Racial Differences in Switching, Augmentation, and Titration of Lipid-lowering Agents by Medicare/Medicaid Dual-eligible Patients

Published on: 
, , ,
Supplements and Featured Publications, Health-related Consequences of and Treatment Strategies for Dyslipidemia, Volume 13, Issue 3 Suppl

Research Objective: The goal of this study was to examine prescription fill patterns of

Conclusion: These results suggest that African Americans may be receiving less aggressive treatment than other patients, which in turn may explain why many studies find that African Americans are less likely to reach lipid goals. These treatment disparities merit further study, because they may impact dual-eligible patients moving into Medicare Part D plans.

Demographic variables: Age (<65 years or =65 years), sex, and ethnicity (African American, White, Hispanic, Asian, Other, Missing).

Comorbidities: The Chronic Disease Score (CDS) was used to control for comorbidities. The CDS uses patterns of prescription utilization data previously identified by a consensus judgment process as indicative of chronic disease to construct a weighted score that predicts healthcare cost and utilization. The maximum possible score is 35.14-16

Number of healthcare visits: Healthcare utilization was defined by the number of outpatient visits.

Days on therapy: Lipid-lowering agent use was defined by total days on therapy during the study period.

The Medicaid states could not be identified for privacy reasons. However, they are blinded and included in the multivariate models as controls for any variation in size of Medicaid population or potential practice pattern differentials between the states. This was a broad measure at best, but it was anticipated that some of the variation might be accounted for by including the state controls in the regression models.

Two types of analyses were performed in this study. The first analysis described the sample. Descriptive characteristics are presented for the sample stratified by each of the adherence outcomes of interest. The second analysis was a set of logistic regression models specified to estimate the probability of switching, augmenting or titrating a lipid-lowering agent dose upward. The models controlled for age, sex, state of residence, total days on therapy, number of outpatient visits, and health status as proxied by use of the CDS.


The study sample consisted of 239 530 patients, of whom 62.8% were women. Of the total sample, 102 695 white, 44 221 Asian, 32 670 Hispanic, 22 570 African American, 37 374 other ethnicity lipid-lowering agent users who were dually eligible for Medicare and Medicaid. Table 1 shows the demographics of the population stratified by ethnicity and age. All ethnicities within the “Other†categories were merged into 1 group in subsequent analyses. The study sample consisted of a greater proportion of whites relative to the other thnicities. The majority of the patients were 65 years or older (187 151).

The descriptive results show that the majority of women in this study did not switch, augment, or have their lipid-lowering agent titrated upward. The same pattern was seen among patients aged 65 years or older. Regarding the patterns of change among patients stratified by ethnicity, fewer African Americans (9%) switched lipid-lowering agents than Asians, Hispanics, whites, or other ethnicities (14%, 13%, 12%, and 13%, respectively). Also, fewer African Americans (3%) augmented with another lipid-lowering agent compared with Asians, Hispanics, whites, or other ethnicities (6%, 5%, 6%, and 5%, respectively). The percentage of patients who titrated their lipid-lowering agent upward was greater than those who switched or augmented. This may be the result of several factors. It may indicate that clinicians have been more likely to increase the dosage rather than switch or augment to improve effectiveness. However, it may also show that the physician prescribed the correct agent and rather than switch, the dose was titrated up for improved effectiveness. The average CDS score is about 8, with higher scores among those who switched, augmented, or titrated.

The CDS was reported for each of the outcomes groups of interest. This score also serves as a predictor of healthcare cost and utilization. We therefore reviewed the healthcare utilization measure in this study (outpatient visits) by levels of the CDS. The outpatient visits are reported by the CDS stratified into quartiles within each ethnicity included in this study. Patients with missing data were excluded from this analysis, which reduced the study sample size from 239 530 to 232 229. The results are shown in Table 3. This was not a temporal analysis but was a cross-sectional picture by quartile. The sample n values are shown in the Table to demonstrate that there are different patients in each quartile.

The logistic regressions showed that African Americans were statistically significantly less likely to switch lipid-lowering agents (OR, 0.68; 95% CI, 0.60-0.78), augment lipid-lowering agents (OR, 0.53; 95% CI, 0.43-0.66), or titrate upward (OR, 0.75; 95% CI, 0.67-0.84) than whites. In most of the models, the Hispanic and Asian patients were significantly more likely to augment, switch, or titrate upward. The exception is that Hispanic patients were less likely to augment than whites, but this variable was not significant in the model. The "Other" group did not attain statistical significance. Relative to whites, only the African Americans had a lower probability of a change in lipid-lowering agents. To test for the possibility that this group of patients may be healthier than the others, thus perhaps being the cause of the decrease in lipid-lowering agent change, the CDS data were interacted with African American ethnicity to control for health in this specific patient group. The ORs in all 3 models are greater than 1, indicating a greater likelihood of change than whites, but statistical significance is not achieved. This interaction term therefore serves as a control for health of this ethnicity group, and may mean that the decrease in the probability of lipid-lowering agent change is caused by factors other than health status.

There were a few effects seen in the other independent covariates included in the models.

Human Participant Protection: No human participants were involved, thus approval by an institutional review board was not required.

Address correspondence to: Kirsten J. Axelsen, MS, Director, Economic & Policy Research, Pfizer Inc, 235 E. 42nd Street 235-12-35, New York, NY 10017. E-mail:

1. American Heart Association. Heart Disease and Stroke Statistics—2005 Update. Dallas, Tex: American Heart Association; 2006.

2. Schuster H. Improving lipid management—to titrate, combine or switch. Int J Clin Pract. 2004;58:689-694.

Am J Cardiol.


3. Ferdinand KC. Coronary artery disease in minority racial and ethnic groups in the United States. 2006;97(suppl 2A):12A-19A.

4. Trivedi AN, Zaslavsky AM, Schneider EC, Ayanian JZ. Relationship between quality of care and racial disparities in Medicare health plans. JAMA. 2006;296:1998-2004.

Ethn Dis.

5. Williams ML, Morris MT 2nd, Ahmad U, Yousseff M, Li W, Ertel N;VANJHCS research group. Veterans Affairs of New Jersey Healthcare System. Racial differences in compliance with NCEP-II recommendations for secondary prevention at a Veterans Affairs medical center. 2002;12:S1-58-S1-62.

6. Massing MW, Foley KA, Carter-Edwards L, Sueta CA, Alexander DM, Simpson RJ. Disparities in lipid management for African Americans and whites with coronary artery disease: a national cross-section study. BMC Cardiovasc Disord. 2004;4:15.

Am Heart J.

7. Yood MU, McCarthy BD, Kempf J, et al. Racial differences in reaching target low-density lipoprotein goal among individuals treated with prescription statin therapy. 2006;152:777-784.

8. Chong PH, Tzallas-Pontikes PJ, Seeger JD, Stamos TD. The low-density lipoprotein cholesterol-lowering effect of pravastatin and factors associated with achieving targeted low-density lipoprotein levels in an African-American population. Pharmacother. 2000;20:1454-1463.

Treat Guidel Med Lett.

9. Drugs for lipids. 2005;3:15-22.

10. Pearson TA, Denke MA, McBride PE, et al. Effectiveness of ezetimibe added to ongoing statin therapy in modifying lipid profiles and low-density lipoprotein cholesterol goal attainment in patients of different races and ethnicities: a substudy of the Ezetimibe Add-On to Statin for Effectiveness trial.Mayo Clin Proc. 2006;81:1177-1185.

Am J Cardiol.

11. Stein EA. Managing dyslipidemia in the high-risk patient. 2002;89(suppl 5A):50C-57C.

12. McKenney JM. Optimizing LDL-C lowering with statins. Am J Ther. 2004;11:54-59. 13. Miller M. Niacin as a component of combination therapy for dyslipidemia. Mayo Clin Proc. 2003;78:735-742.

J Clin Epidemiol.

14. Von Korff M, Wagner EH, Saunders K. A Chronic Disease Score from automated pharmacy data. 1992;45:197-203.

15. Johnson RE, Hornbrook MC, Nichols GA. Replicating the Chronic Disease Score (CDS) from automated pharmacy data. J Clin Epidemiol. 1994;47:1191-1199.

Med Care.

16. Clark DO, Von Korff M, Saunders K, Baluch WM, Simon GE. A Chronic Disease Score with empirically derived weights. 1995;33:783-795.

17. Boulos DL, Groome PA, Brundage MD, et al. Predictive validity of five comorbidity indices in prostate carcinoma patients treated with curative intent. Cancer.2006;106:1804-1814.

J Am Geriatr Soc.

18. Di Bari M, Virgillo A, Matteuzzi D, et al. Predictive validity of measures of comorbidity in older community dwellers: the Insufficienza Cardiaca negli Anziani Residenti a Dicomano Study. 2006;54:210-216.

19. Parker JP, McCombs JS, Graddy EA. Can pharmacy data improve prediction of hospital outcomes? Comparisons with a diagnosis-based comorbidity measure. Med Care. 2003;41:407-419.

20. Gross DJ, Schondelmeyer SW, Raetzman SO. Trends in manufacturer prices of brand name prescription drugs used by older Americans—second quarter 2005 update. Washington, DC: AARP; 2005. Available at: Accessed February 8, 2007.

21. Taira DA, Safran DG, Seto TB, et al. Do patient assessments of primary care differ by patient ethnicity? Health Serv Res. 2001;36:1059-1071.


22. Cooper-Patrick L, Gallo JJ, Gonzales JJ, et al. Race, gender, and partnership in the patient-physician relationship. . 1999;282:583-589.

23. Johnson RL, Saha S, Arbelaez JJ, Beach MC, Cooper LA. Racial and ethnic differences in patient perceptions of bias and cultural competence in health care. J Gen Intern Med. 2004;19:101-110.

Med Care.

24. Murray-Garcia JL, Selby JV, Schmittdiel J, Grumbach K, Quesenberry CP Jr. Racial and ethnic differences in a patient survey: patients’ values, ratings, and reports regarding physician primary care performance in a large health maintenance organization. 2000;38:300-310.

25. Lurie N, Zhan C, Sangl J, Bierman AS, Sekscenski ES. Variation in racial and ethnic differences in consumer assessments of health care. Am J Manag Care. 2003;9:502-509.

26. Department of Health and Human Services. 30 million Medicare beneficiaries now receiving prescription drug coverage [press release]. Washington, DC: Department of Health and Human Services; April 20, 2006. Available at: Accessed February 6, 2007.

27. Donohue J, Hanlon JT. Helping Medicare patients benefit from the new Medicare drug benefit: an overview with practice tips. Am J Geriatr Cardiol. 2005;14:291-297.

Am J Geriatr Cardiol.

28. Usdin S. Medicare Part D: an overview of its practical and clinical impact. 2005;14:284-288.