Variation in statin prescription among veterans with HIV and known atherosclerotic cardiovascular disease

Abstract

Background: People with HIV have increased atherosclerotic cardiovascular disease (ASCVD) risk, worse outcomes following incident ASCVD, and experience gaps in cardiovascular care, highlighting the need to improve delivery of preventive therapies in this population.

Objective: Assess patient-level correlates and inter-facility variations in statin prescription among Veterans with HIV and known ASCVD.

Methods: We studied Veterans with HIV and existing ASCVD, ie, coronary artery disease (CAD), ischemic cerebrovascular disease (ICVD), and peripheral arterial disease (PAD), who received care across 130 VA medical centers for the years 2018-2019. We assessed correlates of statin prescription using two-level hierarchical multivariable logistic regression. Median odds ratios (MORs) were used to quantify inter-facility variation in statin prescription.

Results: Nine thousand six hundred eight Veterans with HIV and known ASCVD (mean age 64.3 ± 8.9 years, 97% male, 48% Black) were included. Only 68% of the participants were prescribed any-statin. Substantially higher statin prescription was observed for those with diabetes (adjusted odds ratio [OR] = 2.3, 95% confidence interval [CI], 2.0-2.6), history of coronary revascularization (OR = 4.0, CI, 3.2-5.0), and receiving antiretroviral therapy (OR = 3.0, CI, 2.7-3.4). Blacks (OR = 0.7, CI, 0.6-0.9), those with non-coronary ASCVD, ie, ICVD and/or PAD only, (OR 0.53, 95% CI: 0.48-0.57), and those with history of illicit substance use (OR=0.7, CI, 0.6-0.9) were less likely to be prescribed statins. There was significant variation in statin prescription across VA facilities (10th, 90th centile: 55%, 78%), with an estimated 20% higher likelihood of difference in statin prescription practice for two clinically similar individuals treated at two comparable facilities (adjusted MOR = 1.21, CI, 1.18-1.24), and a greater variation observed for Blacks or those with non-coronary ASCVD or history of illicit drug use.

Conclusion: In an analysis of large-scale VA data, we found suboptimal statin prescription and significant interfacility variation in statin prescription among Veterans with HIV and known ASCVD, particularly among Blacks and those with a history of non-coronary ASCVD.

Keywords: CVD; Cardiovascular disease; HIV; Health care quality; Health services; Human immunodeficiency virus; Statin.

Figure 1

Multivariable odds ratios of A, any statin and B, recommended intensity statin prescription among Veterans with HIV and known ASCVD for several individual characteristics (9,680 patients across 130 facilities). ∗∗∗ P < .005, ∗∗ P < .01, ∗ P < .05. The black boxes represent odds ratios, the horizontal bars represent 95% confidence intervals ^aAll the variables in the table were included as fixed effects in the multivariate hierarchical model, except for prior ICVD, PAD and CAD procedures which were not included in the main model as they are subsets of history of ICVD, PAD and CAD, respectively. ASCVD was included as ICVD, PAD, and CAD (three variables) in the main model. The odds ratios of ASCVD types were determined in a secondary model that combined ICVD, PAD and CAD into one variable (ASCVD type) with 3 categories - ASCVD type: all CAD, ASCVD type: ICVD only, ASCVD type: PAD +/−ICVD. Similarly, Odds ratios for prior ICVD, PAD and CAD procedures were estimated in a secondary model adjusting for history of prior procedure instead of history of ICVD, PAD and CAD. Four facility level characteristics (facility size, complexity, rurality, and VA region) were also included as fixed effect. VA facility was included as random effects. Illicit drug use includes opioids, sedatives, cannabis, hallucinogens, cocaine, amphetamine drug use. ART, antiretroviral therapy; ASCVD, atherosclerotic cardiovascular disease; CAD, coronary artery disease; CHF, congestive heart failure; CKD, chronic kidney disease; ECI, Elixhauser comorbidity index; ICVD, ischemic cerebrovascular disease; PAD, peripheral artery disease.

Figure 2

Percentage of Veterans with HIV and known ASCVD with any statin prescription, A, waterfall plot by VA facility, and B, heat map by US states In Figure 2A, each bar graph represents the percentage of station prescription in each facility. In Figure 2B the various shades represent percentage of Veterans with statin prescription by state. ASCVD, atherosclerotic cardiovascular disease; VAMC, VA Medical Center.