) | 1406 (29.4%) | 2598 (30.7%) | ||
| Northeast | 1134 (8.6%) | 409 (10.0%) | 725 (7.9%) | 432 (9.0%) | 702 (8.3%) | ||
| South | 5392 (40.7%) | 1562 (38.2%) | 3830 (41.8%) | 1862 (38.9%) | 3530 (41.7%) | ||
| West | 2716 (20.5%) | 865 (21.1%) | 1851 (20.2%) | 1085 (22.7%) | 1631 (19.3%) | ||
| Race/ethnicity, N (%) | <0.0001 | <0.001 | |||||
| Asian | 284 (2.1%) | 88 (2.2%) | 196 (2.1%) | 102 (2.1%) | 182 (2.2%) | ||
| Black | 883 (6.7%) | 187 (4.6%) | 696 (7.6%) | 199 (4.2%) | 684 (8.1%) | ||
| Hispanic | 969 (7.3%) | 242 (5.9%) | 727 (7.9%) | 316 (6.6%) | 653 (7.7%) | ||
| White | 10 127 (76.5%) | 3233 (79.0%) | 6894 (75.3%) | 3626 (75.8%) | 6501 (76.8%) | ||
| Other/unknown | 983 (7.4%) | 343 (8.4%) | 640 (7.0%) | 542 (11.3%) | 441 (5.2%) | ||
| Annual household income, US$, N (%) | <0.0001 | <0.001 | |||||
| <US$40,000 | 1071 (8.1%) | 240 (5.9%) | 831 (9.1%) | 272 (5.7%) | 799 (9.4%) | ||
| US$40,000–US$74,999 | 2286 (17.3%) | 555 (13.6%) | 1731 (18.9%) | 640 (13.4%) | 1646 (19.5%) | ||
| US$75,000–US$124,999 | 3545 (26.8%) | 1070 (26.1%) | 2475 (27.0%) | 1184 (24.7%) | 2361 (27.9%) | ||
| US$125,000–US$199,999 | 2632 (19.9%) | 891 (21.8%) | 1741 (19.0%) | 1008 (21.1%) | 1624 (19.2%) | ||
| ≥US$200,000 | 2422 (18.3%) | 940 (23.0%) | 1482 (16.2%) | 1061 (22.2%) | 1361 (16.1%) | ||
| Other/unknown | 1290 (9.7%) | 397 (9.7%) | 893 (9.8%) | 620 (13.0%) | 670 (7.9%) | ||
| Hyperglycemic crisis | <0.0001 | <0.001 | |||||
| ≥2 | 205 (1.5%) | 52 (1.3%) | 153 (1.7%) | 68 (1.4%) | 137 (1.6%) | ||
| 1 | 2278 (17.2%) | 823 (20.1%) | 1455 (15.9%) | 932 (19.5%) | 1346 (15.9%) | ||
| None | 10 763 (81.3%) | 3218 (78.6%) | 7545 (82.4%) | 3785 (79.1%) | 6978 (82.5%) | ||
CGM, continuous glucose monitor.
Insulin pump use
Overall, 31.9% of white youth had an insulin pump fill, compared with 25.0% of Hispanic and 21.2% of black youth (p<0.001 for both). Asian youth had a higher rate of insulin pump fills (31.0%) compared with Hispanic (p=0.04) and black youth (p<0.001). Youth from high-income families (ie, those earning ≥US$200,000 annually) more often filled insulin pumps than those from low-income families (earning less than US$40,000 per year): 38.8% vs 22.4%, p<0.001. Insulin pump utilization rates increased between 2011 and 2021 across all groups except for black youths (figure 1).
Enlarge this image.Figure 1. Trends of utilization of insulin pumps and CGM among races and income levels throughout the study period. The figure illustrates an increase in utilization of CGMs and pumps throughout the study period. The x-axis represents the follow-up years, y-axis represents the percent of patients using CGM and insulin pump, respectively. For CGM use, this increase was statistically significant for all income levels and races (p<0.05 among all groups). For insulin pump use, this increase was statistically significant for all incomes and races, except for black youth (p<0.05 for all groups, except for blacks p=0.06). CGM, continuous glucose monitor.
When rates of insulin pump use were examined for each racial/ethnic group within the annual household income strata (tables 2–3; online supplemental figure 2), racial/ethnic disparities in insulin pump use were apparent primarily in youth from low-income households. Specifically, among youth from families earning less than US$40,000 per year, 25.5% of white youth used insulin pumps compared with 12.9% of black youth (p<0.001); similarly, among youth from families earning between US$40,000 and US$74,999 annually, 25.9% of white youth and 19.1% of black youth used an insulin pump (p=0.02) (table 3; online supplemental figure 2). Rates of insulin pump use were also lower among Hispanic youth from low-income families than white youth from families with similar income brackets, though the differences were statistically significant only for youth from families earning US$40,000–US$74,999 annually: 17.7% of Hispanic youth compared with 25.9% of white youth (p=0.006).
Table 2Utilization of CGMs and insulin pumps by US youth with type 1 diabetes, stratified by race/ethnicity and annual household income
| Income (US$) | Asian N=270* | Black N=799* | Hispanic N=886* | White N=9694* | ||||
| CGM | Pump | CGM | Pump | CGM | Pump | CGM | Pump | |
| <US$40 000 | NA | NA | 15.17% | 12.92% | 25.00% | 20.27% | 27.90% | 25.46% |
| US$40,000–US$74,999 | 32.50% | 30.00% | 21.16% | 19.09% | 24.57% | 17.67% | 29.45% | 25.90% |
| US$75,000–US$124,999 | 35.44% | 26.58% | 23.04% | 26.27% | 32.35% | 27.21% | 33.92% | 30.82% |
| US$125,000–US$199,999 | 37.88% | 34.85% | 29.41% | 21.57% | 42.75% | 28.99% | 38.56% | 34.73% |
| ≥US$200,000 | 50.00% | 42.59% | 47.54% | 45.90% | 48.96% | 40.63% | 43.10% | 38.26% |
| P value | 0.01 | 0.01 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 |
The p value reflects the statistical significance of the observed increase in the utilization of insulin pumps and continuous glucose monitors with rising annual household income.
* Individuals with missing income are excluded from this analysis.
*The count of patients included in the analysis, excluding those with unknown household income.
CGM, continuous glucose monitor; NA, Not reported due to cell sample size <11 individuals, per data de-identification requirement.
Table 3Differences in insulin pump utilization among racial groups, stratified by annual household income
| Race comparison | Income | % with pump group 1 | %with pump group 2 | P value |
| Group 1: Asian vs Group 2: Black | Overall | 31.0 | 21.2 | <0.001 |
| <US$40,000 | NA | 12.9 | NA | |
| US$40,000–US$74,999 | 30.0 | 19.1 | 0.11 | |
| $US$75,000–US$124,999 | 26.6 | 26.3 | 0.96 | |
| US$125,000–US$199,999 | 34.8 | 21.6 | 0.06 | |
| ≥US$200,000 | 42.6 | 45.9 | 0.72 | |
| Group 1: Asian vs Group 2: Hispanic | Overall | 31.0 | 225.0 | 0.04 |
| <US$40,000 | NA | 20.3 | NA | |
| US$40,000–US$74,999 | 30.0 | 17.7 | 0.07 | |
| US$75,000–US$124,999 | 26.6 | 27.2 | 0.91 | |
| US$125,000–US$199,999 | 34.8 | 29.0 | 0.40 | |
| ≥US$200,000 | 42.6 | 40.6 | 0.81 | |
| Group 1: Asian vs Group 2: White | Overall | 31.0 | 31.9 | 0.74 |
| <US$40,000 | NA | 25.5 | NA | |
| US$40,000–US$74,999 | 30.0 | 25.9 | 0.56 | |
| US$75,000–US$124,999 | 26.6 | 30.8 | 0.42 | |
| US$125,000–US$199,999 | 34.8 | 34.7 | 0.98 | |
| ≥US$200,000 | 42.6 | 38.3 | 0.52 | |
| Race comparison | Income | % with pump group 1 | %with pump group 2 | P value |
| Group 1: Black vs Group 2: White | Overall | 21.2 | 31.9 | <0.001 |
| <US$40,000 | 12.9 | 25.5 | <0.001 | |
| US$40,000–US$74,999 | 19.1 | 25.9 | 0.02 | |
| US$75,000–US$124,999 | 26.3 | 30.8 | 0.16 | |
| US$125,000–US$199,999 | 21.6 | 34.7 | 0.006 | |
| ≥US$200,000 | 45.9 | 38.3 | 0.23 | |
| Group 1: Black vs Group 2: Hispanic | Overall | 21.2 | 25.0 | 0.05 |
| <US$40,000 | 12.9 | 20.3 | 0.07 | |
| US$40,000–US$74,999 | 19.1 | 17.7 | 0.69 | |
| US$75,000–US$124,999 | 26.3 | 27.2 | 0.82 | |
| US$125,000–US$199,999 | 21.6 | 29.0 | 0.19 | |
| ≥US$200,000 | 45.9 | 40.6 | 0.51 | |
| Race comparison | Income | % with pump group 1 | %with pump group 2 | P value |
| Group 1: Hispanic vs Group 2: White | Overall | 25.0 | 31.9 | <0.001 |
| <US$40,000 | 20.3 | 25.5 | 0.18 | |
| US$40,000–US$74,999 | 17.7 | 25.9 | 0.006 | |
| US$75,000–US$124,999 | 27.2 | 30.8 | 0.22 | |
| US$125,000–US$199,999 | 29.0 | 34.7 | 0.17 | |
| ≥US$200,000 | 40.6 | 38.3 | 0.64 |
NA, Not reported due to cell sample size <11 individuals, per data de-identification requirement.
Importantly, the gaps in insulin pump among black and Hispanic youth compared with white youth reversed in high-income families (ie, those earning ≥US$200,000 annually), with 45.90% of black, 40.63% of Hispanic, and 38.26% of white youth having an insulin pump (table 3; online supplemental figure 2). The small number of Asian youth in the cohort limited the power for direct comparisons based on different income brackets to youth from other racial/ethnic groups. However, overall, rates of insulin pump use among Asian youth increased with increasing annual household income and were higher than among white youth for families making US$40,000–US$74 999 and ≥US$200 000 annually, but lower for families making US$75,000–US$124,999 annually.
In multivariate analysis (online supplemental table 2), odds of insulin pump utilization were lower among black (OR 0.68, 95% CI 0.57 to 0.81) and Hispanic (OR 0.79, 95% CI 0.68 to 0.93) youth compared with white youth after controlling for known variables. Odds of insulin pump use also increased progressively with higher annual household incomes, with youth from families making ≥US$200,000 annually over twice as likely to use an insulin pump than those with annual household income <US$45,000 (OR 2.07, 95% CI 1.75 to 2.45). Youth under age 15, women, and those with a history of hyperglycemic crisis were also more likely to use an insulin pump.
CGM use
White youth had a higher overall CGM utilization (35.8%) than black youth (22.5%) (p<0.001) and Hispanic youth (32.6%) (p=0.046). They, however, had a similar rate of CGM utilization compared with Asian youth (35.9%) (p=0.97). While Hispanic youth demonstrated lower CGM utilization than white youth, they had significantly greater fill rates than black youth (p<0.001). CGM utilization rates increased between 2011 and 2021 across all groups (figure 1).
Youth from high-income families (≥US$200,000) showed significantly higher CGM utilization compared with those from low-income families (<US$40,000) (43.8% vs 25.4%, p<0.001), a trend consistent across white, Hispanic, and black youths (table 2). Among income brackets below US$125,000, black youth had lower CGM utilization than white youth, with this difference becoming statistically insignificant in the US$125,000–US$199,999 income range and disappearing entirely at incomes ≥US$200,000.
Within the same income brackets, Hispanic youth had higher CGM fill rates than black youth in the <US$40,000, US$75,000–US$124,999, and US$125,000–US$199,999 income ranges. Although white youth had the highest aggregate CGM utilization, when stratified per income, no statistically significant differences were found between Hispanic and white youth within each income bracket (table 4 and online supplemental figure 3). Asian youth had higher CGM utilization than black youth within the US$75,000–US$124,999 income range, with no significant differences observed with other racial groups at other income levels when the sample size was sufficient to accurately perform the analysis.
Table 4Differences in CGM utilization among racial groups, stratified by annual household income
| Race comparison | Income | % with CGM group 1 | %with CGM group 2 | P value |
| Group 1: Asian vs Group 2: Black | Overall | 35.9 | 22.5 | <0.001 |
| <US$40 000 | NA | 15.2 | NA | |
| US$40,000–US$74,999 | 32.5 | 21.2 | 0.11 | |
| US$75,000–US$124,999 | 35.4 | 23.0 | 0.03 | |
| US$125,000–US$199,999 | 37.9 | 29.4 | 0.25 | |
| ≥US$200,000 | 50.0 | 47.5 | 0.79 | |
| Group 1: Asian vs Group 2: Hispanic | Overall | 35.9 | 32.6 | 0.29 |
| <US$40,000 | NA | 25.0 | NA | |
| US$40,000–US$74,999 | 32.5 | 24.6 | 0.29 | |
| US$75,000–US$124,999 | 35.4 | 32.4 | 0.61 | |
| US$125,000–US$199,999 | 37.9 | 42.8 | 0.51 | |
| ≥US$200,000 | 50.0 | 49.0 | 0.90 | |
| Group 1: Asian vs Group 2: White | Overall | 35.9 | 35.8 | 0.97 |
| <US$40,000 | NA | 27.9 | NA | |
| US$40,000–US$74,999 | 32.5 | 29.5 | 0.68 | |
| US$75,000–US$124,999 | 35.4 | 33.9 | 0.78 | |
| US$125,000–US$199,999 | 37.9 | 38.6 | 0.91 | |
| ≥US$200,000 | 50.0 | 43.1 | 0.31 | |
| Race comparison | Income | % with CGM group 1 | %with CGM group 2 | P value |
| Group 1: Black vs Group 2: White | Overall | 22.5 | 35.8 | <0.001 |
| <US$40,000 | 15.2 | 27.9 | <0.001 | |
| US$40,000–US$74,999 | 21.2 | 29.5 | 0.008 | |
| US$75,000–US$124,999 | 23.0 | 33.9 | 0.001 | |
| US$125,000–US$199,999 | 29.4 | 38.6 | 0.06 | |
| ≥US$200,000 | 47.5 | 43.1 | 0.49 | |
| Group 1: Black vs Group 2: Hispanic | Overall | 22.5 | 32.6 | <0.001 |
| <US$40,000 | 15.2 | 25.0 | 0.03 | |
| US$40,000–US$74,999 | 21.2 | 24.6 | 0.38 | |
| US$75,000–US$124,999 | 23.0 | 32.4 | 0.02 | |
| US$125,000–US$199,999 | 29.4 | 42.8 | 0.03 | |
| ≥US$200,000 | 47.5 | 49.0 | 0.86 | |
| Race comparison | Income | % with CGM group 1 | %with CGM group 2 | P value |
| Group 1: Hispanic vs Group 2: White | Overall | 32.6 | 35.8 | 0.047 |
| <US$40K | 25.0 | 27.9 | 0.47 | |
| US$40 k–US$74,999 | 24.6 | 29.5 | 0.12 | |
| US$75K–US$124,999 | 32.4 | 33.9 | 0.60 | |
| US$125K–US$199,999 | 42.8 | 38.6 | 0.33 | |
| ≥US$200K | 49.0 | 43.1 | 0.26 |
CGM, continuous glucose monitor; NA, Not reported due to cell sample size <11 individuals, per data de-identification requirement.
In multivariate analysis (online supplemental table 2), odds of CGM utilization continued to be lower among black youth (OR 0.60, 95% CI 0.51 to 0.71) and increased progressively with higher annual household incomes, with youth from families making ≥US$200,000 annually over twice as likely to use a CGM than those with annual household income <US$45,000 (OR 2.17, 95% CI 1.84 to 2.55). Youth under age 15 and those with a history of hyperglycemic crisis were also more likely to use a CGM.
Discussion
Both insulin pump and CGM use can improve glycemic management and quality of life among youth with T1D.17 There is robust literature demonstrating the strong association between SES and attainment of glycemic goals by youth with T1D,18 19 with generally higher glucose levels among youth from racial and ethnic minoritized backgrounds.20 21 In this nationwide analysis of diabetes technology use among commercially insured youth with T1D between 2011 and 2021, black and Hispanic youth were consistently less likely to receive insulin pump or CGM compared with white youth, but this disparity was no longer present among high-income families (those earning ≥US$200,000 per year).
In multivariate analysis, black youth were significantly less likely to use insulin pumps and CGM compared with white youth, while those with annual household incomes ≥US$200,000 were twice as likely to use these technologies than those from families earning under US$40,000 when controlling for other variables. Thus, both race/ethnicity and SES are important factors that operate independently to impact the utilization of diabetes technologies, which are increasingly regarded as the standard of care for T1D management.
Our results build on prior studies conducted in single centers and specialty group collaboratives, which have demonstrated gaps in diabetes technology use by both racial and ethnic minoritized youth and by youth from low-income households.22–24 However, those studies often had relatively small sample sizes, combined adult and pediatric populations, and included multiple insurance types.22–24 As a result, knowledge gaps remained regarding the effects of household income and racial/ethnic disparities in diabetes technology utilization, which were the primary focus areas of our study.
Within the same annual income brackets, race was a strong driver of insulin pump and CGM use among youth from families with incomes <US$75,000 annually. This is consistent with a multicenter study by Lin et al, who found a significant impact of race and annual household income on prescribing insulin pumps to youth with T1D within the first year of diagnosis.21 Our findings, therefore, underscore the important independent roles of race/ethnicity and low household income in affecting utilization of advanced diabetes care.
The reduction and eventual disappearance of racial disparities in utilization of diabetes technology at higher income levels was noted to occur gradually within racial groups, except among black youth, who experienced a substantial increase in utilization of diabetes technology within the high-income bracket (≥US$200,000 annually). Possible explanations for these gaps in technology utilization include implicit bias among healthcare professionals and systemic racism disproportionately affecting low-income black people.23 This is consistent with an earlier study conducted in adults with T1D, which detected racial disparities in the utilization of diabetes technologies identified through review of electronic health records. In that study, even after adjusting for social determinants of health, glycemic control, mental health, and diabetes outcomes, black patients remained less likely to receive information about or prescriptions for diabetes technologies.25 Another potential contributing factor to gaps in technology use is a lack of trust in healthcare systems among black individuals of lower SES.24 Wealthy black families may be better able to advocate for the receipt of evidence-based, guideline-recommended care compared with those with lower incomes.
The SEARCH for Diabetes in Youth study reported similar findings, showing a twofold increase in diabetes technology use from 2001 to 2019. However, racial, ethnic, and socioeconomic disparities persisted in that study, with no improvement over two decades. Additionally, higher SES—defined in that study by household income above US$75,000, parental college education, and private insurance—was associated with greater access to diabetes technologies.26 These data align with our findings of increasing trends in the utilization of diabetes technologies over the years across all racial groups and income levels, with the exception of insulin pump use among black youth that did not see such an increase.
While all youth included in our study were privately insured, household income affects the ability to afford copayments and deductibles associated with healthcare services. Despite having health insurance, lower-income families may find themselves forced to ration their utilization of healthcare, including diabetes technologies. Further studies are needed to fully explore the causes of limited utilization of diabetes technology among low-income families, as well as ways to mitigate these barriers.
To our knowledge, this is the first nationwide study using claims data for insulin pump and CGM therapy from a large private insurance provider. By focusing on insured youth, we eliminated the impact of insurance access on the ability to access diabetes care and to obtain and use diabetes technologies, enabling us to more directly assess the effect of household income (independent of insurance coverage) and to examine racial and ethnic disparities in care potentially driven by systemic biases and racism. By including a national cohort of youth across a long time frame, we sought to maximize the sample size and enable subgroup analyses stratified by race/ethnicity and income that were previously infeasible.
Nevertheless, our findings must be considered in the context of the study’s limitations. The use of claims data precludes us from examining the potential reasons for our findings, including the attitudes of the patients, families, and healthcare teams toward diabetes technology; differences in the availability of diabetes specialists with expertise in technology use; and clinicians’ comfort levels in prescribing and managing these technologies. The study’s timeline spans over a decade—a necessity to ensure the largest possible sample size to conduct the subgroup analyses—but this time frame also spans a period in which substantial advancements in diabetes technology (eg, the hybrid closed-loop system was introduced to the market in 2016) and changes in clinicians’ and patients’ understanding of the importance of these technologies in diabetes management. Other studies indicate that although insulin pump usage increased over time between 2005 and 2019, disparities in prescribing patterns and utilization persisted.26 These disparities were characterized by higher usage among individuals with private insurance, household incomes exceeding US$100,000, and those identifying as non-Hispanic white.21
Our findings underscore the importance of targeted interventions aimed at the equitable distribution of diabetes technology across all societal groups, rather than focusing solely on expanding diabetes technology without attention to equitable reach. Data from the T1D Exchange (T1DeX) Clinic Registry demonstrated that utilization of diabetes technology can mitigate, though not entirely eliminate, the negative association between higher A1C levels, lower SES (defined in that study as household income below US$50,000), and African American race within a mixed-age cohort receiving care in the T1DeX collaborative network.27 Addressing disparities in the utilization of evidence-based and guideline-recommended technology to all youth with T1D must be a priority, with specific attention to youth from low-income racially minoritized backgrounds who experience the greatest barriers to care.
Multilevel interventions for better and more equitable access to insulin pump and CGM technology are therefore needed to alleviate racial/ethnic and income-based disparities in their utilization. In 2024, the US Internal Revenue Service classified CGMs as “preventive” treatments for people with diabetes, thereby allowing for their inclusion on preventive drug lists implemented by commercial payors and to be covered by insurance even before the annual deductible is met.28 The impact of this important policy change will need to be examined. Insulin pumps are not classified this way, however, with persistent financial barriers to accessing this essential technology. Health systems also need to invest in culturally sensitive and tailored educational interventions to enhance utilization of diabetes technology. Increasing the availability of CDCES in underserved areas and fostering individuals from diverse backgrounds to enter the CDCES and endocrinology workforce may also help eliminate bias and improve access to care. Institutional initiatives and societal campaigns are needed to raise awareness, educate, and build trust in the effectiveness of advanced diabetes technology across diverse racial, ethnic, and socioeconomic groups. Finally, empowering youth and families with the information necessary for shared decision-making and equipping them with skills to manage diabetes technology effectively are also vital for successful diabetes management.17
Conclusions
In a large national cohort of commercially insured youth with T1D, both race and income significantly influenced utilization of diabetes technology. Utilization of insulin pumps and CGMs increased with higher household income, and racial disparities in usage were ameliorated with higher household incomes. Potential barriers contributing to this disparity include healthcare professionals’ biases, lower trust in the healthcare system among individuals from lower SES, and gaps in insurance coverage, knowledge, and technical skills. Addressing these disparities requires coordinated efforts at the individual, community, institutional, and governmental levels. Promising solutions include expanding research, raising awareness, enhancing education, increasing access to outreach specialty clinics, and expanding insurance coverage. Further investigation and targeted interventions are essential to reduce these inequities in diabetes technology access.
Data availability statement
Data may be obtained from a third party and are not publicly available. This study was conducted using deidentified data from Optum Labs Data Warehouse and linked 100% sample of Medicare fee-for-service claims. These data are third party data owned by Optum Labs and contain sensitive patient information; therefore, the data are only available on request. Interested researchers engaged in HIPAA compliant research may contact [email protected] for data access requests. The data use requires researchers to pay for rights to use and access the data. These data are subject to restrictions on sharing as a condition of access.
Ethics statements
Patient consent for publication
Not applicable.
Ethics approval
This study involves human participants. This study was exempt from review by the Mayo Clinic Institution Review Board (IRB) because OLDW data are deidentified consistent with HIPAA expert deidentification determination.
X @@RozalinaMD
Contributors AAN interpreted study results, drafted the initial manuscript, and participated in its review and revision. DH drafted the initial manuscript and participated in its review and revision. TR participated in the review and revision of the manuscipt. HCH conducted data analysis and participated in the review and revision of the manuscript. RGM conceptualized the study, secured funding, interpreted the results, reviewed and revised the manuscript, and supervised the study. All authors approved the final manuscript. RGM is the guarantor of this work and accepts full responsibility for the finished work and the conduct of the study, had access to the data, and controlled the decision to publish.
Funding This effort was funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health (NIH) grant number K23DK114497. RGM is an investigator at the University of Maryland-Institute for Health Computing, which is supported by funding from Montgomery County, Maryland, and the University of Maryland Strategic Partnership: MPowering the State, a formal collaboration between the University of Maryland, College Park, and the University of Maryland, Baltimore.
Disclaimer Study contents are the sole responsibility of the authors and do not necessarily represent the official views of the NIH.
Competing interests In the last 36 months, RGM has received unrelated research support from NIDDK of the NIH, the National Institute on Aging (NIA) of the NIH, the Patient Centered Outcomes Research Institute (PCORI), National Center for Advancing Translational Sciences (NCATS), and the American Diabetes Association. She serves as a consultant to Emmi Educate (Wolters Kluwer) and the Yale-New Haven Health System’s Center for Outcomes Research and Evaluation and has received speaking honoraria and travel support from the American Diabetes Association. Other authors have no completing interests to disclose.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
1 Patterson CC, Gyürüs E, Rosenbauer J, et al. Trends in childhood type 1 diabetes incidence in Europe during 1989-2008: evidence of non-uniformity over time in rates of increase. Diabetologia 2012; 55: 2142–7. doi:10.1007/s00125-012-2571-8
2 Green A, Hede SM, Patterson CC, et al. Type 1 diabetes in 2017: global estimates of incident and prevalent cases in children and adults. Diabetologia 2021; 64: 2741–50. doi:10.1007/s00125-021-05571-8
3 Lawrence JM, Divers J, Isom S, et al. Trends in Prevalence of Type 1 and Type 2 Diabetes in Children and Adolescents in the US, 2001-2017. JAMA 2021; 326: 717–27. doi:10.1001/jama.2021.11165
4 Nathan DM, DCCT/EDIC Research Group. The diabetes control and complications trial/epidemiology of diabetes interventions and complications study at 30 years: overview. Diabetes Care 2014; 37: 9–16. doi:10.2337/dc13-2112
5 Cengiz E, Danne T, Ahmad T, et al. ISPAD Clinical Practice Consensus Guidelines 2022: Insulin treatment in children and adolescents with diabetes. Pediatr Diabetes 2022; 23: 1277–96. doi:10.1111/pedi.13442
6 American Diabetes Association Professional Practice Committee. 14. Children and Adolescents: Standards of Care in Diabetes-2024. Diabetes Care 2024; 47: S258–81. doi:10.2337/dc24-S014
7 Benioudakis E, Karlafti E, Kalaitzaki A, et al. Technological Developments and Quality of Life in Type 1 Diabetes Mellitus Patients: A Review of the Modern Insulin Analogues, Continuous Glucose Monitoring and Insulin Pump Therapy. Curr Diabetes Rev 2022; 18: e031121197657. doi:10.2174/1573399818666211103163208
8 Sherr JL, Schoelwer M, Dos Santos TJ, et al. ISPAD Clinical Practice Consensus Guidelines 2022: Diabetes technologies: Insulin delivery. Pediatr Diabetes 2022; 23: 1406–31. doi:10.1111/pedi.13421
9 Tauschmann M, Forlenza G, Hood K, et al. ISPAD Clinical Practice Consensus Guidelines 2022: Diabetes technologies: Glucose monitoring. Pediatr Diabetes 2022; 23: 1390–405. doi:10.1111/pedi.13451
10 Fantasia KL, Wirunsawanya K, Lee C, et al. Racial Disparities in Diabetes Technology Use and Outcomes in Type 1 Diabetes in a Safety-Net Hospital. J Diabetes Sci Technol 2021; 15: 1010–7. doi:10.1177/1932296821995810
11 Singh P, Grishman E, Naranjo D, et al. 1018-P: Disparities in Diabetes Technology Use in Youth with Diabetes. Diabetes 2023; 72. doi:10.2337/db23-1018-P
12 Addala A, Auzanneau M, Miller K, et al. A Decade of Disparities in Diabetes Technology Use and HbA 1c in Pediatric Type 1 Diabetes: A Transatlantic Comparison. Diabetes Care 2021; 44: 133–40. doi:10.2337/dc20-0257
13 Wallace PJ, Shah ND, Dennen T, et al. Optum Labs: building a novel node in the learning health care system. Health Aff (Millwood) 2014; 33: 1187–94. doi:10.1377/hlthaff.2014.0038
14 McCoy RG, Lipska KJ, Van Houten HK, et al. Association of Cumulative Multimorbidity, Glycemic Control, and Medication Use With Hypoglycemia-Related Emergency Department Visits and Hospitalizations Among Adults With Diabetes. JAMA Netw Open 2020; 3: e1919099. doi:10.1001/jamanetworkopen.2019.19099
15 Benning TJ, Heien HC, Herges JR, et al. Glucagon fill rates and cost among children and adolescents with type 1 diabetes in the United States, 2011–2021. Diabetes Res Clin Pract 2023; 206: 111026. doi:10.1016/j.diabres.2023.111026
16 McCoy RG, Galindo RJ, Swarna KS, et al. Sociodemographic, Clinical, and Treatment-Related Factors Associated With Hyperglycemic Crises Among Adults With Type 1 or Type 2 Diabetes in the US From 2014 to 2020. JAMA Netw Open 2021; 4: e2123471. doi:10.1001/jamanetworkopen.2021.23471
17 Addala A. Making a Good Thing Even Better: Expanding Access and Applicability of Automated Insulin Delivery Systems to Benefit All Youth With Type 1 Diabetes. Diabetes Care 2023; 46: 2126–8. doi:10.2337/dci23-0057
18 Ebekozien O, Mungmode A, Sanchez J, et al. Longitudinal Trends in Glycemic Outcomes and Technology Use for Over 48,000 People with Type 1 Diabetes (2016-2022) from the T1D Exchange Quality Improvement Collaborative. Diabetes Technol Ther 2023; 25: 765–73. doi:10.1089/dia.2023.0320
19 Majidi S, Ebekozien O, Noor N, et al. Inequities in Health Outcomes in Children and Adults With Type 1 Diabetes: Data From the T1D Exchange Quality Improvement Collaborative. Clin Diabetes 2021; 39: 278–83. doi:10.2337/cd21-0028
20 Demeterco-Berggren C, Ebekozien O, Noor N, et al. Factors Associated With Achieving Target A1C in Children and Adolescents With Type 1 Diabetes: Findings From the T1D Exchange Quality Improvement Collaborative. Clin Diabetes 2022; 41: 68–75. doi:10.2337/cd22-0073
21 Lin MH, Connor CG, Ruedy KJ, et al. Race, socioeconomic status, and treatment center are associated with insulin pump therapy in youth in the first year following diagnosis of type 1 diabetes. Diabetes Technol Ther 2013; 15: 929–34. doi:10.1089/dia.2013.0132
22 Walker AF, Haller MJ, Addala A, et al. Not all healthcare inequities in diabetes are equal: a comparison of two medically underserved cohorts. BMJ Open Diabetes Res Care 2024; 12: e004229. doi:10.1136/bmjdrc-2024-004229
23 Odugbesan O, Addala A, Nelson G, et al. Implicit Racial-Ethnic and Insurance-Mediated Bias to Recommending Diabetes Technology: Insights from T1D Exchange Multicenter Pediatric and Adult Diabetes Provider Cohort. Diabetes Technol Ther 2022; 24: 619–27. doi:10.1089/dia.2022.0042
24 Agarwal S, Schechter C, Gonzalez J, et al. Racial-Ethnic Disparities in Diabetes Technology use Among Young Adults with Type 1 Diabetes. Diabetes Technol Ther 2021; 23: 306–13. doi:10.1089/dia.2020.0338
25 Kanbour S, Jones M, Abusamaan MS, et al. Racial Disparities in Access and Use of Diabetes Technology Among Adult Patients With Type 1 Diabetes in a U.S. Academic Medical Center. Diabetes Care 2023; 46: 56–64. doi:10.2337/dc22-1055
26 Everett EM, Wright D, Williams A, et al. A Longitudinal View of Disparities in Insulin Pump Use Among Youth with Type 1 Diabetes: The SEARCH for Diabetes in Youth Study. Diabetes Technol Ther 2023; 25: 131–9. doi:10.1089/dia.2022.0340
27 Miller KM, Beck RW, Foster NC, et al. HbA1c Levels in Type 1 Diabetes from Early Childhood to Older Adults: A Deeper Dive into the Influence of Technology and Socioeconomic Status on HbA1c in the T1D Exchange Clinic Registry Findings. Diabetes Technol Ther 2020; 22: 645–50. doi:10.1089/dia.2019.0393
28 Preventive care for purposes of qualifying as a high deductible health plan under section 223. notice 2024-75. United States Internal Revenue Service; 2024. Available: https://www.irs.gov/pub/irs-drop/n-24-75.pdf [Accessed 29 Apr 2025 ].
© 2025 Author(s) (or their employer(s)) 2025. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ Group. http://creativecommons.org/licenses/by-nc/4.0/ This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ . Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.