Antiplatelets and statins are efficacious for preventing future cardiovascular events in patients with coronary heart disease. Disparity in cardiovascular outcomes exists by race/ethnicity and gender; however, few studies have explored potential disparities in long-term antiplatelet and statin use by race/ethnicity and gender. We conducted a repeated cross-sectional analysis using the nationally representative Medical Expenditure Panel Survey from 2003 to 2012. The sample consisted of 14,334 men and women >29 years with coronary heart disease. We identified antiplatelet and statin use, medical conditions, and sociodemographic characteristics. Rates of use did not change for statins or the combination of statins and antiplatelets from 2003 to 2012 but decreased for antiplatelets (p = 0.015). Of the total sample, 70.9% (95% confidence interval [CI] 69.7 to 72.1) reported use of antiplatelets, 52.5% (95% CI 51.1 to 53.8) reported statin use, and 43.1% (95% CI 41.8 to 44.4) reported the combination. Use of antiplatelets and statins were associated with one another (odds ratio 3.22; 95% CI 2.87 to 3.62). From 2009 to 2012, black and Hispanic men along with all race/ethnicities of women were less likely to report use of statins, antiplatelets, and the combination of the 2 compared with white men, even after controlling for sociodemographics. Changing the definition of a medication use, inclusion of cardiovascular risk factors, or the inclusion of warfarin in the antiplatelet category did not substantially change the results. Future practice and policy goals should focus on increasing the number of high-risk patients on appropriate preventative medications while focusing particular attention on decreasing the identified disparity.
Race/ethnicity and gender disparities have previously been identified in cardiovascular care and coronary heart disease (CHD) outcomes and long-term medication use. The purpose of this article is to further explore race/ethnic and gender disparities in the use of antiplatelet and statin medication in a high-risk sample with known CHD. This analysis furthers the literature by investigating long-term outpatient use of statins and antiplatelets in a contemporary nationally representative sample. We use a sample of 10 years of the nationally representative Medical Expenditure Panel Survey (MEPS) data to delineate the use of antiplatelets, statins, and their combination by race/ethnicity and gender while controlling for the traditional confounding factors.
Methods
We conducted a repeated cross-sectional analysis using the nationally representative 2003 to 2012 MEPS. The survey is produced by the Agency for Health Research and Quality and consists of 5 interviews for >2 years and has 2 concurrent panels. The interviewee is the member of the house who is most knowledgeable about family’s medical problems. The survey is designed to produce nationally representative estimates for each separate year. The MEPS includes demographics, prescription drugs, insurance coverage, health care expenditures, and self-reported medical conditions.
The MEPS includes around 15,000 households in each survey year and is derived from a sample of the previous year’s National Health Interview Survey. This analysis included all adults >29 years with CHD. This age cutoff was selected because of low levels of CHD at younger ages and concern for false-positive reporting of CHD. Exclusion criteria included those ineligible for statin therapy (pregnancy or severe liver disease). Patients were asked if they were not taking aspirin because of a medical reason. We did not exclude patients who reported being unable to take aspirin in the main analysis given the availability of other antiplatelet options (i.e., clopidogrel) and risk reduction agents (i.e., proton pump inhibitors).
The primary outcomes were the use of antiplatelets, statins, or their combination (antiplatelets and statins). Prescription drug information was collected and verified with pharmacy data. Prescription drug fills in MEPS for long-term medications have been found to be valid and not biased by sociodemographic variables, such as race/ethnicity and gender. We quantified the use of statins and 5 antiplatelet drugs (aspirin, dipyridamole/aspirin, clopidogrel, ticlopidine, and prasugrel). A medication user was classified as someone who reported >179 tablets (>359 for dipyridamole/aspirin) or >2 medication fills within a class of medications over a 1-year period. Warfarin data were also quantified for sensitivity analysis, which a user was classified if they reported >2 prescriptions. Dipyridamole/aspirin was included given the aspirin component of the medication. Given that aspirin and other over the counter drugs are under-reported within the prescription data, we also included respondents as antiplatelet users who reported aspirin use (daily/every other day) over the past year when directly questioned.
CHD was identified by subjects who reported a history of coronary artery disease or myocardial infarction. There was a change in survey methodology starting in 2007, which increased the number of times a subject was asked about diagnoses from once a year to at every interview. Cardiovascular risk conditions (hyperlipidemia, hypertension, diabetes, and tobacco use) were identified by direct questioning regarding these conditions or in association with a medical expense.
For the descriptive longitudinal analysis, survey years were grouped into 2-year intervals. Adjusted Wald tests, chi-square statistics, and logistic regression were used to determine statistical significance in the bivariate analysis. Simple or multivariable logistic regression with year as a continuous independent variable was used to determine trends.
A multivariable logistic regression using data from 2009 to 2012 regressed antiplatelet, statin, and combined antiplatelet and statin use in patients with CHD on age, gender, insurance status (any private, public, or uninsured), race/ethnicity (white, black, Hispanic, or other), region of residence (northeast, midwest, south, or west), educational attainment (no high school, high school/General Education Development, or more than high school), metropolitan resident, and poverty category (poor/near poor, low/middle, or high). We present post-prediction average marginal effects as a proportion and 95% confidence interval (CI). We tested for statistical interaction between race/ethnicity and gender in the multivariable logistic regression models. A p value <0.05 was considered statistically significant for all analyses.
Numerous sensitivity analyses were conducted. First, a medication user was changed to any statin and any antiplatelet or an inability to take aspirin. Second, warfarin users were qualified as an antiplatelet user. Third, hypertension, diabetes, tobacco use, and hyperlipidemia were included in the multivariable logistic regression. Complex survey weighting was for all analyses (STATA, version 13; STATA Corporation, College Station, Texas). The study was judged to be exempt by The Ohio State University Institutional Review Board.
Results
The sample included 185,627 adults >29 years, of which 2,865 met exclusion criteria. In total, 14,334 subjects who reported CHD and met inclusion criteria. The number of patients who reported CHD increased from 6.6% (95% CI 6.2 to 7.0) in 2003 to 2004 to 9.3% (95% CI 8.8 to 9.9) in 2011 to 2012 (p <0.001). There was an increase in the portion of patients reporting CHD in 2007 that coincided with changes in survey methodology. Whites had the highest level of reported CHD (9.1%; 95% CI 8.8 to 9.4) compared with blacks (7.1%; 95% CI 6.6 to 7.6) and Hispanics (4.9%; 95% CI 4.5 to 5.4; p <0.001 for all). Men reported higher levels of CHD (9.9%; 95% CI 9.5 to 10.3) compared with women (6.5%; 95% CI 6.2 to 6.9; p <0.001). Of the sample with CHD, 70.9% (95% CI 69.7 to 72.1) reported use of antiplatelets. Aspirin use was reported by 67.6% (95% CI 66.3 to 68.8) of patients with CHD in the sample, whereas 16.7% (95% CI 15.8 to 17.7) reported clopidogrel/ticlopidine/prasugrel use. Statin use was reported by 52.6% (95% CI 51.2 to 53.9) of the patients with CHD, whereas the combination of antiplatelets and statins was reported by 43.1% (95% CI 41.8 to 44.4). Use of antiplatelets and statins were associated with one another (odds ratio 3.22; 95% CI 2.87 to 3.62). There was a decreasing trend of antiplatelet users (p = 0.015) during 2003 to 2012, whereas there was not a significant trend in statin (p = 0.78) or combination users (p = 0.83). Table 1 lists the bivariate analysis.
| Variable(%±SE) | Anti-platelet n=4447 | No Anti-platelet n=2325 | Statin n=3270 | No Statin n=3502 | Combination n=2590 | No Combination n=4182 |
|---|---|---|---|---|---|---|
| Proportion | 69.3±0.8 | 30.7±0.8 | 52.8±1.0 | 47.2±1.0 | 42.8±1.0 | 57.2±1.0 |
| Age (years) ∗ , † , ‡ | 68.6±0.30 | 64.8±0.47 | 69.0±0.34 | 65.6±0.39 | 68.7±0.38 | 66.4±0.36 |
| Men ∗ , † , ‡ | 62.4±1.0 | 51.4±1.4 | 64.0±1.1 | 53.4±1.2 | 65.9±1.2 | 53.9±1.1 |
| Race ∗ , † , ‡ | ||||||
| White (non-Hispanic) | 80.4±1.0 | 70.3±1.5 | 82.0±1.0 | 72.1±1.3 | 83.3±1.1 | 72.8±1.3 |
| Black (non-Hispanic) | 9.0±0.8 | 12.4±1.0 | 7.9±0.8 | 12.5±0.9 | 7.4±0.8 | 12.1±0.9 |
| Hispanic | 6.7±0.7 | 11.5±1.3 | 5.7±0.6 | 10.9±1.1 | 5.3±0.6 | 10.3±1.0 |
| Other | 3.9±0.4 | 5.7±0.8 | 4.4±0.5 | 4.5±0.5 | 4.0±0.5 | 4.8±0.6 |
| Insurance ∗ , † , ‡ | ||||||
| Private | 55.9±1.2 | 49.3±1.7 | 58.3±1.3 | 48.9±1.7 | 58.6±1.4 | 50.3±1.5 |
| Public Only | 39.6±1.1 | 42.0±1.6 | 38.4±1.2 | 42.5±1.5 | 37.9±1.3 | 42.1±1.4 |
| None | 4.5±0.4 | 8.7±0.8 | 3.3±0.4 | 8.5±0.7 | 3.4±0.4 | 7.5±0.6 |
| Region ∗ , † , ‡ | ||||||
| West | 16.4±1.0 | 21.6±1.4 | 17.2±1.1 | 18.9±1.2 | 16.2±1.1 | 19.4±1.2 |
| Northeast | 19.1±1.3 | 19.3±1.4 | 19.6±1.7 | 18.7±1.2 | 19.9±1.8 | 18.7±1.2 |
| Midwest | 23.8±1.3 | 18.9±1.5 | 23.9±1.4 | 20.5±1.4 | 24.8±1.6 | 20.4±1.3 |
| South | 40.7±1.5 | 40.2±1.7 | 39.3±1.6 | 42.0±1.6 | 39.2±1.7 | 41.6±1.5 |
| Metro-area | 79.4±1.8 | 81.7±2.1 | 79.6±1.9 | 80.5±1.8 | 80.5±1.8 | 79.5±2.1 |
| Degree ∗ , † , ‡ | ||||||
| No High School | 20.1±1.0 | 25.8±1.6 | 20.0±1.2 | 24.1±1.2 | 18.8±1.3 | 24.3±1.1 |
| High School/GED | 50.5±1.4 | 48.5±1.7 | 50.1±1.6 | 49.7±1.5 | 50.8±1.7 | 49.2±1.3 |
| More than High School | 29.4±1.3 | 25.6±1.6 | 30.0±1.5 | 26.2±1.3 | 30.4±1.6 | 26.5±1.2 |
| Poverty Category ∗ , † , ‡ | ||||||
| Poor/Near Poor | 23.8±1.2 | 19.5±0.9 | 18.2±0.8 | 23.8±1.2 | 18.2±0.9 | 22.8±1.0 |
| Low/Middle | 48.9±1.5 | 45.9±1.0 | 45.6±1.3 | 48.2±1.2 | 44.7±1.4 | 48.4±1.1 |
| High | 27.3±1.5 | 34.6±1.2 | 36.2±1.4 | 28.1±1.3 | 37.1±1.5 | 28.8±1.2 |
∗ p-value is <0.05 for the anti-platelet comparison.
† p-value is <0.05 for the statin comparison.
Figure 1 shows the proportion of combined antiplatelet/statin users by race/ethnicity and gender from 2003 to 2012. There were no significant trends by race/ethnicity, but Hispanics nearly had a significant decrease (p = 0.051). Since 2009, use of the combination varied by race/ethnicity with a significantly higher percentage of non-Hispanic whites reporting use (46.1%; 95% CI 43.7 to 48.4) compared with non-Hispanic blacks (31.3%; 95% CI 27.6 to 35.3) and Hispanics (27.7%; 95% CI 23.8 to 32.0; p <0.001 for both).
There were no significant trends by gender from 2003 to 2012. Since 2009, men (47.7%; 95% CI 45.2 to 50.2) were more likely to be treated with the combination than women (35.6%; 95% CI 33.1 to 38.2, p <0.001; Figure 1 ).
Because all the models had a significant interaction between race and gender, separate multivariable logistic regressions were used for men and women. Complete results are in Table 2 . Notably, non-Hispanic blacks and Hispanic men were less likely to report medication use, but there was no difference in use by race/ethnicity in women. There were also differences in all categories of use by region but only in men. A lack of health insurance was associated with less statin and combination use in both men and women. In post-prediction estimates using the model with the interaction term, non-Hispanic white men (0.47; 95% CI 0.43 to 0.50) were more likely to be on treatment than non-Hispanic white women (0.34; 95% CI 0.30 to 0.39). Non-Hispanic black men (0.30; 95% CI 0.25 to 0.37) and women (0.31; 95% CI 0.25 to 0.37) were less likely to use the combination compared with non-Hispanic white men (p <0.001). Both Hispanic men (0.32; 95% CI 0.26 to 0.37) and women (0.28; 95% CI 0.21 to 0.35) were less likely to use the combination than non-Hispanic white men (p <0.001).
| Male | ||||||
|---|---|---|---|---|---|---|
| Variable (95% CI) | Anti-platelet | p-value | Statin | p-value | Combination | p-value |
| Age (per 10 years) | 1.16 (1.06-1.27) | 0.001 | 1.27 (1.16-1.38) | <0.001 | 1.13 (1.03-1.23) | 0.004 |
| Race | ||||||
| White (non-Hispanic) | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| Black (non-Hispanic) | 0.62 (0.48-0.82) | 0.001 | 0.56 (0.41-0.75) | <0.001 | 0.48 (0.36-0.65) | <0.001 |
| Hispanic | 0.71 (0.52-0.97) | 0.031 | 0.56 (0.42-0.75) | <0.001 | 0.54 (0.40-0.73) | <0.001 |
| Other | 0.79 (0.49-1.28) | 0.33 | 0.81 (0.53-1.26) | 0.35 | 0.72 (0.45-1.17) | 0.19 |
| Insurance | ||||||
| Private | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| Public Only | 0.92 (0.70-1.21) | 0.56 | 0.73 (0.57-0.93) | 0.012 | 0.83 (0.65-1.06) | 0.14 |
| None | 0.79 (0.53-1.18) | 0.25 | 0.54 (0.36-0.80) | 0.003 | 0.65 (0.43-0.99) | 0.045 |
| Region | ||||||
| West | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| Northeast | 1.39 (0.93-2.05) | 0.10 | 1.40 (0.97-2.02) | 0.08 | 1.60 (1.11-2.28) | 0.011 |
| Midwest | 1.68 (1.15-2.45) | 0.007 | 1.57 (1.14-2.17) | 0.005 | 1.57 (1.10-2.24) | 0.033 |
| South | 1.46 (1.07-1.99) | 0.017 | 1.14 (0.87-1.49) | 0.35 | 1.27 (0.94-1.73) | 0.14 |
| Metro-area | 0.95 (0.70-1.28) | 0.71 | 1.14 (0.85-1.52) | 0.38 | 1.15 (0.85-1.56) | 0.36 |
| Degree | ||||||
| No High School | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| High School/GED | 1.14 (0.87-1.51) | 0.35 | 1.16 (0.88-1.53) | 0.29 | 1.17 (0.90-1.52) | 0.24 |
| More than High School | 1.49 (1.02-2.18) | 0.041 | 1.04 (0.81-1.33) | 0.74 | 1.42 (1.01-1.99) | 0.042 |
| Poverty Category | ||||||
| Poor/Near Poor | 1 (reference) | 1 (reference) | 1 (reference) | |||
| Low/Middle | 0.84 (0.65-1.08) | 0.17 | 0.90 (0.70-1.15) | 0.39 | 0.87 (0.69-1.11) | 0.26 |
| High | 1.28 (0.93-1.74) | 0.13 | 1.18 (0.88-1.57) | 0.27 | 1.24 (0.94-1.66) | 0.13 |
| Female | ||||||
|---|---|---|---|---|---|---|
| Variable (95% CI) | Anti-platelet | p-value | Statin | p-value | Combination | p-value |
| Age (per 10 years) | 1.39 (1.27-1.53) | <0.001 | 1.17 (1.06-1.29) | 0.003 | 1.18 (1.07-1.30) | 0.001 |
| Race | ||||||
| White (non-Hispanic) | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| Black (non-Hispanic) | 1.01 (0.76-1.32) | 0.97 | 0.81 (0.58-1.11) | 0.18 | 0.83 (0.60-1.15) | 0.27 |
| Hispanic | 0.86 (0.60-1.25) | 0.44 | 0.68 (0.45-1.01) | 0.053 | 0.68 (0.45-1.02) | 0.062 |
| Other | 0.71 (0.44-1.16) | 0.18 | 1.30 (0.82-2.05) | 0.26 | 1.00 (0.59-1.68) | 0.99 |
| Insurance | ||||||
| Private | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| Public Only | 0.88 (0.67-1.16) | 0.37 | 0.94 (0.69-1.28) | 0.710 | 0.93 (0.70-1.24) | 0.62 |
| None | 0.70 (0.43-1.13) | 0.14 | 0.46 (0.27-0.80) | 0.006 | 0.49 (0.28-0.88) | 0.017 |
| Region | ||||||
| West | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| Northeast | 1.03 (0.71-1.50) | 0.86 | 0.79 (0.52-1.20) | 0.27 | 0.78 (0.51-1.21) | 0.27 |
| Midwest | 1.37 (0.97-1.92) | 0.07 | 0.87 (0.62-1.22) | 0.43 | 1.10 (0.77-1.56) | 0.61 |
| South | 1.32 (0.98-1.79) | 0.06 | 1.00 (0.75-1.32) | 0.98 | 1.07 (0.78-1.46) | 0.69 |
| Metro-area | 0.91 (0.65-1.27) | 0.57 | 0.95 (0.71-1.27) | 0.73 | 0.95 (0.69-1.32) | 0.77 |
| Degree | ||||||
| No High School | 1 (reference) | – | 1 (reference) | – | 1 (reference) | – |
| High School/GED | 1.40 (1.04-1.88) | 0.025 | 0.90 (0.68-1.19) | 0.46 | 1.12 (0.84-1.48) | 0.23 |
| More than High School | 1.14 (0.82-1.61) | 0.43 | 0.72 (0.50-1.04) | 0.08 | 0.84 (0.59-1.21) | 0.36 |
| Poverty Category | ||||||
| Poor/Near Poor | 1 (reference) | 1 (reference) | 1 (reference) | |||
| Low/Middle | 0.96 (0.75-1.23) | 0.74 | 1.11 (0.86-1.42) | 0.42 | 1.02 (0.81-1.29) | 0.85 |
| High | 0.85 (0.62-1.16) | 0.30 | 1.20 (0.83-1.73) | 0.34 | 0.95 (0.66-1.37) | 0.79 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
