Abstract
In 2014, an estimated 30.3 million people (9.4%) in the United States had diabetes.1 Improving glycemic control reduces the risk of diabetes-related vascular complications.2 Studies have shown improvement in glycemic control in the United States from 1998 through 2010, as measured by hemoglobin A1c (HbA1c).3,4 We examined recent trends in glycemic control and patient awareness of HbA1c test results and targets.
Methods
The study was exempted from review by the institutional review board at Tufts Medical Center. We used data from 4 survey periods of the National Health and Nutrition Examination Survey (NHANES; survey periods: 2007-2008, 2009-2010, 2011-2012, and 2013-2014), a representative sample of the US noninstitutionalized population with average response rates of 75.4% for the study period.5 A subsample of surveyed participants was asked to complete laboratory tests and physical examinations. Respondents 20 years and older who reported having diabetes diagnosed by a clinician were included in this analysis.
We calculated the proportion of survey participants with an HbA1c level less than 7% (good control), less than 8% (moderate control), and greater than 9% (poor control). We also calculated the proportion of participants who reported having a test for HbA1c in the prior year and who were aware of their HbA1c result and their target HbA1c level. We analyzed results by subgroups of age (20-44, 45-65, and >65 years), race (white vs nonwhite), and sex. Survey weights were used to calculate proportions. P values for trend over the 4 survey periods were calculated using a multilevel logistic regression in STATA (StataCorp), version 14. Age, sex, body mass index, race, insurance status, education, and income3 were included as covariates of the first level, and survey period as the dependent variable of the second level. No covariates were included for the subgroup trend analyses. Significance was defined as a 2-tailed P value less than .05.
Results
Of 23 482 participants, 2908 (12.40%) reported having diabetes (mean age, 62 years [SD, 12.96]; women, 50.40%; white race, 35.80% [unweighted]). Median HbA1c level was 6.90% (interquartile range [IQR], 6.20%-8.15%) in 2007-2008 and 6.95% (IQR, 6.30%-8.20%) in 2013-2014 (P = .39) (Table 1). Glycemic control did not change between 2007-2008 and 2013-2014 overall or in any subgroup (Table 1). Over time, a higher proportion of patients with diabetes reported having an HbA1c test within the past year; 55.10% (95% CI, 49.69%-60.50%) in 2007-2008 to 77.78% (95% CI, 75.07%-80.48%) in 2013-2014 (P value for trend <.001) (Table 2). All subgroups showed a similar change. In 2007-2008, 52.32% (95% CI, 47.63%-57.00%) of patients reported being aware of their past year HbA1c result. This proportion increased to 74.31% (95% CI, 69.63%-78.99%) in the 2013-2014 survey (P value for trend <.001). The change was statistically significant in all subgroups except participants aged 20 to 44 years (P value for trend = .16). The proportion of participants who were aware of the target HbA1c level set by their clinician increased from 2007-2008 (74.07% [95% CI, 70.48%-77.66%]) to 2013-2014 (89.70% [95% CI, 86.11%-93.29%]) (P value for trend <.001). The change was statistically significant in all subgroups except participants aged 20 to 44 years (P value for trend = .81) (Table 2).