Guidelines have been established for the treatment of patients with heart failure (HF) and left ventricular dysfunction, but renal dysfunction might limit adherence to these guidelines. Few data have characterized the use of guideline-recommended therapy for patients with HF, left ventricular dysfunction, and renal dysfunction who are treated in outpatient settings. The Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF) was a prospective study of patients receiving treatment as outpatients in cardiology practices in the United States. The rates of adherence to 7 guideline-recommended therapies were evaluated for patients with a left ventricular ejection fraction of ≤35%. The estimated glomerular filtration rate was estimated using the Modification of Diet in Renal Disease formula for 13,164 patients who were categorized as having stage 1 through stage 4/5 chronic kidney disease (CKD). More than 1/2 (52.2%) of the patients had stage 3 or 4/5 CKD. Older patients and women were at increased risk of higher stage CKD, and the rates of co-morbid health conditions were significantly greater among patients with more severe CKD. The patients with more severe CKD were significantly less likely to receive all interventions except cardiac resynchronization therapy. However, multivariate analysis controlling for patient characteristics revealed that the severity of CKD was an independent predictor of adherence to angiotensin-converting enzyme inhibitor/angiotensin receptor blocker therapy but not to any of the 6 other guideline-recommended measures. In conclusion, these results confirm that CKD is common in patients with HF and left ventricular dysfunction but is not independently associated with adherence to guideline-recommended therapy in outpatient cardiology practices, with the exception of angiotensin-converting enzyme inhibitor/angiotensin receptor blocker therapy.
A significant proportion of patients with heart failure (HF) also have renal dysfunction, which has been attributed to the common risk factors for HF and renal impairment, including hypertension, diabetes mellitus, and atherosclerosis. Renal impairment in patients with HF is recognized as an independent risk factor for morbidity and mortality. Despite the greater risk of mortality in patients with HF and chronic kidney disease (CKD), evidence has suggested that guideline-recommended therapies for HF are less likely to be provided to patients with co-morbid HF and CKD. However, many of these studies involved hospitalized patients or were conducted before the release of contemporary therapeutic guidelines for HF. Thus, information about the rates and predictors of use of guideline-recommended therapies for outpatients with HF and reduced left ventricular ejection fraction who also have renal dysfunction is limited. Identification of the rates of adherence to guideline-recommended medical therapies and the factors associated with variations for patients with HF and CKD might provide important information about patterns of care and facilitate strategies to increase adherence. The present study was undertaken to determine the rates and severity of CKD in a population of patients with HF receiving care at outpatient cardiology practices in the United States and to assess the effect of CKD on adherence to guideline-recommended therapies.
IMPROVE HF was a prospective, longitudinal cohort study undertaken to characterize the current treatment of patients receiving care in outpatient cardiology practice settings for chronic HF or prior myocardial infarction and left ventricular dysfunction. The overall study objectives, design, and methods, including definitions of the 7 process measures, have been previously described. In brief, community-based practices, including single-specialty and multispecialty cardiology practices representing all geographic regions of the United States, were invited to participate. Patient eligibility criteria included a diagnosis of HF documented by physician assessment on ≥2 separate visits for HF treatment in the current practice setting during the 2-year period before the initiation of IMPROVE HF. Left ventricular dysfunction was confirmed by quantitative or qualitative determination of a left ventricular ejection fraction of ≤35%, as measured from the most recent echocardiogram, nuclear multiple-gated acquisition scan, contrast ventriculogram, or magnetic resonance imaging scan.
The baseline data entered into the IMPROVE HF registry between 2005 and 2007 were used for the present analysis. The baseline data for each patient were abstracted from the medical chart by trained, independent, chart review specialists. The patient data included demographic and clinical characteristics, medical history, previous cardiac treatments, New York Heart Association functional class, laboratory and diagnostic tests and results, and current pharmacologic and device-based HF treatments. Documented contraindications, intolerance, or other reasons (e.g., economic, social, religious, refusal, or patient nonadherence) for not prescribing evidence-based HF therapies were recorded by the chart review specialists when noted in the medical record. QRS duration was obtained from the most recent electrocardiogram (computerized reading or physician measurement). A representative sample of the patient medical records was screened to yield a median of 90 patients with HF for each practice (twenty-fifth and seventy-fifth percentiles, 58 and 107, respectively) for the study baseline period using the method described in the trial design report. Analyses examining the effect of renal function on the delivery of guideline-recommended therapies for HF were prespecified in the study protocol.
All practices participating in the IMPROVE HF study were approved by a local or central institutional review board or received institutional review board waivers. Data quality was addressed by developing prespecified definitions for each variable, using consistent chart review specialists to collect data, and conducting regular, centralized retraining and testing of the chart review specialists to maintain accuracy in data abstraction. The average inter-rater reliability (κ) was 0.82. Automated quality checks (1.7 per data field) were performed to check the internal accuracy of the data, and monthly reports were generated to assess the completeness and accuracy of the data submitted. The registry coordinating center was Outcome Sciences (Cambridge, Massachusetts).
Seven care measures were prospectively selected by the IMPROVE HF Steering Committee to quantify the quality of outpatient delivery of guideline-recommended HF therapy. All 7 measures were designated as class I therapies (useful and effective) by the American College of Cardiology/American Heart Association guidelines. Of the 7 measures, 4 were American College of Cardiology/American Heart Association outpatient performance measures, and 3 were not. The specific measures included (1) angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (ACE-I/ARB) therapy; (2) β-blocker therapy; (3) aldosterone receptor antagonist therapy; (4) anticoagulation therapy for atrial fibrillation or flutter; (5) implantable cardioverter-defibrillator (ICD) therapy; (6) cardiac resynchronization therapy; and (7) patient education about HF.
Patients eligible for inclusion in the calculations for each of the 7 care measures included only those who met the defined criteria for each specific care measure and for whom no contraindications, intolerance, refusal, or other documented rationale explaining the reason the guideline-recommended therapy should not be provided were present. Documentation of New York Heart Association functional status was a prerequisite for eligibility for the aldosterone antagonist, ICD, and cardiac resynchronization therapy measures. Thus, patients for whom no quantitative or qualitative documentation of New York Heart Association functional status consistent with the prespecified definitions were excluded from the analyses of these 3 measures. The estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease formula: eGFR = 186.3 × (serum creatinine) −1.154 × (age) −0.203 × (1.212 if black) × (0.742 if female). Patients with missing race data were considered nonblack for these calculations. Patients for whom the eGFR was not calculated because of missing data were not included in any between-group comparisons or multivariate analyses.
All statistical analyses were performed by independent biostatisticians contracted by Outcome Sciences. Descriptive statistics for the patient and practice characteristics were calculated and reported for all practices in the IMPROVE HF registry that had completed a practice survey at baseline. These included the sample size, mean, median, and standard deviation for continuous variables and the sample size and percentage for categorical variables. The proportion and 95% confidence intervals (CIs) or median and twenty-fifth and seventy-fifth interquartile percentages were calculated for patient characteristics and the 7 measures stratified by stage of CKD. An overall test of differences in the mean values between the 4 CKD groups was performed using analysis of variance. Pairwise comparisons of patient groups were done for overall tests that were significant using the Bonferroni adjustment. Univariate general estimating equation hierarchical models were then calculated for the patient clinical and demographic characteristics that might be associated with stage of CKD, controlling for possible intrapractice data correlations. Multivariate logistic regression general estimating equation analysis was completed to determine whether the eGFR (analyzed as a continuous variable) was an independent predictor of adherence to each of the 7 guideline-recommended medical therapies after adjusting for patient characteristics and signs and symptoms of HF. The analyses were completed using Statistical Analysis Systems statistical software, version 9.1 (SAS Institute, Cary, North Carolina). All statistical tests were 2-sided, and results were considered statistically significant at p <0.05.
The baseline IMPROVE HF data set included medical records for 15,381 patients from 167 outpatient cardiology practices. The serum creatinine levels were missing for 2,217 patients, preventing calculation of the eGFR using the Modification of Diet in Renal Disease method and leaving 13,164 patients for the present analysis. The patients were categorized into 1 of 4 groups according to CKD stage, which determined from the eGFR: group 1, stage 1, eGFR of ≥90 ml/min/1.73 m 2 (n = 1,346; 10.2%); group 2, stage 2, eGFR of ≥60 to 89 ml/min/1.73 m 2 (n = 4,941; 37.5%); group 3, stage 3, eGFR 30 to 59 ml/min/1.73 m 2 (n = 5,809; 44.1%); and group 4, stages 4 and 5, eGFR <29 ml/min/1.73 m 2 (n = 1,068; 8.1%).
The demographic and clinical characteristics for each group are listed in Tables 1 and 2 . Older patients, women, and those with an ischemic HF etiology were more likely to have advanced CKD, as were patients with co-morbid conditions such as atrial fibrillation or flutter, diabetes, chronic obstructive pulmonary disease, peripheral vascular disease, hypertension, and prior coronary artery bypass grafting. The New York Heart Association class was greater for patients with more severe CKD. The mean ejection fraction was relatively constant among the 4 groups (≈25%). The blood urea nitrogen levels and B-type natriuretic peptide levels were significantly greater among patients with more severe renal dysfunction. As renal function decreased, serum hemoglobin levels decreased, from 13.7 mg/dl in group 1 to 12.0 mg/dl in group 4 (p <0.001). QRS duration was significantly longer for patients with stage 4–5 CKD than for those with stage 1 (135.5 vs 118.2 ms, p <0.001).
|Characteristic||eGFR (ml/min/1.73 m 2 )||Pairwise Comparison p Value|
|≥90(n = 1,346)||60–89(n = 4,941)||30–59(n = 5,809)||≤29(n = 1,068)||p Value||Stage 2 vs 1||Stage 3 vs 1||Stage 4/5 vs 1|
|Mean ± SD||57.0 ± 14.3||66.2 ± 13.1||72.9 ± 11.0||73.9 ± 11.4|
|Not documented / missing||40.8%||46.1%||46.9%||44.8%|
|Not documented / missing||6.4%||6.2%||6.1%||5.6%|
|Ischemic heart failure etiology||51.7%||62.6%||70.7%||74.2%||<0.001||<0.001||<0.001||<0.001|
|Prior myocardial infarction||36.3%||39.5%||40.6%||40.4%||0.028||0.087||0.009||0.118|
|Prior coronary bypass||21.6%||27.9%||35.9%||40.0%||<0.001||<0.001||<0.001||<0.001|
|History of atrial fibrillation/flutter||18.1%||28.1%||36.3%||36.6%||<0.001||<0.001||<0.001||<0.001|
|Prior peripheral vascular disease||7.3%||9.0%||13.2%||19.7%||<0.001||0.137||<0.001||<0.001|
|Chronic obstructive pulmonary disease||14.0%||15.9%||17.9%||18.2%||0.001||0.261||0.002||0.018|
|Prior percutaneous coronary intervention||22.1%||26.6%||25.3%||25.1%||0.011||0.003||0.043||0.267|
|New York Heart Association class||<0.001||0.032||0.041||<0.001|
|Left ventricular ejection fraction (%)||0.031||1.000||0.611||1.000|
|Mean ± SD||25.5 ± 7.1||25.6 ± 7.0||25.2 ± 7.0||25.3 ± 7.0|
|Body mass index (kg/m 2 )||<0.001||<0.001||<0.001||<0.001|
|Mean ± SD||30.3 ± 7.4||29.1 ± 6.7||27.9 ± 6.2||27.3 ± 6.7|
|Systolic blood pressure (mm Hg)||<0.001||0.405||1.000||1.000|
|Mean ± SD||120.1 ± 18.3||121.1 ± 18.6||119.3 ± 19.1||120.7 ± 20.3|
|Diastolic blood pressure (mm Hg)||<0.001||0.001||<0.001||<0.001|
|Mean ± SD||72.9 ± 11.3||71.6 ± 11.2||68.6 ± 11.0||67.7 ± 12.2|
|Heart rate at rest (beats/min)||<0.001||<0.001||<0.001||0.004|
|Mean ± SD||73.9 ± 12.2||72.0 ± 11.5||71.8 ± 11.2||72.3 ± 11.1|
|Characteristic||eGFR (ml/min/1.73 m 2 )||Pairwise Comparison p Value|
|≥90(n = 1,346)||60–89(n = 4,941)||30–59(n = 5,809)||≤29(n = 1,068)||p Value||Stage 2 vs 1||Stage 3 vs 1||Stage 4 vs 1|
|Mean ± SD||138.8 ± 3.4||139.3 ± 3.3||139.3 ± 4.2||138.8 ± 3.9||<0.001||<0.001||<0.001||1.000|
|Mean ± SD||4.3 ± 1.9||4.4 ± 1.2||4.5 ± 2.5||4.5 ± 1.2||<0.001||1.000||0.003||0.093|
|Blood urea nitrogen (mg/dl)|
|Mean ± SD||15.2 ± 6.4||18.8 ± 6.4||29.3 ± 12.6||52.4 ± 22.1||<0.001||<0.001||<0.001||<0.001|
|Mean ± SD||0.8 ± 0.1||1.1 ± 0.2||1.5 ± 0.3||3.2 ± 1.8||<0.001||<0.001||<0.001||<0.001|
|Mean ± SD||4.1 ± 1.9||4.0 ± 1.1||3.9 ± 0.9||3.7 ± 0.6||<0.001||0.326||<0.001||<0.001|
|Mean ± SD||13.7 ± 1.8||13.7 ± 1.8||13.0 ± 2.0||12.0 ± 1.7||<0.001||1.000||<0.001||<0.001|
|B-type natriuretic peptide (pg/dl)|
|Mean ± SD||446.2 ± 674.8||549.6 ± 795.8||775.1 ± 932.3||1,169.5 ± 1,157.3||<0.001||0.219||<0.001||<0.001|
|QRS duration (ms)|
|Mean ± SD||118.2 ± 35.3||126.5 ± 38.3||133.6 ± 40.2||135.5 ± 42.8||<0.001||<0.001||<0.001||<0.001|