Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Long-Term Outcomes of Medicare Beneficiaries With Worsening Renal Function During Hospitalization for Heart Failure

2010; Elsevier BV; Volume: 105; Issue: 12 Linguagem: Inglês

10.1016/j.amjcard.2010.01.361

ISSN

1879-1913

Autores

Robb D. Kociol, Melissa A. Greiner, Bradley G. Hammill, Hemant Phatak, Gregg C. Fonarow, Lesley H. Curtis, Adrian F. Hernandez,

Tópico(s)

Acute Kidney Injury Research

Resumo

We examined whether worsening renal function (RF) was associated with long-term mortality, readmission, and inpatient costs in Medicare beneficiaries hospitalized with heart failure (HF). Baseline renal insufficiency in patients hospitalized for HF is associated with increased risk of morbidity and mortality. However, the relation between worsening RF and long-term clinical outcomes is unclear. We linked clinical registry data to Medicare inpatient claims to identify 1-year outcomes of patients ≥65 years of age hospitalized with HF. Worsening RF was defined as a change in serum creatinine ≥0.3 mg/dl. Relations between worsening RF and 1-year mortality and readmission were evaluated with multivariable Cox proportional hazards models with robust SEs; associations with inpatient costs were evaluated with generalized linear models with a log-link and Poisson distribution. Of 20,063 patients hospitalized with HF and discharged alive, 3,581 (17.8%) had worsening RF during the index hospitalization. One year after discharge, 35.4% of these patients died, 64.5% were readmitted, and average costs at 1 year were $14,829 (interquartile range 0 to 19,366). After adjustment for patient characteristics, baseline RF, and comorbid conditions, worsening RF was independently associated with 1-year mortality (hazard ratio 1.12, 95% confidence interval 1.04 to 1.20) but not readmission or total inpatient costs. In conclusion, worsening RF in patients hospitalized with HF was independently associated with long-term mortality. We examined whether worsening renal function (RF) was associated with long-term mortality, readmission, and inpatient costs in Medicare beneficiaries hospitalized with heart failure (HF). Baseline renal insufficiency in patients hospitalized for HF is associated with increased risk of morbidity and mortality. However, the relation between worsening RF and long-term clinical outcomes is unclear. We linked clinical registry data to Medicare inpatient claims to identify 1-year outcomes of patients ≥65 years of age hospitalized with HF. Worsening RF was defined as a change in serum creatinine ≥0.3 mg/dl. Relations between worsening RF and 1-year mortality and readmission were evaluated with multivariable Cox proportional hazards models with robust SEs; associations with inpatient costs were evaluated with generalized linear models with a log-link and Poisson distribution. Of 20,063 patients hospitalized with HF and discharged alive, 3,581 (17.8%) had worsening RF during the index hospitalization. One year after discharge, 35.4% of these patients died, 64.5% were readmitted, and average costs at 1 year were $14,829 (interquartile range 0 to 19,366). After adjustment for patient characteristics, baseline RF, and comorbid conditions, worsening RF was independently associated with 1-year mortality (hazard ratio 1.12, 95% confidence interval 1.04 to 1.20) but not readmission or total inpatient costs. In conclusion, worsening RF in patients hospitalized with HF was independently associated with long-term mortality. Patients admitted with acute heart failure (HF) may have impaired hemodynamics, maladaptive neurohumoral effects, and increased circulating inflammatory cytokines that influence cardiovascular function and renal function (RF).1Ferguson D.W. Berg W.J. Roach P.J. Oren R.M. Mark A.L. Effects of heart failure on baroreflex control of sympathetic neural activity.Am J Cardiol. 1992; 69: 523-531Abstract Full Text PDF PubMed Scopus (177) Google Scholar, 2Schrier R.W. Abraham W.T. Hormones and hemodynamics in heart failure.N Engl J Med. 1999; 341: 577-585Crossref PubMed Scopus (978) Google Scholar, 3Francis G.S. Siegel R.M. Goldsmith S.R. Olivari M.T. Levine T.B. Cohn J.N. Acute vasoconstrictor response to intravenous furosemide in patients with chronic congestive heart failure Activation of the neurohumoral axis.Ann Intern Med. 1985; 103: 1-6Crossref PubMed Scopus (487) Google Scholar Patients with HF often have renal insufficiency at the time of admission or develop worsening RF during hospitalization. Patients with worsening RF during a HF hospitalization have worse inpatient outcomes, including greater rates of mortality and complications, longer length of stay, and higher costs. Previous evaluations of the impact of worsening RF during HF admissions have focused primarily on short-term outcomes or have been conducted in the context of clinical trials, which do not necessarily reflect real-world practice.4Dries D.L. Sweitzer N.K. Drazner M.H. Stevenson L.W. Gersh B.J. The prognostic implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction.J Am Coll Cardiol. 2000; 35: 681-689Abstract Full Text Full Text PDF PubMed Scopus (733) Google Scholar, 5Forman D.E. Butler J. Wang Y. Abraham W.T. O'Connor C.M. Gottlieb S.S. Loh E. Massie B.M. Rich M.W. Stevenson L.W. Young J.B. Krumholz H.M. Incidence, predictors at admission, and impact of worsening renal function among patients hospitalized with heart failure.J Am Coll Cardiol. 2004; 43: 61-67Abstract Full Text Full Text PDF PubMed Scopus (751) Google Scholar, 6Hillege H.L. Nitsch D. Pfeffer M.A. Swedberg K. McMurray J.J. Yusuf S. Granger C.B. Michelson E.L. Ostergren J. Cornel J.H. de Zeeuw D. Pocock S. van Veldhuisen D.J. Candesartan in Heart Failure: Assessment of Reduction In Mortality and Morbidity (CHARM) InvestigatorsRenal function as a predictor of outcome in a broad spectrum of patients with heart failure.Circulation. 2006; 113: 671-678Crossref PubMed Scopus (769) Google Scholar, 7Nohria A. Hasselblad V. Stebbins A. Pauly D.F. Fonarow G.C. Shah M. Yancy C.W. Califf R.M. Stevenson L.W. Hill J.A. Cardiorenal interactions: insights from the ESCAPE trial.J Am Coll Cardiol. 2008; 51: 1268-1274Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar, 8Weinfeld M.S. Chertow G.M. Stevenson L.W. Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure.Am Heart J. 1999; 138: 285-290Abstract Full Text Full Text PDF PubMed Scopus (231) Google Scholar The relation between worsening RF and long-term outcomes is poorly understood, especially in older patients hospitalized with HF. We used data from the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients With Heart Failure (OPTIMIZE-HF) registry linked to Medicare claims data to evaluate associations between worsening RF and postdischarge mortality, readmission, and inpatient costs at 1 year. We accessed clinical data for the study from the OPTIMIZE-HF registry. We also obtained research-identifiable Medicare claims data from the Centers for Medicare and Medicaid Services. The OPTIMIZE-HF registry contains data for patients admitted with HF from January 1, 2003, through December 31, 2004, at 259 participating hospitals.8Weinfeld M.S. Chertow G.M. Stevenson L.W. Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure.Am Heart J. 1999; 138: 285-290Abstract Full Text Full Text PDF PubMed Scopus (231) Google Scholar, 9Fonarow G.C. Abraham W.T. Albert N.M. Gattis W.A. Gheorghiade M. Greenberg B. O'Connor C.M. Yancy C.W. Young J. Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF): rationale and design.Am Heart J. 2004; 148: 43-51Abstract Full Text Full Text PDF PubMed Scopus (277) Google Scholar, 10Fonarow G.C. Lukas M.A. Robertson M. Colucci W.S. Dargie H.J. Effects of carvedilol early after myocardial infarction: analysis of the first 30 days in Carvedilol Post-Infarct Survival Control in Left Ventricular Dysfunction (CAPRICORN).Am Heart J. 2007; 154: 637-644Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar Eligible patients were those who presented with symptoms of HF during a hospitalization for which HF was the primary discharge diagnosis or for whom worsening HF was the primary reason for hospital admission. The registry included hospitals in a range of sizes throughout the United States, and Medicare beneficiaries enrolled in OPTIMIZE-HF are similar to the broader Medicare population with HF.11Curtis L.H. Greiner M.A. Hammill B.G. DiMartino L.D. Shea A.M. Hernandez A.F. Fonarow G.C. Representativeness of a national heart failure quality-of-care registry: comparison of OPTIMIZE-HF and non-OPTIMIZE-HF Medicare patients.Circ Cardiovasc Qual Outcomes. 2009; 2: 377-384Crossref PubMed Scopus (75) Google Scholar Variables available for this study included gender, race, date of birth, hospital admission and discharge dates, American Hospital Association hospital identifier, medical history variables, and laboratory measurements and pharmacy indicators from admission and discharge. For race, we used the category "black" and combined all others as "nonblack." Medicare claims data in this study included inpatient claims and corresponding denominator files for all Medicare beneficiaries discharged from a hospitalization from 2002 through 2005. These files included institutional claims for facility costs covered by Medicare Part A; beneficiary, physician, and hospital identifiers; admission and discharge dates; and International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis and procedure codes. Corresponding denominator files included beneficiary identifier, date of birth, gender, race, date of death, and information about program eligibility and enrollment. Data from Medicare claims files also allowed us to calculate length of stay and Medicare payments during the year before the index hospitalization. We converted Medicare payment amounts to 2005 US dollars. We included patients in the analysis if we could link a patient's OPTIMIZE-HF registry record to an inpatient Medicare claim. Because OPTIMIZE-HF and Medicare claims data do not include direct patient identifiers, we linked the files by gender, admission date, discharge date, and hospital identifier. In combination, the fields could be used to identify unique hospitalizations.12Hammill B.G. Hernandez A.G. Peterson E.D. Fonarow G.C. Schulman K.A. Curtis L.H. Linking inpatient clinical registry data to Medicare claims data using indirect identifiers.Am Heart J. 2009; 157: 995-1000Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar The earliest hospitalization for each patient served as the index HF hospitalization. Several patients were excluded from the analysis because they had missing data for serum creatinine level at admission (n = 166, 0.6%) or discharge (n = 3,650, 14%) or they had a history of dialysis (n = 485, 1.9%). Eligible patients lived in the United States, were ≥65 years of age, were enrolled in fee-for-service Medicare for ≥12 months before the index hospitalization, and were alive at the time of discharge from the index hospitalization. We obtained patient demographic characteristics, medical history variables, laboratory and examination variables, discharge medications, and procedural information from the OPTIMIZE-HF registry. We imputed mean values of the overall cohort for patients who were missing values for systolic blood pressure (n = 18, 0.1%), serum sodium level (n = 54, 0.3%), or hemoglobin level (n = 322, 1.6%). We imputed the "no" value for missing values on several dichotomous variables, including left ventricular systolic dysfunction (n = 3,015, 15.0%), smoker within the previous year (n = 634, 3.2%), rales (n = 307, 1.5%), lower extremity edema (n = 375, 1.9%), and β-blocker prescription at discharge (n = 178, 0.9%). We defined 3 categories of RF by serum creatinine level at admission. Normal RF at admission was defined as a serum creatinine level or =65 years of age with heart failure.Am J Cardiol. 2000; 85: 1110-1113Abstract Full Text Full Text PDF PubMed Scopus (339) Google Scholar, 14Damman K. Navis G. Voors A.A. Asselbergs F.W. Smilde T.D. Cleland J.G. van Veldhuisen D.J. Hillege H.L. Worsening renal function and prognosis in heart failure: systematic review and meta-analysis.J Card Fail. 2007; 13: 599-608Abstract Full Text Full Text PDF PubMed Scopus (479) Google Scholar Patients were followed for up to 1 year after discharge from the index hospitalization. We calculated all-cause mortality and hospital readmission within 1 year. Mortality information was obtained from Centers for Medicare and Medicaid Services denominator files. We also calculated time to first hospital readmission within 1 year after the index discharge date as the number of days between the index discharge date and subsequent readmission date. We did not count transfers to or from another hospital and admissions for rehabilitation (Diagnosis Related Group 462 or an admitting diagnosis code of V57.xx) as readmissions. In addition, we calculated total inpatient costs to Medicare by summing payment amounts and per diem adjustments from all inpatient claims (including transfer and rehabilitation claims) within 1 year of the index discharge date. We present categorical variables as frequencies and continuous variables as means ± SDs or medians with interquartile ranges. In assessing differences in baseline characteristics between patients with worsening RF and those without, we used chi-square tests for categorical variables and Kruskal-Wallis tests for continuous variables. We present 1-year outcomes stratified by whether patients had worsening RF. We also calculated 1-year inpatient costs by renal impairment group and compared groups using Kruskal-Wallis tests. We calculated unadjusted 1-year mortality rates using Kaplan-Meier methods and tested for differences between patients with worsening RF and those without using log-rank tests. We calculated unadjusted 1-year hospital readmission rates using the cumulative incidence function to account for the competing risk of death.15Prentice R.L. Kalbfleisch J.D. Peterson Jr, A.V. Flournoy N. Farewell V.T. Breslow N.E. The analysis of failure times in the presence of competing risks.Biometrics. 1978; 34: 541-554Crossref PubMed Scopus (1110) Google Scholar We assessed differences in readmission between patients with worsening RF and those without using Gray tests. Cox proportional hazards models with robust SEs accounted for site clustering when we examined unadjusted and adjusted relations between worsening RF and mortality.16Lin D.Y. Wei L.J. The robust inference for the Cox proportional hazards model.J Am Stat Assoc. 1989; 84: 1074-1078Crossref Scopus (1975) Google Scholar In multivariable analysis, we modeled 1-year mortality as a function of worsening RF, age, gender, race, medical history, admission laboratory and examination variables, medications at discharge, procedures during the index hospitalization, inpatient costs in the year before the index hospitalization, intensive care unit length of stay, and a variable indicating whether the length of stay for the index hospitalization was >7 days.17Krumholz H.M. Parent E.M. Tu N. Vaccarino V. Wang Y. Radford M.J. Hennen J. Readmission after hospitalization for congestive heart failure among Medicare beneficiaries.Arch Intern Med. 1997; 157: 99-104Crossref PubMed Google Scholar We also used Cox proportional hazards models to examine relations between worsening RF and 1-year readmission. We used generalized linear models with a log-link and Poisson distribution to examine the unadjusted relation between worsening RF and 1-year inpatient costs. Cost ratios, when using exponentiation, estimate the proportional increase in costs attributable to the variable. Generalized estimating equations accounted for correlations resulting from clustering of similar patients within hospitals. We also used this approach to examine adjusted relations and included the same baseline variables from the mortality and readmission models. We used SAS 9.2 (SAS Institute, Cary, North Carolina) for all analyses. The institutional review board of the Duke University Health System (Durham, North Carolina) approved the study. After we linked OPTIMIZE-HF hospitalizations to Medicare inpatient claims and applied the exclusion criteria, the study population included 20,063 patients hospitalized with HF, of whom 3,581 (17.8%) had worsening RF at discharge. Median age was 80 years, and 56.2% were women. Cause of HF was ischemic in 9,671 patients (48.2%), and 7,246 patients (36.1%) had systolic dysfunction. Table 1 lists baseline demographic characteristics of patients with worsening RF and those without. Patients with worsening RF were more likely to be black. These patients also had higher rates of hypertension, hyperlipidemia, peripheral vascular disease, anemia, history of cerebrovascular accident or transient ischemic attack, and chronic renal dysfunction. Mean length of stay was longer for patients with worsening RF. Patients with worsening RF were less likely to have systolic dysfunction and to be discharged with angiotensin-converting enzyme (ACE) inhibitors or angiotensin II receptor blockers (ARBs), digoxin, and diuretics. In contrast, patients with worsening RF were more likely to be discharged with antiplatelet medications and lipid-lowering agents.Table 1Baseline characteristics of study populationCharacteristicChange in Serum Creatinine Level⁎Patients with a change in serum creatinine ≥0.3 mg/dl during the index hospitalization were considered to have worsening renal function.p Value<0.3 mg/dl≥0.3 mg/dl(n = 16,482)(n = 3,581)Age (years)79.6 ± 7.879.8 ± 7.70.42 65–691,918 (11.6%)384 (10.7%)0.12 70–742,662 (16.2%)592 (16.5%)0.58 75–793,553 (21.6%)770 (21.5%)0.94 ≥808,349 (50.7%)1,835 (51.2%)0.52Men7,249 (44.0%)1,544 (43.1%)0.34Black1,660 (10.1%)405 (11.3%)0.03Nonblack14,822 (89.9%)3,176 (88.7%)0.03Anemia3,038 (18.4%)738 (20.6%)0.003Atrial arrhythmia6,194 (37.6%)1,169 (32.6%)<0.001Chronic obstructive pulmonary disease4,717 (28.6%)1,060 (29.6%)0.24Chronic renal insufficiency2,849 (17.3%)967 (27.0%)<0.001Depression1,779 (10.8%)378 (10.6%)0.68Diabetes mellitus6,340 (38.5%)1,471 (41.1%)0.004Heart failure cause Nonischemic heart failure8,561 (51.9%)1,831 (51.1%)0.38 Ischemic heart failure with myocardial infarction3,791 (23.0%)780 (21.8%)0.11 Ischemic heart failure without myocardial infarction4,130 (25.1%)970 (27.1%)0.01 Heart failure with left ventricular systolic dysfunction6,052 (36.7%)1,194 (33.3%)<0.001Hyperlipidemia5,372 (32.6%)1,258 (35.1%)0.003Hypertension11,641 (70.6%)2,719 (75.9%)<0.001Peripheral vascular disease2,398 (14.5%)596 (16.6%)0.001Previous cerebrovascular accident or transient ischemic attack2,831 (17.2%)685 (19.1%)0.005Pulmonary reactive airway disease777 (4.7%)184 (5.1%)0.28Smoker in previous year1,529 (9.3%)318 (8.9%)0.46Thyroid abnormality2,859 (17.3%)680 (19.0%)0.02Admission characteristics Hemoglobin level (g/dl)12.0 ± 2.011.8 ± 1.9<0.001 Lower extremity edema10,481 (63.6%)2,317 (64.7%)0.21 Rales10,548 (64.0%)2,391 (66.8%)0.002 Serum creatinine level (mg/dl)1.5 ± 0.91.6 ± 1.0<0.001 <1.59,786 (59.4%)1,995 (55.7%)<0.001 1.5–<2.03,474 (21.1%)799 (22.3%)0.10 ≥2.03,222 (19.5%)787 (22.0%)0.001Serum sodium level (mEq/L)137.7 ± 4.9138.0 ± 4.8<0.001Systolic blood pressure (mm Hg)140.9 ± 30.8149.5 ± 32.5<0.001Discharge medications Angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker10,189 (61.8%)2,076 (58.0%)<0.001 Aldosterone antagonist1,900 (11.5%)392 (10.9%)0.32 Antiplatelet agent8,659 (52.5%)1,955 (54.6%)0.03 β blocker10,360 (62.9%)2,272 (63.4%)0.51 Digoxin4,939 (30.0%)816 (22.8%)<0.001 Diuretic13,712 (83.2%)2,899 (81.0%)0.001 Lipid-lowering agent5,847 (35.5%)1,408 (39.3%)<0.001Characteristics of index hospitalization Coronary angiography1,121 (6.8%)269 (7.5%)0.13 Implantable cardioverter–defibrillator278 (1.7%)45 (1.3%)0.06Intensive care unit length of stay (days)1.2 ± 3.01.3 ± 3.40.22 Length of stay (days)5.6 ± 4.76.3 ± 4.8 $21,0003,439 (20.9%)690 (19.3%)0.03Missing OPTIMIZE-HF clinical measurements Hemoglobin level at admission266 (1.6%)56 (1.6%)0.83 Left ventricular systolic dysfunction2,492 (15.1%)523 (14.6%)0.43 Lower extremity edema at admission317 (1.9%)58 (1.6%)0.22 Prescription of β blocker at discharge143 (0.9%)35 (1.0%)0.53 Rales at admission253 (1.5%)54 (1.5%)0.90 Serum sodium at admission45 (0.3%)9 (0.3%)0.82 Smoker in previous year521 (3.2%)113 (3.2%)0.99 Systolic blood pressure at admission13 (0.1%)5 (0.1%)0.27Values are means ± SDs or numbers of patients (percentages). Patients with a change in serum creatinine ≥0.3 mg/dl during the index hospitalization were considered to have worsening renal function. Open table in a new tab Values are means ± SDs or numbers of patients (percentages). Unadjusted all-cause mortality at 1 year in patients discharged with worsening RF was 35.4% versus 34.2% in patients without (p = 0.07; Figure 1,Table 2). Multivariable predictors of mortality are presented in Table 3. After adjustment for baseline characteristics, including serum creatinine level at admission, worsening RF was a significant independent predictor of 1-year all-cause mortality (hazard ratio 1.12, 95% confidence interval 1.04 to 1.20). Other variables were also significant predictors. After adjustment for other variables in the model, a serum creatinine level ≥2 mg/dl at admission was associated with a 46% higher hazard of mortality compared with the reference value (serum creatinine 7 days increased the hazard of 1-year mortality. Beta blockers and ACE inhibitors/ARBs at discharge lowered the hazard of mortality. Diuretic prescription at discharge also lowered the hazard of mortality, whereas digoxin at discharge was associated with a slightly higher hazard of mortality.Table 2Outcomes at one year by change in serum creatinine levelOutcomeChange in Serum Creatinine Level⁎Patients with a change in serum creatinine ≥0.3 mg/dl during the index hospitalization were considered to have worsening RF.p Value<0.3 mg/dl≥0.3 mg/dlUnadjusted mortality rate5,601 (34.2%)1,261 (35.4%)0.07Cumulative incidence all-cause readmission rate10,625 (64.7%)2,301 (64.5%)0.05Inpatient Medicare costs†Costs adjusted to 2005 US dollars.0.75 Mean ± SD$14,957 ± 24,624$14,829 ± 22,931 Median (interquartile range)$6,589 (0–19,401)$6,941 (0–19,366) Patients with a change in serum creatinine ≥0.3 mg/dl during the index hospitalization were considered to have worsening RF.† Costs adjusted to 2005 US dollars. Open table in a new tab Table 3Predictors of mortality, readmission, and inpatient Medicare costs at one yearVariableMortalityReadmissionInpatient Medicare CostsHR (95% CI)p ValueHR (95% CI)p ValueCost Ratio (95% CI)p ValueAge (years) 65–691.00 (reference)1.00 (reference)1.00 (reference) 70–741.20 (1.07–1.34)0.0011.04 (0.97–1.11)0.310.92 (0.83–1.03)0.14 75–791.37 (1.24–1.51)<0.0011.04 (0.97–1.11)0.250.85 (0.78–0.93)<0.001 ≥801.91 (1.73–2.10)<0.0011.05 (0.99–1.12)0.130.69 (0.63–0.76)<0.001Male gender1.07 (1.01–1.13)0.020.98 (0.94–1.02)0.371.04 (0.98–1.09)0.20Black0.93 (0.84–1.04)0.211.24 (1.17–1.33)<0.0011.27 (1.16–1.38)<0.001Nonblack1.00 (reference)1.00 (reference)1.00 (reference)Anemia1.09 (1.02–1.17)0.021.04 (0.99–1.10)0.140.98 (0.93–1.03)0.42Atrial arrhythmia1.06 (1.01–1.12)0.031.05 (1.01–1.09)0.011.00 (0.96–1.05)0.88Chronic obstructive pulmonary disease1.20 (1.14–1.26)<0.0011.17 (1.12–1.22)<0.0011.09 (1.04–1.14)<0.001Chronic renal insufficiency1.10 (1.02–1.19)0.021.06 (0.99–1.12)0.081.03 (0.96–1.10)0.46Depression1.11 (1.03–1.19)0.0091.03 (0.98–1.09)0.250.99 (0.93–1.04)0.62Diabetes mellitus1.02 (0.96–1.08)0.481.11 (1.07–1.15)<0.0011.09 (1.04–1.14)<0.001Heart failure cause, ischemic1.00 (0.94–1.05)0.921.06 (1.02–1.11)0.0060.98 (0.94–1.03)0.45Hyperlipidemia0.84 (0.79–0.90)<0.0010.92 (0.88–0.96)<0.0011.02 (0.96–1.07)0.57Left ventricular systolic dysfunction1.21 (1.14–1.29)<0.0011.02 (0.98–1.07)0.251.02 (0.97–1.07)0.43Peripheral vascular disease1.15 (1.08–1.22)<0.0011.11 (1.06–1.17)<0.0011.09 (1.02–1.16)0.01Previous cerebrovascular accident or transient ischemic attack1.16 (1.08–1.25)<0.0011.10 (1.05–1.15)<0.0011.05 (0.99–1.10)0.08Smoker in previous year1.09 (0.99–1.19)0.071.07 (1.01–1.13)0.030.99 (0.93–1.06)0.86Pulmonary reactive airway disease1.04 (0.93–1.17)0.451.05 (0.96–1.14)0.271.04 (0.90–1.20)0.57Thyroid abnormality1.04 (0.97–1.10)0.261.05 (1.01–1.10)0.021.02 (0.97–1.07)0.42Admission characteristics Hemoglobin level0.96 (0.94–0.97)<0.0010.97 (0.96–0.98)<0.0010.98 (0.97–0.99)0.001 Lower extremity edema0.95 (0.90–1.00)0.040.98 (0.95–1.02)0.401.00 (0.96–1.05)0.83 Rales1.15 (1.09–1.21)<0.0011.01 (0.97–1.06)0.651.04 (1.00–1.08)0.06Serum creatinine level <1.5 mg/dl1.00 (reference)1.00 (reference)1.00 (reference) 1.5–<2.0 mg/dl1.22 (1.14–1.31)<0.0011.19 (1.13–1.25)<0.0011.08 (1.03–1.14)0.004 ≥2.0 mg/dl1.46 (1.36–1.57)<0.0011.25 (1.17–1.34)<0.0011.12 (1.02–1.22)0.02Serum sodium level0.98 (0.98–0.99)<0.0011.00 (0.99–1.00)0.101.00 (1.00–1.00)0.96Systolic blood pressure, per 10 mm Hg0.92 (0.91–0.93)<0.0010.99 (0.98–0.99)<0.0011.00 (0.99–1.01)0.52Change in serum creatinine ≥0.3 mg/dl1.12 (1.04–1.20)0.0031.02 (0.98–1.07)0.300.98 (0.94–1.03)0.50Medications at discharge Angiotensin-converting enzyme inhibitor or angiotensin receptor blocker0.81 (0.77–0.86)<0.0010.97 (0.94–1.01)0.211.04 (0.99–1.10)0.13 Aldosterone antagonist0.96 (0.88–1.04)0.310.98 (0.92–1.03)0.401.03 (0.96–1.11)0.45 Antiplatelet agent0.99 (0.94–1.04)0.631.01 (0.98–1.05)0.571.00 (0.95–1.04)0.85 β blocker0.80 (0.76–0.84)<0.0010.93 (0.89–0.97)<0.0010.97 (0.93–1.02)0.20 Digoxin1.07 (1.01–1.13)0.021.03 (0.99–1.07)0.201.01 (0.96–1.07)0.67 Diuretic0.92 (0.86–0.99)0.021.04 (0.99–1.10)0.121.01 (0.94–1.08)0.81 Lipid-lowering agent0.73 (0.69–0.77)<0.0010.98 (0.94–1.02)0.371.08 (1.02–1.14)0.006Characteristics of index hospitalization Coronary angiography0.62 (0.55–0.70)<0.0010.97 (0.90–1.06)0.531.17 (1.08–1.27)<0.001 Implantable cardioverter–defibrillator0.70 (0.54–0.90)0.0060.81 (0.69–0.96)0.020.81 (0.67–0.98)0.03 Inpatient costs to Medicare in previous year $01.00 (reference)1.00 (reference)1.00 (reference) ≤$8,0001.23 (1.15–1.31)<0.0011.27 (1.21–1.35)<0.0011.19 (1.12–1.27)<0.001 $8,001–$21,0001.30 (1.22–1.38)<0.0011.40 (1.33–1.47)<0.0011.34 (1.27–1.41) $21,0001.26 (1.18–1.33)<0.0011.58 (1.49–1.67)<0.0011.78 (1.67–1.90) 7 days1.37 (1.29–1.46)<0.0011.10 (1.05–1.15)<0.0010.99 (0.94–1.05)0.83 Mechanical ventilation1.16 (0.96–1.40)0.131.07 (0.93–1.23)0.371.19 (1.03–1.38)0.02CI = confidence interval; HR = hazard ratio. Open table in a new tab CI = confidence interval; HR = hazard ratio. Cumulative incidence of all-cause readmission is presented in Figure 2 and Table 2. Nearly 2/3 of patients discharged with worsening RF were readmitted within 1 year. Multivariable predictors of readmission are listed in Table 3. After adjustment for baseline characteristics and comorbid conditions, black race, diabetes mellitus, peripheral vascular disease, chronic obstructive pulmonary disease, and serum creatinine level ≥2 mg/dl at admission increased the hazard of readmission. Worsening RF was not a significant independent predictor of all-cause readmission after multivariable adjustment (Table 3). Mean and median 1-year inpatient Medicare costs are listed in Table 2. Multivariable predictors of 1-year inpatient costs are presented in Table 3. There was no significant difference in observed mean 1-year costs by worsening RF and after multivariable adjustment for baseline characteristics; worsening RF was not a predictor of 1-year inpatient costs. Diabetes mellitus, peripheral arterial disease, chronic obstructive pulmonary disease, and increased serum creatinine level at admission were associated with increased 1-year costs after multivariable adjustment. Strongest predictors of future costs were inpatient costs during the year before the index hospitalization. Older age was inversely associated with 1-year inpatient costs. This analysis is the first large study of patients hospitalized with acute HF outside clinical trial settings to show a significant association between worsening RF and 1-year all-cause mortality. Although we found several predictors of long-term mortality, worsening RF remained an independent predictor after adjustment for baseline characteristics and comorbid conditions. Weinfeld et al8Weinfeld M.S. Chertow G.M. Stevenson L.W. Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure.Am Heart J. 1999; 138: 285-290Abstract Full Text Full Text PDF PubMed Scopus (231) Google Scholar reported an association between worsening RF during HF hospitalization and long-term mortality; however, their study was small (n = 44) and the population was dissimilar to the patients in our study. A more recent, larger analysis did not confirm the association.7Nohria A. Hasselblad V. Stebbins A. Pauly D.F. Fonarow G.C. Shah M. Yancy C.W. Califf R.M. Stevenson L.W. Hill J.A. Cardiorenal interactions: insights from the ESCAPE trial.J Am Coll Cardiol. 2008; 51: 1268-1274Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar That chronic kidney disease increases the risk of adverse outcomes in chronic and acute HF is well established. In a retrospective analysis of the Studies of Left Ventricular Dysfunction prevention and treatment trials, creatinine clearance 80,000 patients found a significant association between renal insufficiency and mortality.18Smith G.L. Lichtman J.H. Bracken M.B. Shlipak M.G. Phillips C.O. DiCapua P. Krumholz H.M. Renal impairment and outcomes in heart failure: systematic review and meta-analysis.J Am Coll Cardiol. 2006; 47: 1987-1996Abstract Full Text Full Text PDF PubMed Scopus (689) Google Scholar Previous research has established that worsening RF during hospitalization for HF increases the risk of adverse outcomes during the hospitalization, including death, complications, and increased length of stay.5Forman D.E. Butler J. Wang Y. Abraham W.T. O'Connor C.M. Gottlieb S.S. Loh E. Massie B.M. Rich M.W. Stevenson L.W. Young J.B. Krumholz H.M. Incidence, predictors at admission, and impact of worsening renal function among patients hospitalized with heart failure.J Am Coll Cardiol. 2004; 43: 61-67Abstract Full Text Full Text PDF PubMed Scopus (751) Google Scholar, 8Weinfeld M.S. Chertow G.M. Stevenson L.W. Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure.Am Heart J. 1999; 138: 285-290Abstract Full Text Full Text PDF PubMed Scopus (231) Google Scholar Baseline renal insufficiency has been associated with increased mortality in the index hospitalization in the Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness registry,10Fonarow G.C. Lukas M.A. Robertson M. Colucci W.S. Dargie H.J. Effects of carvedilol early after myocardial infarction: analysis of the first 30 days in Carvedilol Post-Infarct Survival Control in Left Ventricular Dysfunction (CAPRICORN).Am Heart J. 2007; 154: 637-644Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar the OPTIMIZE-HF registry,19Abraham W.T. Fonarow G.C. Albert N.M. Stough W.G. Gheorghiade M. Greenberg B.H. O'Connor C.M. Sun J.L. Yancy C.W. Young J.B. OPTIMIZE-HF Investigators and CoordinatorsPredictors of in-hospital mortality in patients hospitalized for heart failure: insights from the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF).J Am Coll Cardiol. 2008; 52: 347-356Abstract Full Text Full Text PDF PubMed Scopus (453) Google Scholar and the Acute Decompensated heart failure registry.20Heywood J.T. Fonarow G.C. Costanzo M.R. Mathur V.S. Wigneswaran J.R. Wynne J. ADHERE Scientific Advisory Committee and InvestigatorsHigh prevalence of renal dysfunction and its impact on outcome in 118,465 patients hospitalized with acute decompensated heart failure: a report from the ADHERE database.J Card Fail. 2007; 13: 422-430Abstract Full Text Full Text PDF PubMed Scopus (617) Google Scholar The mechanism by which worsening RF during a HF hospitalization relates to long-term mortality is not clear. Chronic kidney disease and progressive renal dysfunction in the general population are generally associated with increased risk of death. In the HF population, in particular, the risk is combined with and likely contributes to progressive cardiac dysfunction, the so-called cardiorenal syndrome, resulting in increased maladaptive neurohumoral and hemodynamic perturbations. Indeed, worsening RF may be a marker of upregulation of maladaptive neurohormones resulting in renal afferent arteriolar vasoconstriction.21Hillege H.L. Girbes A.R. de Kam P.J. Boomsma F. de Zeeuw D. Charlesworth A. Hampton J.R. van Veldhuisen D.J. Renal function, neurohormonal activation, and survival in patients with chronic heart failure.Circulation. 2000; 102: 203-210Crossref PubMed Scopus (858) Google Scholar In addition, worsening RF during hospitalization may limit diuresis and result in discontinuation of evidence-based therapies, particularly ACE inhibitors, ARBs, and aldosterone antagonists. Indeed, our analysis demonstrated that patients with worsening RF were less likely to be discharged on ACE inhibitors or ARBs and diuretics. Worsening RF may also reflect poor hemodynamics, such as a decreased cardiac index and increased systemic vascular resistance, indicating overall worse tissue perfusion as shown by the significant decrease in RF in patients receiving left ventricular assist device support.22Butler J. Geisberg C. Howser R. Portner P.M. Rogers J.G. Deng M.C. Pierson III, R.N. Relationship between renal function and left ventricular assist device use.Ann Thorac Surg. 2006; 81: 1745-1751Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar Worsening RF as reflected by increased serum creatinine level indicates a decreased glomerular filtration rate that may be a result of prerenal azotemia or acute kidney injury. In patients with acute kidney injury, this may be irreversible, resulting in chronic renal dysfunction. Thus, it is unclear whether it is worsening RF per se that results in worse outcomes rather than a new worse baseline renal insufficiency. In contrast to long-term mortality, we did not find an independent association between worsening RF and readmission or future inpatient costs at 1 year. This finding may be attributable to the fact that increased mortality competes with readmission and costs, because patients who die do not accrue additional hospital admissions or costs. Patients with worsening RF and those without had very high readmission rates and substantial inpatient costs. This analysis has several limitations. First, we defined worsening RF as a change in serum creatinine level ≥0.3 mg/dl from admission to discharge. We did not use daily serum creatinine level during the index hospitalization to define worsening RF; thus, we would have missed patients who sustained worsening RF during the hospitalization but recovered by the time of discharge. Similarly, we do not know the trajectory of patients' RF at discharge, whether the serum creatinine level increased or decreased. Second, a 0.3-mg/dl change in serum creatinine reflects a greater decrease in glomerular filtration rate in patients with lower serum creatinine at baseline than in patients with worse baseline renal insufficiency. For example, a 1.0- to 1.3-mg/dl change in serum creatinine indicates a decrease in glomerular filtration rate of approximately 30 ml/min. A 4.0- to 4.3-mg/dl change in serum creatinine indicates a 2-ml/min decrease in glomerular filtration rate. Other measurements of RF may be more sensitive and allow a more homogenous comparison. Estimation of glomerular filtration rate by Cockroft-Gault equation, Modification of Diet in Renal Disease equation, or percent change in serum creatinine level would be candidate methods.23Smilde T.D. van Veldhuisen D.J. Navis G. Voors A.A. Hillege H.L. Drawbacks and prognostic value of formulas estimating renal function in patients with chronic heart failure and systolic dysfunction.Circulation. 2006; 114: 1572-1580Crossref PubMed Scopus (259) Google Scholar Use of cystatin C, a novel marker of RF, has also been proposed as a more sensitive method to monitor changes in RF with improved prognostic utility.24Shlipak M.G. Sarnak M.J. Katz R. Fried L.F. Seliger S.L. Newman A.B. Siscovick D.S. Stehman-Breen C. Cystatin C and the risk of death and cardiovascular events among elderly persons.N Engl J Med. 2005; 352: 2049-2060Crossref PubMed Scopus (1047) Google Scholar, 25Lassus J. Harjola V.P. Sund R. Siirilä-Waris K. Melin J. Peuhkurinen K. Pulkki K. Nieminen M.S. FINN-AKVA Study GroupPrognostic value of cystatin C in acute heart failure in relation to other markers of renal function and NT-proBNP.Eur Heart J. 2007; 28: 1841-1847Crossref PubMed Scopus (179) Google Scholar Further, there is accumulating evidence that biomarkers such as natriuretic peptides and cardiac troponins are strong predictors of outcomes in HF. Inclusion of these biomarkers into the model may have improved discrimination. However, these biomarkers were not routinely collected at all centers participating in OPTIMIZE-HF and could not be included in the model. Also, worsening RF can be multifactorial, resulting from decreased renal perfusion (prerenal azotemia), intrinsic renal injury (acute kidney injury), or, less likely in our population, postrenal obstruction. Our study does not discriminate among these different causes. It is biologically plausible that patients with worsening RF but no intrinsic injury have better outcomes than patients with acute kidney injury. Perhaps patients with normal RF who develop worsening RF during acute decompensated HF but return to baseline normal function do better than patients with kidney injury who subsequently have chronic renal dysfunction. Stratifying patients with worsening RF by the presence or absence of biomarkers of acute kidney injury, such as neutrophil gelatinase-associated lipocalin or kidney injury molecule-1, may be useful in this regard.26Bennett M. Dent C.L. Ma Q. Dastrala S. Grenier F. Workman R. Syed H. Ali S. Barasch J. Devarajan P. Urine NGAL predicts severity of acute kidney injury after cardiac surgery: a prospective study.Clin J Am Soc Nephrol. 2008; 3: 665-673Crossref PubMed Scopus (626) Google Scholar, 27Bonventre J.V. Kidney injury molecule-1 (KIM-1): a specific and sensitive biomarker of kidney injury.Scand J Clin Lab Invest Suppl. 2008; 241: 78-83Crossref PubMed Scopus (130) Google Scholar, 28Damman K. van Veldhuisen D.J. Navis G. Voors A.A. Hillege H.L. Urinary neutrophil gelatinase associated lipocalin (NGAL), a marker of tubular damage, is increased in patients with chronic heart failure.Eur J Heart Fail. 2008; 10: 997-1000Crossref PubMed Scopus (171) Google Scholar, 29Mishra J. Dent C. Tarabishi R. Mitsnefes M.M. Ma Q. Kelly C. Ruff S.M. Zahedi K. Shao M. Bean J. Mori K. Barasch J. Devarajan P. Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery.Lancet. 2005; 365: 1231-1238Abstract Full Text Full Text PDF PubMed Scopus (1967) Google Scholar We thank Damon M. Seils, MA, Duke University, for assistance with preparation of this report. Mr. Seils did not receive compensation for his assistance apart from his employment at the institution where the study was conducted.

Referência(s)