Evaluation of the Cascade of Diabetes Care in the United States, 2005-2016

Abstract

Importance: Treatment advances in diabetes can meaningfully improve outcomes only if they effectively reach the populations at risk.

Objectives: To evaluate whether the cascade of US diabetes care, defined as diabetes diagnosis, linkage to care, and achievement of individual and combined treatment targets, improved from 2005 to 2016 and to investigate potential disparities in US diabetes care.

Design, setting, and participants: Nationally representative, serial cross-sectional studies included in the 2005-2016 National Health and Nutrition Examination Survey were evaluated. Data on nonpregnant US adults (age ≥18 years) with diabetes who had reported fasting for 9 or more hours (n = 1742 diagnosed and 746 undiagnosed) were included. Data analysis was performed from August 1, 2018, to May 10, 2019.

Exposures: Time period (2005-2008, 2009-2012, and 2013-2016), age, sex, race/ethnicity, health insurance, and educational level incorporated into logistic regression models predicting odds of target achievement.

Main outcomes and measures: Proportion of participants overall and stratified by age, sex, and race/ethnicity who were linked to diabetes care and met glycemic (hemoglobin A1c <7.0%-8.5%, depending on age and complications), blood pressure (<140/90 mm Hg), cholesterol level (low-density lipoprotein cholesterol <100 mg/dL), and smoking abstinence targets and a composite of all targets.

Results: In 2013-2016, of 1742 US adults with diagnosed diabetes, 94% (95% CI, 92%-96%) were linked to diabetes care; 64% (95% CI, 58%-69%) met hemoglobin A1c level, 70% (95% CI, 64%-75%) met blood pressure level, and 57% (95% CI, 51%-62%) met cholesterol level targets; 85% were nonsmokers (95% CI, 82%-88%); and 23% (95% CI, 17%-29%) achieved the composite goal. Results were similar in 2005-2008 (composite 23%) and in 2009-2012 (composite 25%). There was no significant improvement in diagnosis or target achievement during the study period. Compared with middle-aged adults (45-64 years) with diagnosed diabetes, older patients (≥65 years) had higher odds (adjusted odds ratio [aOR], 1.70; 95% CI, 1.17-2.48) and younger adults (18-44 years) had lower odds (aOR, 0.53; 95% CI, 0.29-0.97) of meeting the composite target. Women had lower odds of achieving the composite target than men (aOR, 0.60; 95% CI, 0.45-0.80). Non-Hispanic black individuals vs non-Hispanic white individuals had lower odds of achieving the composite target (aOR, 0.57; 95% CI, 0.39-0.83). Having health insurance was the strongest predictor of linkage to diabetes care (aOR, 3.96; 95% CI, 2.34-6.69).

Conclusions and relevance: It appears that the diabetes care cascade in the United States has not significantly improved between 2005 and 2016. This study's findings suggest that gaps in diabetes care that were present in 2005, particularly among younger adults (18-44 years), women, and nonwhite individuals, persist.

Figure 1.. Cascade of Diabetes Care in the United States for 2005-2016

The cascade of diabetes care (also known as the diabetes care continuum) for US nonpregnant adults with diagnosed diabetes in 2005-2008, 2009-2012, and 2013-2016. The cascade of care illustrates the population-level steps of linkage to care, meeting individual treatment targets, and attaining the combined targets of care, and helps reveal the gaps in diabetes care. Glycemic control was defined as hemoglobin A_1c (HbA_1c) level less than or equal to an individualized target. Blood pressure (BP) control was defined as systolic/diastolic BP less than 140/90 mm Hg. Cholesterol control was defined as low-density lipoprotein cholesterol level less than 100 mg/dL (to convert to millimoles per liter, multiply by 0.0259). BC control refers to BP and cholesterol level control; ABC control refers to HbA_1c, BP, and cholesterol level control; and ABCN control was defined as HbA_1c, BP, and cholesterol level control and not smoking tobacco. Error bars indicate 95% CI.

Figure 2.. Cascade of Diabetes Care in the United States for 2005-2016 by Age

Cascade of diabetes care for US nonpregnant adults with diagnosed diabetes by age, shown as diagnosed diabetes in the base cohort linked to diabetes care and glycemic control (A), blood pressure (BP) and cholesterol control in the nonsmoking cohort (B), and combined control (C). Glycemic control was defined as HbA_1c less than or equal to an individualized target. Blood pressure control was defined as systolic/diastolic BP less than 140/90 mm Hg. Cholesterol control was defined as low-density lipoprotein cholesterol level less than 100 mg/dL (to convert to millimoles per liter, multiply by 0.0259). BC control refers to BP and cholesterol level control; ABC control refers to HbA_1c, BP, and cholesterol level control; and ABCN control was defined as HbA_1c, BP, and cholesterol level control and not smoking tobacco. Error bars indicate 95% CI.

Figure 3.. Cascade of Diabetes Care in the United States for 2005-2016 by Race/Ethnicity

Cascade of diabetes care for US nonpregnant adults with diagnosed diabetes by race/ethnicity, shown as diagnosed diabetes in the base cohort linked to diabetes care and glycemic control (A), blood pressure (BP) and cholesterol control in the nonsmoking cohort (B), and combined control (C). Glycemic control was defined as hemoglobin A_1c (HbA_1c) level less than or equal to an individualized target. Blood pressure (BP) control was defined as systolic/diastolic BP less than 140/90 mm Hg. Cholesterol control was defined as low-density lipoprotein cholesterol level less than 100 mg/dL (to convert to millimoles per liter, multiply by 0.0259). BC control refers to BP and cholesterol level control; ABC control refers to HbA_1c, BP, and cholesterol level control; and ABCN control was defined as HbA_1c, BP, and cholesterol level control and not smoking tobacco. Error bars indicate 95% CI.