Undertreatment of hyperlipidemia in patients with coronary artery disease and heart failure☆☆☆

Methods 

Results 

Of the 41,487 patients included in the study, about one third had documentation of HF. Measurement of left ventricular function was documented in 74% of patients with CAD and HF and in 52% of patients with CAD alone. Systolic dysfunction (left ventricular ejection fraction < 40%, or qualitatively moderate or severe) was recorded in 50% of patients with CAD and HF who had documented left ventricular function. The mean ejection fraction was 41% ± 0.2% for those with HF (n = 8,760) compared with 59 ± 0.1% for patients without HF (n = 11,108). ACE inhibitor prescription was higher in patients with CAD and HF compared with patients with CAD alone, 51% versus 21%.

Characteristics of CAD patients with and without HF are shown in Table 1.

Table 1. Patient characteristics

		CAD without Heart Failure (n = 27,285)	CAD with Heart Failure (n = 14,202)
	Mean age (yr ± SEM)	67 ± 0.1	72 ± 0.1*
	Male (%)	64	60*
	History of myocardial infarction (%)	41	49*
	History of revascularization (%)	58	51*
	History of hypertension (%)	55	55
	History of diabetes (%)	20	30*
	Cardiology clinic (%)	53	57*
	Lipid-lowering drug (%)	52	36*
	LDL-C documentation (%)	69	53*
	Mean LDL-C (mg/dL)	120 ± 0.3	119 ± 0.4
	LDL ≤ 100 mg/dL (%)	30	31

P<.01 CAD, coronary artery disease; SEM, standard error of the mean; LDL-C, low-density lipoprotein cholesterol.

Lipid medication prescription rates declined rapidly with age for those 65 years and older (Fig 1) among CAD patients with and without HF.

View Large Image Download to PowerPoint

Fig. 1. Percent of patients treated with lipid-lowering therapy and mean low-density lipoprotein cholesterol levels in coronary artery disease patients with and without heart failure.

Predictors of lipid-lowering drug prescriptions for the entire study population are shown in Table 2.

Table 2. Predictors of the prescription of lipid-lowering agents in CAD patients (n = 38,467)

		Odds Ratio*	95% Confidence Limits
	Heart Failure	0.68	0.64-0.71
	Age (Years)†
	75+	0.38	0.35-0.40
	65-74	0.74	0.70-0.79
	45-54	1.03	0.94-1.12
	21-44	0.94	0.81-1.09
	Male	0.93	0.89-0.98
	LDL-C Documented	3.83	3.65-4.02
	History of Myocardial Infarction	1.18	1.12-1.23
	History of Revascularization‡	1.99	1.90-2.08
	History of Hypertension	1.02	0.98-1.07
	History of Diabetes	0.97	0.92-1.02
	Cardiology Specialty	1.14	1.09-1.20
	Region#
	Northeast	1.10	1.03-1.18
	Midwest	1.12	1.05-1.19
	West	1.26	1.17-1.35
	*Odds ratios are adjusted for all variables shown. †Compared with patients aged 55 to 64 years. ‡Bypass surgery or percutaneous revascularization. #Compared with the South.

Discussion 

Our study demonstrates markedly lower treatment rates for hyperlipidemia in CAD patients with HF compared with those without HF. Only about one-third of patients with CAD and HF received a prescription for a lipid-lowering medication, compared with about half of patients without HF. A potential reason for this disparity may be a physician emphasis on the prescription of “classical” HF therapy, including ACE inhibitors, digoxin, and diuretics, and, more recently, β-blockers and a focus on relief of symptoms. Consistent with this hypothesis is the fact that documentation of LDL-C in our study was significantly lower in the patients with combined CAD and HF (53%) compared with patients with CAD alone (69%). Our finding that ACE inhibitor prescription rates were higher and lipid-lowering drug prescription rates were lower among CAD and HF patients compared with patients with CAD alone suggests that the relatively low utilization of lipid-lowering drugs among HF patients reflect barriers specific to lipid management rather than overall lower drug prescription among HF patients.

Evidence of lipid-lowering therapy improving morbidity or mortality in HF patients is not well-established because medically complex, high-risk HF patients have often been excluded from clinical trials. However, CAD is the most common etiology of HF. It may be reasonably hypothesized that the beneficial effects of statins, including lipid-lowering, plaque stabilization, anti-inflammatory properties, and improvement of endothelial function may result in slowing the progression of ischemia that may lead to worsening HF. Animal studies have suggested a beneficial effect of statins on postinfarction remodeling and survival.18 The Scandinavian Simvastatin Survival Study reported a 19.6% long-term reduction in the development of HF in patients with a history of myocardial infarction who were randomized to statin therapy.6 A prospective observational study of 1,410 older patients with a history of myocardial infarction and serum LDL-C > 125 mg/dL found a 26% reduction in the development of heart failure in patients treated with a statin.19 Multivariate analysis confirmed that statin treatment was an independent predictor (risk ratio 0.52) of lowered risk for development of HF. In the CARE trial, pravastatin was found to be equally effective in reducing coronary events in the 706 patients with an ejection fraction between 26% and 40% as in the group with an ejection fraction > 40%.7 Coronary events were reduced by 28% in the low ejection fraction group and 23% in the group with ejection fraction > 40%. A retrospective analysis of the ELITE II trial suggests that patients with symptomatic HF treated with statin therapy had a lower mortality—10.6% versus 17.6%.8 The mortality effect of statin therapy was independent of treatment with either captopril or losartan. Overall, age was an important predictor of lipid-lowering prescription. Patients 65 years of age or older in our study were treated much less aggressively compared with younger patients. Suboptimal care of patients with advanced age was not justified by LDL-C or goal attainment levels. Undertreatment was also not warranted based on clinical trial data.20 Consistent with this, the recently published National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP III) guidelines for CAD patients states that “no hard-and-fast age restrictions appear necessary when selecting persons with established coronary heart disease for LDL-lowering therapy.”10 Despite clinical practice guidelines, more than two-thirds of CAD patients in QAP-II with recorded LDL-C values were above the NCEP ATP II goal.9 Mean LDL-C concentrations were well above goal.

Documentation of LDL-C was the most important predictor associated with increased lipid-lowering prescription both in CAD patients with and without HF. Patients with recorded LDL-C values on the chart were 4-fold more likely to receive a prescription for lipid-lowering medication. Other predictors of lipid-lowering prescription were similar between the 2 groups. Patients with more severe CAD (ie, those with a history of myocardial infarction or who had undergone revascularization) were more likely to receive a lipid-lowering prescription. However, patients with diabetes, which is associated with accelerated atherosclerosis, were not more aggressively managed consistent with our previous findings.

Compared with QAP, few, if any, other data sources are available that provide recent detailed lipid assay results and medication information for such a large number of CAD patients in an ambulatory setting. In the 1991–1992 National Ambulatory Medical Care Survey, only 23% of patients had cholesterol testing and only 23% of patients were prescribed therapy.14 Medication prescription was associated with a history of cardiovascular disease, northeast region of the United States, nonobese patients, and care by an internist. A study of 348 patients in a southern Florida health maintenance organization clinic seen from 1992 through 1996 revealed that only 14% of patients eligible for drug therapy were treated.21 In another study, the temporal trends of cholesterol measurement and lipid-lowering therapy of 5,204 residents of Worcester, Massachusetts, hospitalized with an acute myocardial infarction during 1986–1997 were examined.22 Only 24% of patients admitted during 1997 were tested. In patients with elevated serum cholesterol (≥240 mg/dL), 1.9% received lipid-lowering therapy in 1986 and 36.6% in 1997. We have reported in QAP-I15 that 39% of outpatients with documented CAD were receiving lipid-lowering therapy, which increased to 52% in QAP-II.16

Attainment of target LDL-C is low. The Lipid Treatment Assessment Project studied a national cohort of 4,888 patients already receiving lipid-lowering therapy during 1996–1997.12 Of the 30% of patients with documented CAD, only 18% were at goal LDL. We have reported in QAP-I15 that only 25% of patients with documented CAD were at goal, which has increased to 30% in QAP-II.16 Although there has been a temporal increase in lipid testing, prescription of lipid-lowering therapy, and attainment of LDL goal, care remains suboptimal.

There are several limitations to our study. The results from QAP are not likely generalizable to the US population. Only medical practices with high-volume use of medications were invited to participate. Thus participating practices in QAP probably represent some of the largest and most sophisticated ambulatory practices in the nation. The impact of this site selection on the generalizability of our results to a broader population is difficult to ascertain, but we suspect that the clinical experience and disease management systems available at these large practices would enhance lipid management. If this is true, then LDL-C goal attainment and pharmacologic therapy in QAP were likely better than the national average.

Incomplete medical information is an important limiting factor in this study. Data regarding adverse effects and contraindications are difficult to assess from the medical record. Although significant contraindications and adverse effects are uncommon for statins, these conditions may have resulted in the intentional withdrawal of these medications from a minority of medically complex patients. Additionally, medical information is obviously incomplete for patients seeing physicians outside of the study. Thus lipid assay results and medication information for some patients were not available from the medical practices included in the study. Our data also do not provide information characterizing patient compliance. Effective lipid management would require that patients fill prescriptions and take medications as advised. In one study of patients older than 65 years, lipid medication prescriptions went unfilled for more than one third of the study year on average.23 Thus, in addition to physician undertreatment, lipid management is further impeded by barriers to patient compliance.

Conclusion 

Patients at high risk resulting from combined CAD and HF were markedly undertreated relative to those with CAD alone. Patients with CAD, HF, and advanced age simultaneously experience among the highest risk and the lowest lipid drug treatment rates. This marked undertreatment may contribute to the poor outcome of patients with CAD and HF. The disparities in pharmacologic management among these high-risk CAD patients constitutes an important public health problem and one that is likely to grow as the population ages and the prevalence CAD and HF increases. Documentation of LDL testing was the most important predictor of increased lipid-lowering prescription. The identification of additional barriers to lipid management and the development of interventions to circumvent these barriers should be a priority of future research and public health policy.