Methods

Using the Department of Veterans Affairs (VA) administrative data sources, we identified patients with CVD with primary care clinic visits in the VA health care system (130 facilities or their associated community-based outpatient clinics) from October 1, 2010, to September 30, 2011. Patients with CVD were those with a history of coronary heart disease or peripheral artery disease, including ischemic stroke. We identified patients as having coronary heart disease using International Classification of Diseases, Ninth Revision, Clinical Modification diagnoses and procedure codes for unstable angina or myocardial infarction, or current procedural terminology codes for percutaneous coronary intervention or coronary artery bypass grafting ( Supplementary Table A ). We also used International Classification of Diseases, Ninth Revision, Clinical Modification diagnoses codes ( Supplementary Table A ) to identify patients with peripheral artery disease. As patients with metastatic cancers and those receiving hospice care are not considered candidates for quality measurement, we excluded them from our cohort ( Figure 1 ). An index visit was defined as the most recent primary care visit for each patient from October 1, 2010, to September 30, 2011.

Flow chart of the study cohort identification and exclusions. ^∗ Total adds up to >10,914 as some patients met >1 exclusion criteria. ^† n for final analysis. CHD = coronary heart disease; PAD = peripheral arterial disease.

Using VA administrative pharmacy data sources, we evaluated whether these patients were prescribed a statin or a nonstatin LDL-C lowering medication within 100 days before or 14 days after their index visit. The nonstatin LDL-C lowering medications included bile acid-binding resins, niacin, and ezetimibe. Fibrates were not included as they only marginally lower LDL-C, and fish oil-based therapies were not included as they do not lower LDL-C.

Using VA’s administrative data sets, we assessed several patient characteristics including age, race, and history of diabetes or hypertension ( Supplementary Table A ). To assess the impact of patient’s illness burden on receipt of evidence-based statin therapy, we calculated a mean diagnostic cost group relative risk score (RRS) for each patient. Diagnostic cost group RRS is a ratio of the predicted cost to the mean cost and has been used as a measure of illness burden. For example, a patient with an RRS of 2 is expected to be twice as costly, with an illness burden twice as high as an “average” patient. We also assessed facility, provider, and system-of-care characteristics that could impact receipt of statin (or high-intensity statin) therapy. These included whether the facility was a teaching facility, the type of provider (physician or nonphysician primary care provider, such as a nurse practitioner or a physician assistant), and the number of primary care visits by a patient in the 12 months before the index visit. We first compared the patient, facility, provider, and system-of-care characteristics between female and male patients with CVD. We then compared the most recent levels of various lipid parameters (total cholesterol, LDL-C, high-density lipoprotein cholesterol [HDL-C], triglycerides, non-HDL-C) between male and female patients with CVD.

Our primary outcome was the receipt of statin therapy. We assessed the proportion of female and male patients receiving any statin. We then performed multivariate hierarchical logistic regression to determine whether gender is independently associated with receipt of statins. Covariates used for adjustment included patient, facility, provider, and system-of-care characteristics described previously. We also adjusted for clustering of patients at the facility-level in our regression models using generalized linear latent and mixed models in STATA (College Station, TX). We subsequently performed multivariate hierarchical logistic regression to determine whether gender is independently associated with the receipt of high-intensity statins among patients with CVD. Per the recent ACC/AHA guideline, high-intensity statin therapy was defined as the daily statin intensity associated with approximately ≥50% LDL-C reduction (atorvastatin dose of 40 to 80 mg/day, rosuvastatin 20 to 40 mg/day, or simvastatin >40 mg/day).

To evaluate the proportion of female patients with CVD at each facility receiving statin or high-intensity statin therapy, we calculated median facility-level rates for statin or high-intensity statin therapy with interquartile ranges in female patients with CVD. These analyses were restricted to facilities with at least 10 female patients with CVD (only 1 of the 130 facilities was excluded from these analyses).

As per the recent ACC/AHA guidelines, we also evaluated the proportion of female and male patients with CVD receiving at least a moderate-intensity statin therapy described as the daily dose of statin therapy associated with approximately 30% to <50% LDL-C reduction. We conducted the analyses with SAS version 9.1.3 (SAS Institute Inc, Cary, North Carolina) and STATA version 11 (College Station, Texas). The protocol was approved by the Institutional Review Boards at Baylor College of Medicine and the Michael E. DeBakey VA Medical Center.

Results

We excluded 10,941 (1.1%) patients with documented metastatic care or terminal illness from our initial cohort of 983,476 patients ( Figure 1 ). Our final analyses included 972,532 patients with 13,371 women (1.4%) and 959,161 (98.6%) men. Table 1 lists a comparison of the baseline characteristics between female and male patients with CVD. Female patients with CVD were on average younger, less often White, had a lower prevalence of diabetes and hypertension, and a higher prevalence of peripheral artery disease compared with male patients. Female patients with CVD had a higher overall illness burden (diagnostic cost group RRS) and were more likely to receive care at a teaching facility compared with male patients with CVD. On average, female patients with CVD received more primary care visits than male patients in the 12 months before their index visit. Female patients with CVD were less likely to receive care from a physician primary care provider (as opposed to a nurse practitioner or a physician assistant).

As listed in Table 2 , a lower proportion of female patients with CVD (57.6%) were receiving statin therapy compared with male patients (64.8%). Furthermore, a lower proportion of female patients with CVD (21.1%) were on high-intensity statin therapy compared with male patients with CVD (23.6%). Total cholesterol, LDL-C, HDL-C, triglycerides, and non–HDL-C were on average 25 mg/dl, 14 mg/dl, 9 mg/dl, 6 mg/dl, and 15 mg/dl higher in female patients compared with male patients with CVD. In fully adjusted analyses ( Table 3 ), female gender was independently associated with 32% lesser likelihood of receipt of any statin therapy (odds ratio 0.68, 95% confidence interval 0.66 to 0.71) and a 24% lesser likelihood (odds ratio 0.76, 95% confidence interval 0.73 to 0.80) of receiving high-intensity statin therapy.

Stay updated, free articles. Join our Telegram channel

Join