Association Between Health Insurance Type and Adverse Outcomes for Children and Young Adults With Type 1 Diabetes and Coronavirus Disease 2019

The T1DX-QI sponsored a secondary analysis of a cross-sectional study of 52 adult and pediatric endocrinology clinics across the United States (12). The Western Institutional Review Board deemed the study as exempt. For this study, COVID-19 status occurred before hospital admission and was confirmed with a positive COVID-19 test. Hospitalization was defined as patients admitted to an inpatient unit of a hospital. Both DKA and hypoglycemia were defined using International Society for Pediatric and Adolescent Diabetes definitions (13,14). Each participating site conducted simple chart reviews and submitted data using online survey software via Qualtrics between April 2020 and October 2021 to identify patients with established type 1 diabetes and COVID-19. Details of the questionnaire (Supplementary Material) have been previously reported (15).

Patient outcomes and demographics were analyzed as categorical variables when appropriate. Race/ethnicity was categorized as Non-Hispanic White, Non-Hispanic Black, Hispanic, or other (16). Health insurance type was classified as public, private, or unknown. Use of an insulin pump and use of continuous glucose monitoring (CGM) were analyzed as categorical (yes/no) variables. Patients’ age, limited to <24 years, and last reported A1C values were collected and analyzed as continuous variables, using mean ± SD or median (interquartile range [IQR]) as appropriate based on data distribution. P values were calculated using Fisher exact or χ² tests to examine the association between the categorical variables.

A post hoc power analysis was performed using G*Power to determine the minimum sample size required to achieve a sufficient level of power needed to avoid a type II error. Power calculations demonstrated that a minimum sample size of 210 was needed to achieve a power of 0.95 for an effect size of 0.5. Our study included 619 individuals in the analysis.

Data for a total of 619 individuals were analyzed. Selected patient characteristics are displayed in Table 1. Publicly insured patients were more likely to be younger and female compared with privately insured patients (14.8 vs. 16.3 years of age [P >0.001] and 57 vs. 48% female [P = 0.02]). There was a statistically significant difference in reported race/ethnicity; fewer patients with public insurance identified as non-Hispanic White (35 vs. 81%, P <0.001) and more identified as non-Hispanic Black (23 vs. 5%, P <0.001) or Hispanic (36 vs. 9%, P <0.001). Median A1C values were higher in those with public insurance (9.5 vs. 7.9%, P <0.001). There were fewer publicly insured patients using diabetes technology such as insulin pumps (30 vs. 62%, P <0.001) and CGM (52 vs. 78%, P <0.001) for their baseline diabetes management compared with those with private insurance.

For individuals with public versus private insurance, COVID-19 was associated with higher rates of hospitalization (P <0.001) (Figure 1) and adverse acute diabetes-related outcomes (P <0.001) (Figure 2). Although individuals with public insurance were three times more likely to be hospitalized than those with private insurance (odds ratio [OR] 3.4, 95% CI 2.1–5.4), the odds were reduced and no longer significant after adjusting for baseline A1C (OR 1.7, 95% CI 0.99–2.85). Of the 257 individuals with public insurance, 53 (21%) presented in DKA and 4 (2%) presented with severe hypoglycemia. In comparison, of the 362 individuals with private insurance, 21 (6%) presented in DKA and 4 (1%) had severe hypoglycemia. Rates of adverse outcomes were also higher among publicly insured individuals when controlling for race/ethnicity (10% publicly insured non-Hispanic White vs. 6% privately insured; 40% publicly insured non-Hispanic Black vs. 24% privately insured; 20% publicly insured Hispanic vs. 6% privately insured, P <0.001) (Figure 3). The odds for adverse outcomes were increased among individuals on public insurance (OR 3.6, 95% CI 2.17–6.03). After adjusting for baseline A1C, the odds were still significant (OR 1.8, 95% CI 1.01–3.15, P = 0.04) (Tables 2 and 3).

This secondary analysis revealed that publicly insured children and young adults with type 1 diabetes and COVID-19 were more likely to be hospitalized and experience adverse outcomes. This higher rate of adverse outcomes was correlated with public insurance status regardless of reported race/ethnicity. The increased odds were persistent even after adjusting for baseline A1C. Our findings are consistent with previous studies that have demonstrated socioeconomic inequities contributing to adverse outcomes such as DKA among individuals with type 1 diabetes and COVID-19 (15,17–19). Differential access to health care is a possible explanation for the higher rate of adverse outcomes.

At baseline, those with public insurance had higher median (9.5 vs. 7.9%) and mean (10 vs. 8.3%) A1C values compared with those with private insurance. One possible explanation is decreased access to diabetes technology, which has been associated with fewer adverse outcomes (20). Within our study population, 52% of publicly insured patients were using CGM and 30% were using an insulin pump compared with 78% of privately insured patients using CGM and 62% using an insulin pump. Access to diabetes technology, which can be affected by insurance type (21), may be another contributing factor to the differences in diabetes outcomes found in this study.

Our findings expand and are consistent with earlier work by Alonso et al. (15) demonstrating that baseline A1C was a major contributor to hospitalization. However, because of an extended data collection period, this secondary analysis had an updated dataset with more people, new COVID-19 variants, and a broader age range (0–25 years). This study adds to the literature crucial information on the impact of public insurance status on adverse outcomes, which was still a significant determinant despite adjusting for baseline A1C. One possible explanation for this finding is that, in the later stage of the COVID-19 pandemic, there were fewer hospitalizations because more people could access virtual care for adverse outcomes.

For at least 2 decades, evidence has linked insurance to health outcomes (10,11). Racial/ethnic minority and low-income people with type 1 diabetes are more likely to be on public health insurance. An important contribution of this study is its call for a stronger and more effective public insurance system, especially because this system serves as a lifeline for some of our nation’s most vulnerable people.

A limitation to our study is that it was cross-sectional and cannot establish causality. Hence, the associations between DKA and public insurance are strictly correlational. Data were collected via questionnaire, and there may have been underreporting of COVID-19 cases. It would be helpful to know the baseline insurance status of the study sample because type of insurance can have an impact on type of treatments covered. This study also did not include data on baseline health status or comorbidities between insurance groups.

In this study, rates of hospitalization and adverse outcomes were higher in individuals <24 years of age with type 1 diabetes and COVID-19 who had public versus private health insurance, across races/ethnicities. There is a need for further research that explores the impact of health insurance on quality of care, health care utilization, and outcomes.

O.E. is the principal investigator for research projects funded through his institution by Dexcom, Eli Lilly, and Medtronic; he is also a member of the Medtronic Diabetes Health Inequity Advisory Board. No other potential conflicts of interest relevant to this article were reported.

B.M. and J.K.D. wrote the manuscript. O.E. researched data and contributed to the discussion. S.R. researched and analyzed data and contributed to the methodology. A.O., I.G.-B., A.M., A.G., V.R., and A.S. contributed to discussion and reviewed/edited the manuscript. S.R. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.