Published online before print January 12, 2009, doi: 10.1161/CIRCULATIONAHA.108.820472
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2009;119:398-407.)
© 2009 American Heart Association, Inc.
Health Services and Outcomes Research |
Patient Health Status and Costs in Heart Failure
Insights From the Eplerenone Post–Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS)
From the Mid America Heart Institute, Kansas City, Mo (P.S.C., P.G.J., J.A.S.); Washington University School of Medicine, St Louis, Mo (G.S.); University of Michigan Medical School, Ann Arbor (B.K.N.); and Christiana Healthcare System, Newark, Del (Z.Z., W.S.W.).
Reprint requests to Paul Chan, MD, MSc, St. Luke’s Mid-America Heart Institute, 5th Floor, 4401 Wornall Rd, Kansas City, MO 64111. E-mail paul1chan{at}yahoo.com
Received May 15, 2008; accepted October 24, 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Background— Although a variety of prognostic tools have been shown to predict rehospitalization and mortality in heart failure patients, their utility in assessing future costs is less clear. We assessed whether health status assessment with the Kansas City Cardiomyopathy Questionnaire (KCCQ) predicts future costs in stable heart failure outpatients with left ventricular dysfunction after myocardial infarction.
Methods and Results— We evaluated 12-month cost utilization data from 1516 heart failure outpatients enrolled in the Quality-of-Life Substudy of the Eplerenone Post–Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS). Multivariable hierarchical models assessed whether the KCCQ (categorized as 0 to <25, 25 to <50, 50 to <75, and 75 to 100) was an independent predictor of future resource use. At baseline, 685 patients (45.2%) had good health status (KCCQ scores 
75), whereas 510 (33.6%), 262 (17.3%), and 59 (3.9%) had fair (KCCQ, 50 to 74), poor (KCCQ, 25 to 49), and the worst (KCCQ <25) health status, respectively. After multivariable adjustment, compared with patients with good health status, patients with fair health status incurred incremental 1-year costs of $1520 (cost ratio, 1.23; 95% confidence interval, 1.05 to 1.43), whereas patients with poor and the worst health status incurred incremental 1-year costs of $4265 (cost ratio, 1.63; 95% confidence interval, 1.34 to 1.99) and $8999 (cost ratio, 2.34; 95% confidence interval, 1.62 to 3.38), respectively (P<0.0001 for association with KCCQ). Further adjustment for New York Heart Association class led to only partial attenuation of this relationship (P=0.0002).
Conclusion— Health status assessment predicts resource use and costs over the next year in stable heart failure outpatients with left ventricular dysfunction after myocardial infarction.
Key Words: costs and cost analysis • health status • heart failure • prognosis
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Heart failure is a prevalent chronic condition associated with high morbidity and mortality. In the United States, >5 million people have diagnosed heart failure, and 550 000 incident cases occur annually, accounting for >1 million hospitalizations and 285 000 deaths each year.1 The annual direct and indirect costs of heart failure are estimated to exceed $33 billion annually in the United States1 and account for 1% to 2% of all healthcare expenditures in developed countries.2 Although a variety of biomarkers, noninvasive screening tests, and clinical risk factors have been shown to predict rehospitalization and mortality in heart failure patients,3–13 their utility in assessing healthcare costs has not been described. Predicting subsequent costs is especially important in heart failure in that it may help to identify high-utilization patients who may benefit from targeted disease management programs.
Editorial p 368
Clinical Perspective p 407
The Kansas City Cardiomyopathy Questionnaire (KCCQ) is a well-validated tool for evaluating clinical health status and quality of life in patients with heart failure.11–13 The KCCQ is inexpensive and easy to administer, offers patient-centered insights unavailable with traditional risk factors or screening tests, and can be serially monitored.14 Although the KCCQ predicts future clinical outcomes, its ability to forecast future costs is unknown. If the KCCQ predicts both clinical outcomes and future costs, clinicians and third-party payers alike may be able to use this simple, noninvasive screen routinely to identify patients with the greatest need for intensive disease management interventions.
Accordingly, we examined the relationship of KCCQ scores among heart failure outpatients and 1-year costs using the Eplerenone Post–Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS), a randomized multicenter trial of aldosterone blockade in patients with acute myocardial infarction (MI) complicated by heart failure and left ventricular dysfunction.15 Because of the detailed clinical and cost information collected, serial measurements of patients’ health status, and close trial follow-up,16 EPHESUS represents an excellent opportunity to address this current gap in knowledge.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Study Population
EPHESUS was an international, multicenter, randomized trial that evaluated the effect of the selective aldosterone blocker eplerenone on morbidity and mortality among patients with acute MI complicated by heart failure and left ventricular dysfunction between December 27, 1999, and December 31, 2001. Its study design has been previously described.15 Briefly, 6632 patients with confirmed MI, left ventricular ejection fraction
40%, and either clinical heart failure (as demonstrated by pulmonary rales, venous congestion on chest x-ray, or presence of a third heart sound) or diabetes mellitus were randomized to eplerenone or placebo and were subsequently followed up for a mean of 16 months. For the purposes of this study, because we were interested primarily in predicting future healthcare costs among study participants with stabilized heart failure, we included only those patients from the 10 prespecified countries participating in the EPHESUS quality-of-life substudy who survived their index hospitalization to the 1-month baseline (n=2191).16 The substudy involved serial evaluations of patient health status with the KCCQ, a self-administered 23-item instrument with established validity, reliability, and prognostic utility.14 We excluded patients enrolled in EPHESUS with diabetes mellitus but no clinical heart failure (n=292) and those who did not complete a 1-month KCCQ (233 patients lost to follow-up before 1 month; 147 patients without a 1-month visit, and 3 patients with an incomplete KCCQ), so our final study population comprised 1516 patients.
Health Status Assessment
Patient health status assessment was performed at the 1-month visit with the KCCQ. The KCCQ evaluates distinct health status domains for heart failure (physical limitations, symptoms, social function, self-efficacy, and quality of life), and its value over non–disease-specific instruments has been previously demonstrated.13,14 An overall score based on contributions from each domain quantifies the multiple domains of the KCCQ into a single summary score. Scores from the KCCQ are transformed to a range from 0 to 100, with higher scores denoting better health status. Assessment with the KCCQ also was performed at the 3-month follow-up in 1358 patients (90% of those who completed the 1-month assessment).
Cost Assessment
The EPHESUS economic substudy, which prospectively evaluated direct costs incurred during the follow-up period using a predefined algorithm, has been previously described.17 Briefly, costs from all participating countries were determined by quantifying resource use, applying the US Medicare Fee Schedule, and standardizing costs to US dollars for the year 2001. For inpatient hospitalizations during follow-up, an investigator blinded to patient characteristics and treatment group assigned diagnosis-related groups to each hospitalization. Costs were then estimated from average Medicare reimbursement rates obtained from the Medicare Part A file.18 Outpatient procedures and services were similarly captured in a blinded fashion with trial case report forms. Procedural codes were used to determine Medicare reimbursement rates for procedures, and physician professional costs were estimated as a percentage share of the hospital costs according to the diagnosis-related group.19 Finally, average wholesale prices were used to determine outpatient medication costs.20 In addition, because mortality rates may differ by KCCQ score, we accounted for the societal cost of death (US $8000) derived from prior cost-effectiveness studies to ensure no bias from competing risks because patients who die earlier have less exposure time to incur subsequent costs.18,21,22
Independent Variable
The independent variable of interest was the KCCQ at 1 month after enrollment in EPHESUS. In a secondary analysis, the KCCQ at 3 months was used as the baseline to demonstrate reproducibility. Based on prior work, the KCCQ was evaluated categorically by prespecified cut points (0 to <25, 25 to <50, 50 to <75, and 75 to 100) and as a continuous variable.11,16,23
Outcome Assessment
The main outcome was 1-year costs from the 1-month baseline assessment (ie, from month 1 to month 13 after randomization). All-cause mortality rates during this 1-year period also were examined and accounted for in calculating costs, as described above. All mortality end points were adjudicated by a blinded critical events committee.15 To examine the reproducibility of our findings, we repeated our analyses of 1-year costs from the 3-month assessment (from month 3 to month 15 after randomization).
Statistical Analysis
Baseline characteristics were compared across KCCQ categories through the use of Mantel-Haenszel trend tests for categorical variables and ANOVA linear trend tests for continuous variables. Kaplan-Meier survival curves were used to examine the unadjusted relationship between KCCQ categories and 1-year death from all causes.
Sequential multivariable hierarchical regression models were then constructed to assess whether the prognostic impact of baseline KCCQ on future cost was independent of other patient characteristics. Because of the skewness of the cost data, generalized linear models were used, which allow for nonnormal distributions. Our models used a 
error distribution and log link for expected costs, which were found by residual plots to fit the data reasonably well. Because a log link was used, cost comparisons are expressed in relative rather than absolute differences.
All models included adjustment for treatment randomization group and country as fixed effects and study enrollment center as a random effect. The initial model included only KCCQ categories and study center. The second model added demographic characteristics (age, gender, and race), medical comorbidities (prior heart failure, MI, or atrial fibrillation; prior anginal symptoms; hypertension; dyslipidemia; diabetes mellitus; stroke; and chronic lung disease), disease severity at the index hospitalization (pulmonary edema, Killip class, left ventricular ejection fraction, presence of Q-wave MI, and mode of reperfusion therapy), study measures at 1 month (body mass index, vital signs [heart rate, systolic and diastolic blood pressures], and laboratory values [sodium, estimated glomerular filtration rate using the Cockcroft-Gault formula]), and medications at 1 month (aspirin, thienopyridines, angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers, β-blockers, diuretics, and 3-hydroxy-3-methyl-glutaryl-CoA reductase inhibitors). Continuous variables were assessed for nonlinearity and were transformed when appropriate through the use of restricted cubic splines. A third and final model included all the aforementioned covariates and adjusted further for New York Heart Association (NYHA) class to determine whether the prognostic utility of the KCCQ was independent of all clinical markers of heart failure severity. Least-squares mean estimates of cost by KCCQ groups were calculated, along with cost ratios and confidence intervals (CIs) versus the reference group of 75 to 100 and P values for linear trends across KCCQ groups. Collinearity diagnostics were computed to determine the stability of effect estimates in the regression models. We also assessed the KCCQ as a continuous variable, including the use of 5-knot restricted cubic splines to examine whether a "cutoff" existed at which the association of KCCQ costs was stronger (ie, if P<0.05 for test for nonlinearity).
Of the 1516 patients in the study cohort, 1141 (75%) were alive at the end of 1 year, 130 (9%) died, and 245 (16%) were censored before completion of the 1-year follow-up. Censoring was due almost entirely to early termination of the study (n=239), with only 6 patients censored because of lost follow-up. The mean±SD time to follow-up among censored patients was 10±2 months and did not differ by KCCQ group (P=0.15). To examine the effect of censoring on estimates of 1-year costs, we used the method of Zhao and Tian24 and Zhao et al25 that weights observations by the inverse probability of being censored and integrates over patients’ cost histories to correct for censoring bias. This analysis revealed a negligible bias resulting from censoring (–$167) that also did not vary by KCCQ group (P=0.98). Similar results were found for 1-year costs after the 3-month visit. Therefore, the results presented here are not corrected for censoring.
Data on at least one of the covariates were missing for 7% of patients (only 10 of the 32 covariates had any missing values; missing rates ranged from <0.1% to 3%). We used multiple imputation methods to impute missing values for these covariates on the basis of all other observed data. Imputations were performed with Markov Chain Monte Carlo methods as implemented in SAS PROC MI. Five imputed data sets were generated; analyses were replicated across data sets and pooled to obtain final estimates. The results using these imputed values were virtually identical to those obtained on patients with the complete data; the former are presented here.
To explicitly examine direct medical costs, we repeated the analyses without attributing costs for death. To demonstrate the reproducibility of our findings, models were replicated using KCCQ scores from the 3-month visit (instead of the 1-month visit) and 1-year costs from this visit (costs from 3 to 15 months after randomization).
For all analyses, the null hypothesis was evaluated at a 2-sided significance level of 0.05 with 95% CIs calculated. All analyses were conducted with SAS 9.1 (SAS Institute, Inc, Cary, NC) and R version 2.6.2.26 The institutional review board or ethics committee at each site involved in the EPHESUS trial approved the protocol, and all patients provided written informed consent before enrollment.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Of the initial 1516 patients in the study cohort, 685 (45.2%) had good health status (KCCQ score, 75 to 100), 510 (33.6%) had fair health status (KCCQ score, 50 to <75), 262 (17.3%) had poor health status (KCCQ score, 25 to <50), and 59 (3.9%) had the worst health status (KCCQ, 0 to <25) 1 month after hospitalization. In general, medical therapy for the entire heart failure cohort and across all 4 KCCQ groups was excellent. Patients with lower health status were older and more likely to be female; more likely to have prior heart failure, MI, and angina, as well as coexisting diabetes, cerebrovascular disease, and chronic lung disease; and more likely to have higher Killip class, a lower left ventricular ejection fraction, and no revascularization procedures performed on initial presentation (Table 1
). At the 1-month follow-up, patients with lower heath status also had higher resting heart rates, lower diastolic blood pressures, lower serum sodium levels, and lower renal glomerular filtration rates; were more likely to have been hospitalized since discharge; were less likely to be taking aspirin and β-blockers; and were more likely to be using diuretic agents.
|
|
Kaplan-Meier survival curves showed that an inverse graded relationship existed between the KCCQ score and mortality from all causes (P<0.001) (Figure 1). In unadjusted analysis, patients with good health status (KCCQ 75) incurred the lowest 1-year costs (mean, $9628; median, $4190; interquartile range, $2659 to $10 623), whereas patients with the worst health status (KCCQ <25) incurred the highest costs (mean, $18 476; median, $10 383; interquartile range, $4371 to $25 760) (Table 2). Greater healthcare resource use among those with worse health status was due largely to both higher rates of hospitalization and longer lengths of stay for heart failure and other cardiovascular causes (Table 3). This pattern of resource use across KCCQ categories remained proportionate throughout the 12 months of follow-up (Figure 2).
|
|
|
|
After adjustment for country, enrollment site, and all clinical covariates except NYHA class, an inverse graded relationship was found between KCCQ scores and 1-year costs (P<0.0001) (Table 4). Compared with patients with good health status (KCCQ of 75), patients with fair health status (KCCQ, 50 to 74) incurred incremental 1-year costs of $1520 (cost ratio, 1.23; 95% CI, 1.05 to 1.43). In contrast, patients with poor (KCCQ, 25 to 49) and the worst (KCCQ <25) health status incurred incremental 1-year costs of $4265 (cost ratio, 1.63; 95% CI, 1.34 to 1.99) and $8999 (cost ratio, 2.34; 95% CI, 1.62 to 3.38), respectively. Further adjustment for NYHA class led to only partial attenuation of the effect size, and the KCCQ remained a strong independent predictor of 1-year costs (P=0.0002). In addition, these relationships remained essentially unchanged when we excluded costs for death in our analyses (Table 4). Table 5 lists significant predictors of 1-year costs (P<0.05) in the final fully adjusted model. Variance inflation factors (which measure the degree to which variances are inflated because of collinearity with other model variables) were 1.5 for KCCQ and between 1.1 and 2.0 for the other significant predictors, well below the threshold of 10 that is typically used to signal multicollinearity concerns.27
|
|
When the KCCQ was examined as a continuous variable, in the full model adjusted for all covariates and NYHA class, each 5-point-lower baseline 1-month KCCQ score was associated with a 4.0% increase in 1-year costs (95% CI, 2.2% to 5.9%; P<0.0001). The association was linear on the log scale (P=0.52 for nonlinearity by restricted cubic splines), indicating that the 4.0% rate was constant across the range of KCCQ scores without a discernible cut point at which the association between KCCQ and costs was attenuated. Finally, when we used the 3-month KCCQ as the baseline and followed up patients for 1-year costs from this visit, our findings were essentially unchanged (see Tables 2 through 4
).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Our findings demonstrate that patient health status assessments with the KCCQ can identify heart failure outpatients with left ventricular dysfunction after MI who are likely to have high resource use over the following year despite excellent medical therapy. We found a graded inverse relationship between health status and 1-year costs, with incremental annual costs among patients with the worst health status being nearly $9000 greater than those for patients with the best health status. The KCCQ remained predictive of costs even after adjustment for NHYA class, suggesting that health status, from a patient-centered perspective, carries significant incremental prognostic information. Moreover, we found that these associations between health status and 1-year costs were reproducible at a subsequent outpatient visit. Collectively, these findings suggest that the KCCQ can be used as an effective serial monitoring instrument to help identify post-MI heart failure patients with left ventricular dysfunction likely to have high resource use over the next year.
Although the vast majority of prognostic tools in heart failure have focused on clinical end points such as mortality and rehospitalization,3–13 the ability of these tools to predict costs to society is less clear. Because of competing risks, it is important to demonstrate that a prediction tool has prognostic utility for both clinical outcomes and costs because it is plausible that patients with chronic diseases who are more likely to die may actually incur fewer costs than patients who remain alive. In the present study, we were able to demonstrate that patients with poorer health status incurred significantly higher costs, primarily because of a greater frequency and duration of inpatient hospitalization.
The ability to prognosticate cost in heart failure is of interest to clinicians and payers alike and would allow more appropriate targeting of interventions with established efficacy. Heart failure hospitalizations have increased nearly 3-fold in the past 2 decades, and it has been estimated that 55% of these admissions may be preventable.28–30 Intensive disease management programs in heart failure have been shown to improve use and to decrease morbidity and mortality. These programs remain vastly underused, however. The reasons for this underuse include the labor-intensive process of patient education, empowerment, and monitoring; the incremental cost to hospitals for implementing disease management programs that, although cost-effective, are not cost neutral; and the loss of future revenue to hospitals by successfully preventing readmissions.31 By using a simple and inexpensive tool to identify heart failure patients likely to have high subsequent resource use, clinicians can focus limited disease management resources on the sickest patients who are the most likely to benefit. Such an application may also improve the cost-effectiveness of these programs by targeting those with the greatest risk for high costs. Because it can also be repeated at subsequent visits, the KCCQ would allow clinicians to make real-time decisions in the outpatient setting about which patients may benefit from intensified heart failure management.
Current national guidelines recommend that "an ongoing review of the patient’s clinical status is critical to the appropriate selection and monitoring of treatments,"32 and the use of standardized instruments such as the KCCQ has been cited as a marker of high-quality care.33 The KCCQ is simple to administer and score, is inexpensive and noninvasive, uses standardized questionnaires with minimal interobserver variability compared with conventional measures such as NYHA class, and provides insights into patient-centered domains such as self-efficacy, quality of life, and social limitations.14 With its ability to predict both costs and outcomes and to be administered repeatedly, the KCCQ could allow close monitoring of heart failure patients with left ventricular dysfunction after MI in whom prognostically important health status changes can be feasibly and readily identified, interpreted, and acted on.
Our study should be interpreted in the context of several potential limitations. First, there remains the possibility of residual confounding despite adjustment for a wide variety of demographic, clinical, and patient factors. Second, EPHESUS enrolled patients with symptomatic heart failure and left ventricular dysfunction after acute MI. Whether the KCCQ has similar prognostic significance in heart failure patients with nonischemic origins or preserved left ventricular ejection fraction requires further investigation. Third, our cost analysis evaluated costs associated with inpatient and outpatient utilization, but we did not examine other indirect costs such as loss of productivity or costs associated with heart failure specialty clinics. However, omission of these costs would likely have biased our estimates of effect toward the null, and we may have underestimated the true societal cost difference between those with poor and those with good health status. Fourth, we could not determine what proportion of the cost difference was due to a lower threshold among physicians (eg, primary care physicians or emergency room physicians) to hospitalize patients with worse baseline health status. However, we believe that any threshold differences, should they exist, are likely to reflect the routine practice of medicine in clinical decision making. Fifth, our analyses did not include the cost of device therapy, which was not standard therapy for patients with heart failure and left ventricular dysfunction at the time that the EPHESUS study was conducted. Finally, although our findings suggest an association between those with the lowest KCCQ scores and those with higher costs, whether intensified disease management programs in these high-risk patients can reduce costs remains to be established.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Health status assessment predicts resource use in post-MI heart failure patients with left ventricular dysfunction. Serial assessments of patients’ health status in clinical practice may help identify high-risk patients who may benefit from intensified disease management programs.
|
Acknowledgments |
|---|
Disclosures
Dr Spertus developed and owns the copyrights for the KCCQ, is a consultant for Amgen, and has received research grant support from Amgen and Medtronic. Dr Weintraub receives grant support from Pfizer. The other authors report no conflicts.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
1. Rosamond W, Flegal K, Friday G, Furie K, Go A, Greenlund K, Haase N, Ho M, Howard V, Kissela B, Kittner S, Lloyd-Jones D, McDermott M, Meigs J, Moy C, Nichol G, O'Donnell CJ, Roger V, Rumsfeld J, Sorlie P, Steinberger J, Thom T, Wasserthiel-Smoller S, Hong Y. Heart disease and stroke statistics–2007 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2007; 115: e69–e171.
2. Berry C, Murdoch DR, McMurray JJ. Economics of chronic heart failure. Eur J Heart Fail. 2001; 3: 283–291.
3. Levy WC, Mozaffarian D, Linker DT, Sutradhar SC, Anker SD, Cropp AB, Anand I, Maggioni A, Burton P, Sullivan MD, Pitt B, Poole-Wilson PA, Mann DL, Packer M. The Seattle Heart Failure Model: prediction of survival in heart failure. Circulation. 2006; 113: 1424–1433.
4. Benedict CR, Shelton B, Johnstone DE, Francis G, Greenberg B, Konstam M, Probstfield JL, Yusuf S. Prognostic significance of plasma norepinephrine in patients with asymptomatic left ventricular dysfunction: SOLVD Investigators. Circulation. 1996; 94: 690–697.
5. Bettencourt P, Ferreira A, Dias P, Pimenta J, Frioes F, Martins L, Cerqueira-Gomes M. Predictors of prognosis in patients with stable mild to moderate heart failure. J Card Fail. 2000; 6: 306–313.[CrossRef][Medline] [Order article via Infotrieve]
6. Brooksby P, Batin PD, Nolan J, Lindsay SJ, Andrews R, Mullen M, Baig W, Flapan AD, Prescott RJ, Neilson JM, Cowley AJ, Fox KA. The relationship between QT intervals and mortality in ambulant patients with chronic heart failure: the United Kingdom Heart Failure Evaluation and Assessment of Risk Trial (UK-HEART). Eur Heart J. 1999; 20: 1335–1341.
7. Shah MR, Hasselblad V, Gheorghiade M, Adams KF Jr, Swedberg K, Califf RM, O'Connor CM. Prognostic usefulness of the six-minute walk in patients with advanced congestive heart failure secondary to ischemic or nonischemic cardiomyopathy. Am J Cardiol. 2001; 88: 987–993.[CrossRef][Medline] [Order article via Infotrieve]
8. Chow T, Kereiakes DJ, Bartone C, Booth T, Schloss EJ, Waller T, Chung E, Menon S, Nallamothu BK, Chan PS. Prognostic utility of microvolt T-wave alternans in risk stratifying patients with ischemic cardiomyopathy. J Am Coll Cardiol. 2006; 47: 1820–1827.
9. Vrtovec B, Delgado R, Zewail A, Thomas CD, Richartz BM, Radovancevic B. Prolonged QTc interval and high B-type natriuretic peptide levels together predict mortality in patients with advanced heart failure. Circulation. 2003; 107: 1764–1769.
10. Bloomfield DM, Steinman RC, Namerow PB, Parides M, Davidenko J, Kaufman ES, Shinn T, Curtis A, Fontaine J, Holmes D, Russo A, Tang C, Bigger JT Jr. Microvolt T-wave alternans distinguishes between patients likely and patients not likely to benefit from implanted cardiac defibrillator therapy: a solution to the Multicenter Automatic Defibrillator Implantation Trial (MADIT) II conundrum. Circulation. 2004; 110: 1885–1889.
11. Heidenreich PA, Spertus JA, Jones PG, Weintraub WS, Rumsfeld JS, Rathore SS, Peterson ED, Masoudi FA, Krumholz HM, Havranek EP, Conard MW, Williams RE. Health status identifies heart failure outpatients at risk for hospitalization or death. J Am Coll Cardiol. 2006; 47: 752–756.
12. Luther SA, McCullough PA, Havranek EP, Rumsfeld JS, Jones PG, Heidenreich PA, Peterson ED, Rathore SS, Krumholz HM, Weintraub WS, Spertus JA, Masoudi FA. The relationship between B-type natriuretic peptide and health status in patients with heart failure. J Card Fail. 2005; 11: 414–421.[CrossRef][Medline] [Order article via Infotrieve]
13. Spertus J, Peterson E, Conard MW, Heidenreich PA, Krumholz HM, Jones P, McCullough PA, Pina I, Tooley J, Weintraub WS, Rumsfeld JS. Monitoring clinical changes in patients with heart failure: a comparison of methods. Am Heart J. 2005; 150: 707–715.[CrossRef][Medline] [Order article via Infotrieve]
14. Green CP, Porter CB, Bresnahan DR, Spertus JA. Development and evaluation of the Kansas City Cardiomyopathy Questionnaire: a new health status measure for heart failure. J Am Coll Cardiol. 2000; 35: 1245–1255.
15. Pitt B, Remme W, Zannad F, Neaton J, Martinez F, Roniker B, Bittman R, Hurley S, Kleiman J, Gatlin M. Eplerenone, a selective aldosterone blocker, in patients with left ventricular dysfunction after myocardial infarction. N Engl J Med. 2003; 348: 1309–1321.
16. Spertus JA, Tooley J, Jones P, Poston C, Mahoney E, Deedwania P, Hurley S, Pitt B, Weintraub WS. Expanding the outcomes in clinical trials of heart failure: the quality of life and economic components of EPHESUS (Eplerenone’s Neurohormonal Efficacy and Survival Study). Am Heart J. 2002; 143: 636–642.[CrossRef][Medline] [Order article via Infotrieve]
17. Weintraub WS, Zhang Z, Mahoney EM, Kolm P, Spertus JA, Caro J, Ishak J, Goldberg R, Tooley J, Willke R, Pitt B. Cost-effectiveness of eplerenone compared with placebo in patients with myocardial infarction complicated by left ventricular dysfunction and heart failure. Circulation. 2005; 111: 1106–1113.
18. Gold M, Siegel J, Russell L, Weinstein M. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996.
19. Mitchell JB, Burge RT, Lee AJ. Per case prospective payment for episodes of hospital care: final report to HCFA for Master Contract No. 500-92-0020. October 6, 1995.
20. Drug Topics Red Book. Montvale, NJ: Medical Economics Company; 2000.
21. Gage BF, Cardinalli AB, Albers GW, Owens DK. Cost-effectiveness of warfarin and aspirin for prophylaxis of stroke in patients with nonvalvular atrial fibrillation. JAMA. 1995; 274: 1839–1845.
22. Eckman MH, Falk RH, Pauker SG. Cost-effectiveness of therapies for patients with nonvalvular atrial fibrillation. Arch Intern Med. 1998; 158: 1669–1677.
23. Rumsfeld JS, Havranek E, Masoudi FA, Peterson ED, Jones P, Tooley JF, Krumholz HM, Spertus JA. Depressive symptoms are the strongest predictors of short-term declines in health status in patients with heart failure. J Am Coll Cardiol. 2003; 42: 1811–1817.
24. Zhao H, Tian L. On estimating medical cost and incremental cost-effectiveness ratios with censored data. Biometrics. 2001; 57: 1002–1008.[CrossRef][Medline] [Order article via Infotrieve]
25. Zhao H, Bang H, Wang H, Pfeifer PE. On the equivalence of some medical cost estimators with censored data. Stat Med. 2007; 26: 4520–4530.[CrossRef][Medline] [Order article via Infotrieve]
26. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing V. Available at: http://www.R-project.org. Accessed April 16, 2008.
27. Myers R. Classical and Modern Regression With Applications. 2nd ed. Boston, Mass: PWS-Kent; 1990.
28. Michalsen A, Konig G, Thimme W. Preventable causative factors leading to hospital admission with decompensated heart failure. Heart. 1998; 80: 437–441.
29. Tsuyuki RT, McKelvie RS, Arnold JM, Avezum A Jr, Barretto AC, Carvalho AC, Isaac DL, Kitching AD, Piegas LS, Teo KK, Yusuf S. Acute precipitants of congestive heart failure exacerbations. Arch Intern Med. 2001; 161: 2337–2342.
30. Opasich C, Rapezzi C, Lucci D, Gorini M, Pozzar F, Zanelli E, Tavazzi L, Maggioni AP. Precipitating factors and decision-making processes of short-term worsening heart failure despite "optimal" treatment (from the IN-CHF Registry). Am J Cardiol. 2001; 88: 382–387.[CrossRef][Medline] [Order article via Infotrieve]
31. Seow H, Phillips CO, Rich MW, Spertus JA, Krumholz HM, Lynn J. Isolation of health services research from practice and policy: the example of chronic heart failure management. J Am Geriatr Soc. 2006; 54: 535–540.[CrossRef][Medline] [Order article via Infotrieve]
32. Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats TG, Jessup M, Konstam MA, Mancini DM, Michl K, Oates JA, Rahko PS, Silver MA, Stevenson LW, Yancy CW, Antman EM, Smith SC Jr, Adams CD, Anderson JL, Faxon DP, Fuster V, Halperin JL, Hiratzka LF, Jacobs AK, Nishimura R, Ornato JP, Page RL, Riegel B. ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure): developed in collaboration with the American College of Chest Physicians and the International Society for Heart and Lung Transplantation: endorsed by the Heart Rhythm Society. Circulation. 2005; 112: e154–e235.
33. Bonow RO, Bennett S, Casey DE Jr, Ganiats TG, Hlatky MA, Konstam MA, Lambrow CT, Normand SL, Pina IL, Radford MJ, Smith AL, Stevenson LW, Burke G, Eagle KA, Krumholz HM, Linderbaum J, Masoudi FA, Ritchie JL, Rumsfeld JS, Spertus JA; American College of Cardiology; American Heart Association Task Force on Performance Measures; Heart Failure Society of America. ACC/AHA clinical performance measures for adults with chronic heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures (Writing Committee to Develop Heart Failure Clinical Performance Measures): endorsed by the Heart Failure Society of America. Circulation. 2005; 112: 1853–1887.
|
Footnotes |
|---|
Guest Editor for this article was Paul W. Armstrong, MD.
Related Article:
Circulation 2009 119: 359-361.
This article has been cited by other articles:
 |
|
 |
|
  S. V. Arnold, D. A. Morrow, Y. Lei, D. J. Cohen, E. M. Mahoney, E. Braunwald, and P. S. Chan Economic Impact of Angina After an Acute Coronary Syndrome: Insights From the MERLIN-TIMI 36 Trial Circ Cardiovasc Qual Outcomes, July 1, 2009; 2(4): 344 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 How Are You Doing? Journal Watch Cardiology, February 4, 2009; 2009(204): 2 - 2. [Full Text]
|
|
|
|
 |
|
 J. A. Ezekowitz What Price to Pay?: The Cost of Everything and the Value of Nothing Circulation, January 27, 2009; 119(3): 368 - 370. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||