Abstract
Aims
Hospitalized heart failure (HF) patients are at high risk for death and readmission. We examined the incremental value of data obtained 1 week after HF hospital discharge in predicting mortality and readmission.
Methods and results
In the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with tolvaptan, 1528 hospitalized patients (ejection fraction ≤40%) with a physical examination, laboratories, and health status [Kansas City Cardiomyopathy Questionnaire (KCCQ)] assessments 1 week after discharge were included. The ability to predict 1 year cardiovascular rehospitalization and mortality was assessed with Cox models, c-statistics, and the integrated discrimination improvement (IDI). Not using a beta-blocker, rales, pedal oedema, hyponatraemia, lower creatinine clearance, higher brain natriuretic peptide, and worse health status were independent risk factors for rehospitalization and death. The c-statistic for the base model (history and medications) was 0.657. The model improved with physical examination, laboratory, and KCCQ results, with IDI increases of 4.9, 7.0, and 3.2%, respectively (P < 0.001 each). The combination of all three offered the greatest incremental gain (c-statistic 0.749; IDI increase 10.8%).
Conclusion
Physical examination, laboratories, and KCCQ assessed 1 week after discharge offer important prognostic information, suggesting that all are critical components of outpatient evaluation after HF hospitalization.
Keywords: Heart failure, Prognosis, Health status, Hospitalization, Mortality, BNP
Introduction
Following hospitalization for acute heart failure (HF) syndromes, patients are at a significantly increased risk for adverse outcomes, with mortality rates as high as 10% and rehospitalization rates as high as 30% 3 months after discharge.1,2 Further, patients hospitalized with HF have the highest 30-day readmission rates of any admission diagnosis among Medicare beneficiaries.3 The ability to identify high-risk patients following hospitalization is particularly important given the opportunity to aggressively treat those patients at greatest risk to improve their outcomes. For example, medication titration, more intense follow-up, and the use of HF disease management programmes have been shown to reduce hospitalizations and mortality.4,5 Accordingly, increased attention has been focused on transitions of care for patients with HF.6
While serial monitoring of chronic HF patients is a class I recommendation,7 and guidelines suggest patients are followed-up 7–10 days after hospital discharge for acute HF syndromes,8 the specific components of early follow-up evaluation that are most predictive of adverse outcomes are unknown. As patients are likely to see a different set of providers during a hospitalization, and direct communication between hospital physicians and outpatient providers at the time of hospital discharge occurs <20% of the time,9 information from hospitalization may not be available at the time of outpatient evaluation. The evaluation after hospital discharge may include a clinic visit with history and physical examination, a formal health status assessment,10,11 and/or laboratory studies.12 Although each of these have been shown to be prognostic in HF, they have not been studied for their capacity to stratify patients' risk in the period shortly after discharge, nor have they been compared to determine if they incrementally, or redundantly, contribute to risk stratification. By clarifying which domains of follow-up evaluation are most important, more efficient strategies for following discharged HF patients could be designed. For example, if health status assessments provided the most important information, merely calling patients after discharge might enable accurate risk stratification and guide future treatment decisions. If biomarkers conferred the most discrimination, a laboratory visit with review of results by phone might suffice. If physical examination is needed to discriminate prognosis, than a clinic visit would be required.
To address these gaps in knowledge, we evaluated which components of the 1-week follow-up visit, including physical examination, laboratories, and health status, offered the greatest incremental value in predicting cardiovascular rehospitalization and mortality in a cohort of patients enrolled in the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with Tolvaptan (EVEREST) Trial. These analyses were designed to improve risk stratification and lend support for recommendations about appropriate clinical monitoring of patients recovering from acute HF syndromes.
Methods
Study design and patient population
This study was a retrospective analysis of data from the EVEREST trial. The design of the EVEREST trial has been previously described.13 Briefly, EVEREST examined both the short-14 and long-term15 clinical effects of the oral vasopressin V2-receptor antagonist tolvaptan, relative to placebo, in adults with a left ventricular ejection fraction (LVEF) ≤40% hospitalized for worsening HF. These patients were receiving standard HF therapies, such as angiotensin-converting enzyme (ACE)-inhibitors and beta-blockers. The trial protocol was approved by the appropriate Institutional Review Board at each study site, and enrolled patients provided written informed consent to participate. The investigation conforms to the principles outlined in the Declaration of Helsinki. From October 2003 to February 2006, 4133 patients were enrolled from 359 sites in North America, South America, and Europe, and randomized to receive tolvaptan or placebo. The eligibility criteria for the trial have been previously described.15 EVEREST was an event-driven trial with the primary outcomes of all-cause mortality and cardiovascular death or hospitalization for HF.
Patients were enrolled in the trial within 48 h of HF hospitalization. Following hospital discharge, data were obtained on each participant at regular intervals (weeks 1, 4, 8, and every 8 weeks thereafter). Our analysis focused upon data obtained at the follow-up visit 1 week after discharge. The 1-week visit was scheduled 7 days after discharge or the 17th day after randomization for those still in the hospital on day 10. Patients who were hospitalized at the time of the 1 week visit were excluded from analysis. Because the contribution to risk assessment conferred by health status measurement [Kansas City Cardiomyopathy Questionnaire (KCCQ) score] was needed to compare and contrast elements of a potential examination, we restricted our analysis to participants with a KCCQ score recorded 1 week after discharge. Since the KCCQ was added as an amendment to the study protocol mid-way through recruitment in 2004, only patients enrolled after this time were eligible for analysis.
Study variables
Data that would not be expected to change from hospitalization to 1 week, including demographics, comorbidities, and discharge medications, were abstracted from the hospital record. The remaining variables included in analysis were obtained at the time of the 1-week follow-up visit including physical examination findings, laboratories, and KCCQ score. Physical examination findings evaluated in these analyses included heart rate (normal if ≤100 b.p.m. vs. tachycardic if >100 b.p.m.), systolic blood pressure (SBP, normal if >100 mmHg vs. hypotensive if ≤100 mmHg), the presence of rales, and the presence of moderate or marked pedal oedema. Laboratory data included haemoglobin (g/dL), sodium (mmol/L), creatinine (mg/dL), and BNP (pg/mL). Hyponatraemia and hypernatraemia were defined by serum sodium of <135 and >145 mmol/L, respectively. Creatinine was used to estimate the glomerular filtration rate (GFR) using the modification of diet in renal disease equation. Anaemia was defined according to haemoglobin level using age-, sex-, and race-specific indices.16 BNP was measured at a centralized laboratory using the Triage Assay (Biosite, CA) performed on the Beckman Access automated chemiluminescense instrument. BNP was divided based on distribution; approximately half of patients had BNP <500 pg/mL and the remainder were divided evenly between 500–999 and ≥1000 pg/mL. The KCCQ is a valid, 23-item self-administered questionnaire that quantifies health status in patients with HF.17 The KCCQ overall summary score (range 0–100, higher scores reflect better health status) includes information from the physical limitation, symptom, social limitation, and quality of life scales and was used for analysis. Culturally and linguistically validated versions of the KCCQ were used in all countries. Participants from the following countries were represented in these analyses: Argentina, Belgium, Brazil, Canada, Czech Republic, France, Germany, Italy, Netherlands, Norway, Poland, Russia, Spain, Sweden, and USA.
Clinically related groups of patient data were clustered for analysis. Patient history and medications [age, sex, race, smoking status, hypertension, coronary artery disease (CAD), diabetes, stroke, chronic obstructive pulmonary disease, internal cardiac defibrillator (ICD), LVEF, and discharge medications] were grouped as data that would be available at hospital discharge and stable in the post-discharge visit. Variables obtained at 1-week follow-up were grouped into physical examination findings (heart rate, SBP, rales, and pedal oedema), laboratory data (sodium, GFR, anaemia, and BNP), and health status (KCCQ). These groupings represent distinct types of follow-up that could be used in defining the optimal clinical strategies for transitioning patients from the inpatient to the outpatient setting.
Study endpoints
The primary endpoint for this analysis was the composite of all-cause mortality or cardiovascular rehospitalization, which was analysed as a time-to-event outcome. All outcomes were adjudicated by an independent clinical events committee. Two committee members independently reviewed records from each hospitalization to determine the primary reason for hospitalization; if there was disagreement, a third member reviewed the case. Cardiovascular hospitalization included hospitalizations due to HF, acute myocardial infarction, stroke, arrhythmia, or other cardiovascular cause.
Statistical analysis
Subjects were divided based on their 1-week KCCQ score (0–24, 25–49, 50–74, 75 and above).10,18 Baseline characteristics are presented as frequencies or means with standard deviations. The association between KCCQ score and the combined endpoint of death or cardiovascular rehospitalization was examined using the Kaplan–Meier method with censoring at the time of last follow-up (mean 8.9 months) or at 1 year in all cases. Cox proportional hazard regression analyses, stratified by site of care, were performed to determine the multivariable predictors of death or rehospitalization. Potential predictors in the model included age, sex, race (African American, Caucasian, other), smoking status, hypertension, CAD, diabetes, stroke, chronic obstructive pulmonary disease, ICD, LVEF, discharge medications [beta-blocker, ACE-inhibitor/angiotensin II receptor blocker (ARB), diuretic, lipid-lowering agent], heart rate (≤100, >100 b.p.m.), SBP (≤100, >100 mmHg), rales, pedal oedema, serum sodium (<135, 135–145, >145 mmol/L), GFR (<30, 30–59, ≥60 mL/min), anaemia, BNP (<500, 500–999, ≥1000 pg/mL), and KCCQ score (0–24, 25–49, 50–74, ≥75). Model predictors were chosen based on clinical factors hypothesized to predict poor outcome.
The incremental prognostic value of the 1-week visit components (physical examination, laboratory data, and KCCQ), over and above the baseline hospital discharge model, was assessed using the c-statistic, a measure of the area under the receiver operating characteristic curves,19 and the integrated discrimination improvement (IDI). The IDI, described by Pencina et al.,20 measures the average increase in model sensitivity penalized for average decrease in specificity with the addition of new variables. It provides an estimate of the accuracy of classifying patients’ likelihood of experiencing the observed outcome(s).
Data were ≥97% complete for all variables examined. The most common missing covariates were BNP (n = 48, 3% missing), a history of stroke (n = 27, 2% missing), and haemoglobin (n = 21, 2% missing), with all other covariates missing <1% of their data. The number of covariates missing per individual was minimal; 93.5% had no missing covariates, 5.5% had 1 missing, and only 1% had 2 or more missing. Therefore, the missing data were assumed to be missing at random and imputed using a single imputation dataset. The imputation model was run using imputation and variance estimation software (IVEWARE)21 and included variables in the Cox model and variables collected at surrounding time intervals. All analyses were performed using SAS version 9.1.3 (SAS Institute, Cary, NC) and R version 2.6.0 (Vienna, Austria). A P-value of <0.05 was used as the level of significance.
Results
Study participants
Of the 4133 EVEREST participants, 2005 were enrolled in the study during the time period when the KCCQ was included in the protocol. Of these 2005, 178 were hospitalized at the time of the 1-week visit and 80 died prior to the visit and were excluded from analysis. Among the remaining 1747 participants, 1528 (87.5%) had a KCCQ score recorded 1 week after discharge and comprised the final study population. These 1528 patients were comparable with the rest of the EVEREST population (n = 2605) in terms of age and sex, but were less frequently Caucasian (82.7% vs. 87.2%, P < 0.001). The distribution of comorbidities was largely similar between the two groups. The 1528 patients were from Eastern Europe (35%), Western Europe (15%), North America (27%), and South America (23%).
The prognostic significance of health status 1 week after discharge
Because other clinical factors examined have been previously demonstrated to have prognostic value in HF, we first sought to establish the prognostic significance of the KCCQ in this setting. Baseline characteristics of participants by KCCQ score are shown in Table 1. The KCCQ score was below 25 (worst health status) in 142 patients (9.3%), 25–49 in 491 (32.1%) patients, 50–74 in 607 (39.7%) patients, and 75–100 (best health status) in 288 (18.8%) patients.
Table 1.
Baseline patient characteristics by 1-week KCCQ score
| Patient characteristic | Overall (n = 1528) | KCCQ <25 (n = 142) | KCCQ 25–49 (n = 491) | KCCQ 50–74 (n = 607) | KCCQ 75–100 (n = 288) | P-value (trend) |
|---|---|---|---|---|---|---|
| Age, years | 65.4 (11.6) | 65.5 (11.1) | 66.2 (10.9) | 65.3 (11.8) | 64.0 (12.4) | 0.14 |
| Female | 404 (26.4) | 54 (38.0) | 155 (31.6) | 135 (22.2) | 60 (20.8) | <0.001 |
| Race | <0.001 | |||||
| Caucasian | 1263 (82.7) | 117 (82.4) | 424 (86.4) | 512 (84.3) | 210 (72.9) | |
| African American | 135 (8.8) | 14 (9.9) | 44 (9.0) | 41 (6.8) | 36 (12.5) | |
| Other | 130 (8.5) | 11 (7.7) | 23 (4.7) | 54 (8.9) | 42 (14.6) | |
| Current smoker | 203 (13.3) | 11 (7.7) | 73 (14.9) | 78 (12.9) | 41 (14.2) | 0.36 |
| Comorbidities | ||||||
| Hypertension | 1119 (73.2) | 115 (81.0) | 364 (74.1) | 431 (71.0) | 209 (72.6) | 0.06 |
| Hyperlipidaemia | 755 (49.6) | 75 (53.2) | 261 (53.5) | 286 (47.3) | 133 (46.3) | 0.03 |
| CAD | 1057 (69.2) | 106 (74.6) | 374 (76.2) | 420 (69.2) | 157 (54.5) | <0.001 |
| Prior MI | 795 (52.1) | 81 (57.4) | 280 (57.1) | 322 (53.0) | 112 (39.0) | <0.001 |
| Prior CABG | 312 (20.4) | 36 (25.4) | 122 (24.8) | 103 (17.0) | 51 (17.7) | 0.002 |
| Diabetes mellitus | 596 (39.0) | 77 (54.2) | 201 (40.9) | 213 (35.1) | 105 (36.5) | <0.001 |
| Prior stroke | 188 (12.5) | 29 (20.9) | 77 (16.1) | 60 (10.1) | 22 (7.6) | <0.001 |
| Chronic obstructive pulmonary disease | 126 (8.2) | 18 (12.7) | 47 (9.6) | 46 (7.6) | 15 (5.2) | 0.003 |
| ICD | 224 (14.7) | 33 (23.2) | 82 (16.7) | 87 (14.3) | 22 (7.6) | <0.001 |
| LVEF, % | 27.5 (8.1) | 27.6 (8.0) | 27.8 (8.1) | 27.4 (8.1) | 27.3 (7.9) | 0.58 |
| NYHA functional class (1 week) | <0.001 | |||||
| Class I | 89 (5.9) | 2 (1.4) | 12 (2.5) | 27 (4.5) | 48 (16.7) | |
| Class II | 697 (45.8) | 28 (19.9) | 153 (31.4) | 332 (54.9) | 184 (64.1) | |
| Class III | 683 (45.6) | 79 (56.0) | 286 (58.6) | 234 (38.7) | 54 (18.8) | |
| Class IV | 82 (5.4) | 32 (22.7) | 37 (7.6) | 12 (2.0) | 1 (0.3) | |
| Discharge medications | ||||||
| Beta-blocker | 1189 (77.8) | 109 (76.8) | 385 (78.4) | 477 (78.6) | 218 (75.7) | 0.67 |
| ACE-inhibitor/ARB | 1334 (87.3) | 120 (84.5) | 424 (86.4) | 542 (89.3) | 248 (86.1) | 0.43 |
| Diuretic | 1436 (94.0) | 137 (96.5) | 460 (93.7) | 572 (94.2) | 267 (92.7) | 0.26 |
| Lipid-lowering agent | 595 (38.9) | 56 (39.4) | 197 (40.1) | 228 (37.6) | 114 (39.6) | 0.75 |
| Physical examination findings (1 week) | ||||||
| Heart rate, b.p.m. | 75.7 (13.3) | 77.6 (14.1) | 76.3 (13.1) | 74.6 (13.2) | 76.0 (13.5) | 0.12 |
| SBP, mmHg | 117.4 (18.5) | 115.1 (18.9) | 116.6 (18.7) | 118.3 (18.7) | 117.7 (17.6) | 0.10 |
| DBP, mmHg | 71.8 (11.5) | 69.8 (11.1) | 71.8 (11.6) | 72.0 (11.5) | 72.3 (11.4) | 0.04 |
| Rales | 266 (17.5) | 46 (32.9) | 101 (20.7) | 88 (14.5) | 31 (10.8) | <0.001 |
| Pedal oedema | 154 (10.1) | 38 (27.0) | 64 (13.1) | 42 (6.9) | 10 (3.5) | <0.001 |
| Laboratory data (1 week) | ||||||
| Sodium, mmol/L | 141 (138, 143) | 140 (137, 143) | 141 (138, 143) | 141 (138, 144) | 141 (138, 143) | 0.03 |
| GFR, mL/min | 57.5 (43.1, 73.1) | 53.1 (39.0, 67.3) | 57.4 (42.6, 73.7) | 57.7 (43.6, 72.1) | 60.7 (46.3, 77.3) | 0.009 |
| Haemoglobin, g/dL | 13.7 (12.3, 15.0) | 13.1 (11.4, 14.2) | 13.5 (12.2, 14.9) | 13.8 (12.6, 15.1) | 14.1 (12.9, 15.5) | <0.001 |
| BNP, pg/mL | 502.3 (212.3, 1003.7) | 791.3 (356.5, 1472.1) | 559.5 (215.5, 1138.0) | 453.4 (185.0, 920.0) | 462.5 (211.5, 816.7) | <0.001 |
Age, LVEF, SBP, DBP, and heart rate are given as mean (standard deviation). Laboratory data are given as median (interquartile range). All other variables are given as n (%).
DBP, diastolic blood pressure; NYHA, New York Heart Association.
At study end, 501 (41.3%) patients had been rehospitalized and 230 (20.6%) had died. In total, 533 (43.2%) patients either died or were rehospitalized. Death or cardiovascular rehospitalization was more frequent in patients with lower KCCQ scores (P < 0.001, Figure 1). At 12 months follow-up, an estimated 64.6% of patients with KCCQ < 25 experienced rehospitalization or death, compared with 48.0, 40.9, and 30.0% for those with KCCQ 25–49, 50–74, and ≥75, respectively. After adjustment for other prognostic factors, a graded independent association between lower KCCQ score and increasing risk of rehospitalization/mortality remained, with fully adjusted hazard ratios (HR) of 3.27 [95% confidence interval (CI) 2.13–5.03], 2.09 (1.46–2.99), and 1.45 (1.03–2.04) for those with KCCQ scores of <25, 25–49, and 50–74 as compared with those whose KCCQ scores were ≥75 (Figure 2). Additional independent predictors of cardiovascular rehospitalization/mortality included the presence of rales, pedal oedema, poor renal function, hyponatraemia, and higher BNP, while use of a beta-blocker was protective. The independent predictors of cardiovascular rehospitalization and death (separately) were similar to the combined endpoint.
Figure 1.
Kaplan–Meier curves for cardiovascular rehospitalization or death by 1-week KCCQ scores are shown.
Figure 2.
Adjusted risk of cardiovascular rehospitalization or death hazard ratios and 95% CI derived from multivariable Cox proportional hazard regression analysis are shown. Demographics, history, and medication data were available at the time of hospital discharge. Physical examination findings, laboratory data, and KCCQ scores were obtained at the 1-week follow-up visit.
The relative prognostic value of alternative methods for monitoring HF patients after discharge
The c-statistic for the base model, including patient history and medications, was 0.657. Addition of the domains of physical examination findings, laboratory data, and health status (KCCQ) 1 week after discharge each improved the discriminatory performance of the model (Table 2). Continuous predictors were entered into the model as categorical variables for improved interpretation; sensitivity analysis using continuous variables demonstrated comparable results. While physical examination findings, laboratory data, and health status each contributed important prognostic information, the combination of all three provided the greatest incremental prognostic value, as noted by an increase in c-statistic from 0.658 to 0.749, and an IDI increase of 10.8% (P < 0.001, Table 2).
Table 2.
Prognostic value of 1-week physical examination, laboratories, and KCCQ
| 1-week follow-up componentsa | c-statistic | IDI increase (%) | P-value for IDI increaseb |
|---|---|---|---|
| History/medications | 0.657 | — | — |
| History/medications + physical examination | 0.699 | 4.9 | <0.001 |
| History/medications + laboratories | 0.719 | 7.0 | <0.001 |
| History/medications + KCCQ | 0.686 | 3.2 | <0.001 |
| History/medications + physical examination + laboratories + KCCQ | 0.749 | 10.8 | <0.001 |
aHistory/medications includes age, sex, race, smoking status, hypertension, CAD, diabetes, stroke, chronic obstructive pulmonary disease, ICD, LVEF, discharge medications (beta-blocker, ACE-inhibitor/ARB, diuretic, lipid-lowering agent); physical examination includes heart rate, SBP, rales, pedal oedema; Laboratories include GFR, sodium, anaemia, BNP.
bComparison with model with history/medications only.
Given the time and resource-intensive nature of performing a comprehensive 1-week evaluation, we investigated which of the individual variables from the 1-week visit provided the greatest incremental prognostic value. The increase in the IDI for each variable is shown in Table 3. BNP was associated with the largest increase in prognostic value of all variables examined (IDI increase 5.5%), followed by KCCQ (IDI increase 3.2%), and pedal oedema (IDI increase 2.9%). Including only these three variables (BNP, KCCQ, and pedal oedema), the c-statistic for the model was 0.734 (IDI increase 8.8%), representing only a minimal loss in discriminatory power compared with the full model reported above.
Table 3.
Prognostic value of individual elements of the 1-week follow-up examination
| 1-week follow-up componentsa | IDI increase (%) | P-value for IDI increaseb |
|---|---|---|
| BNP | 5.5 | <0.001 |
| KCCQ | 3.2 | <0.001 |
| Pedal oedema | 2.9 | <0.001 |
| Rales | 2.2 | <0.001 |
| Anaemia | 1.5 | <0.001 |
| GFR (mL/min) | 1.0 | <0.001 |
| SBP (mmHg) | 0.6 | 0.005 |
| Serum sodium (mmol/L) | 0.2 | 0.08 |
| Heart rate (b.p.m.) | 0.03 | 0.48 |
aFrom highest to lowest IDI increase.
bComparison to model with history (age, sex, race, comorbidities, LVEF)/discharge medications only.
Discussion
While early follow-up after HF discharge is strongly endorsed by clinical guidelines8,22 and has been of recent clinical interest,6 to the best of our knowledge this is the first study to rigorously assess the value of the follow-up examination and to identify the components of evaluation that are most important in stratifying patients' risk for death and rehospitalization. We found that information obtained at the 1-week follow-up examination contributed substantially to the ability to risk-stratify patients, improving the c-statistic of a model based on history and medication data available at discharge from 0.657 to 0.749, associated with an increase in IDI of 10.8%. While each component of the 1-week follow-up (physical examination, laboratories, health status) provided important information in determining prognosis, their combined value provided the best discrimination. Given the centrality of risk stratification as the foundation for modifying treatment choices, these findings provide important quantitative information to guide follow-up strategies for delivering optimal care.
Physical examination
The physical examination remains a cornerstone of the HF evaluation. While vital signs, including low SBP, are associated with adverse risk in patients hospitalized with HF,23,24 their prognostic role early after hospital discharge was unclear. In addition, while physical examination findings of volume overload, including pedal oedema and pulmonary rales, have an indisputable role in diagnosing HF and evaluating cardiac performance,25 there are limited data regarding their prognostic value in patients with HF following a hospitalization. Although a physical examination can be obtained by a provider and quickly applied to medical decision-making, it requires an in-person visit. These data demonstrate that a targeted physical examination including measurement of heart rate, SBP, and an evaluation for rales and pedal oedema offers important prognostic information in patients recently hospitalized with HF. In particular, the presence of rales and pedal oedema was associated with increased risk of rehospitalization and mortality.
Laboratory data
Laboratory data are often used to augment the history and physical examination in HF patients. The use of biomarkers in HF risk prediction has been of interest,26 and BNP has important prognostic value in patients hospitalized with worsening HF.27 Additional laboratory findings, including anaemia, reduced renal function, and hyponatraemia, have been demonstrated to be associated with poor prognosis,28–30 although their role individually and in combination after hospital discharge is not well established. These data demonstrate that the presence of hyponatraemia, decreased renal function, and elevated BNP 1 week after hospital discharge is associated with an increased risk of rehospitalization and mortality among patients discharged from the hospital for HF exacerbation. Furthermore, the measurement of HF-specific laboratories provided the largest single incremental increase in predictive ability of the follow-up domains evaluated. Among the laboratories assessed, BNP provided the largest incremental prognostic value, and thus represents a powerful predictor of adverse outcomes in HF patients being evaluated 1 week after discharge. Potential disadvantages of laboratory assessments are cost, and, while methods such as rapid BNP assays exist in some settings, in most clinical situations laboratory results are not available at the time of the clinic visit, limiting their immediate use in medical decision-making.
Health status measurement
HF is a chronic disease associated with a high symptom burden, and poor health status is common.10 Indeed, a thorough history and assessment of symptoms are an integral part of any clinical encounter. The KCCQ provides a quantitative assessment of symptoms and has proven to be a reliable measure of health status in HF,17 whereas other measures of quantifying symptoms such as New York Heart Association (NYHA) functional class have proven to be less reproducible across practices.31 While other health status measures such as the Minnesota Living with Heart Failure Questionnaire and SF-36 have been used in HF,32–34 the KCCQ has been demonstrated to have much greater sensitivity to changes in health status in patients with HF.17 KCCQ scores have been demonstrated to be associated with survival and hospitalization in patients with HF,11 and among chronic HF outpatients who have not been recently hospitalized.10 Despite these prior data, formal measures of health status are rarely done as part of routine clinical practice or used in risk-stratifying patients. The present study extends previous reports by demonstrating that KCCQ scores measured 1 week after hospital discharge independently predict survival free of cardiovascular rehospitalization. In fact, patients with KCCQ scores <25 (worst health status) had a more than threefold increased risk of the combined endpoint of rehospitalization and mortality as compared with those in the best health status tier (KCCQ score ≥75). Furthermore, KCCQ scores provided important independent prognostic information, and its incremental value in risk prediction was second only to BNP. Thus, including the KCCQ in a follow-up visit 1 week after hospital discharge may improve risk stratification for individual patients and may be practical as the costs of its administration are low and the results are immediately available at the time of the clinical encounter.
Comprehensive evaluation strategy
While the physical examination, laboratory data, and KCCQ each provided important individual prognostic information, the combination of all provided the greatest accuracy in risk stratification. Therefore, a comprehensive assessment 1 week after hospital discharge, including patient history, review of medications, targeted physical examination, laboratory, and health status assessments, may represent the best strategy for identifying HF patients at highest risk for adverse outcomes. Our analyses examining each elements' incremental contribution to risk stratification suggested that the most important elements were BNP, KCCQ, and pedal oedema and that a model including only these data had only slightly less discriminatory power than the full model. Therefore, a focus on these elements during follow-up may represent an efficient method for risk stratification.
Study limitations
Some potential limitations of the study should be acknowledged. This is a post-hoc analysis of patients enrolled in a clinical trial and it is known that such patients are younger and more frequently male than a community HF population. Second, this study only included patients with LVEF ≤40%, so the prognostic value of these parameters among HF patients with preserved LVEF is unknown. In addition, this study excluded patients with significant hypotension and severe renal dysfunction. Findings from the 1-week follow-up visit may have contributed to the decision to hospitalize a patient with HF. However, this is not likely to explain the association between BNP and KCCQ scores with outcome, given that these results were not available to the clinicians evaluating the patients. In addition, the independent predictors of mortality, which would not be impacted by the provider's knowledge of patient 1-week findings, were nearly identical to those of the combined endpoint. Finally, these findings would benefit from future external validation using other cohorts of patients following HF hospitalization. Despite these potential limitations, this study has the advantage of including a large amount of clinical information that was systematically collected 1 week after hospital discharge. The presence of a HF-specific physical examination, as well as measurement of BNP and health status, is particularly notable, as these elements are frequently lacking from observational HF registries. In addition, the enrollment of patients from 359 sites on several continents, and from both academic and non-academic centres, increases the generalizability of the findings.
Conclusions
Care of patients in the period after hospital discharge for HF is important, as these patients are at high risk for adverse outcomes, including mortality and rehospitalization. The present study identifies the specific components of an early evaluation after hospital discharge that may aid clinicians in most efficiently identifying HF patients at the highest risk for adverse outcomes. By accurately identifying patients at higher risk after hospital discharge, providers may employ a more aggressive treatment strategy, including medication titration, increased frequency of follow-up, and enrollment in HF disease management programmes that have the potential to optimize patients' prognosis. Further work is needed to determine whether application of these data can lead to improved care and outcomes for HF patients.
Funding
The EVEREST trial was sponsored by the Otsuka Maryland Research Institute, Rockville.
Conflict of interest: S.M.D., P.S.C., and K.J.R. have no disclosures. M.G. has served as a consultant for Otsuka, Solvay Pharma, Novartis, Bayer, Sigma Tau, Debiopharm, Medtronic, Merck, Astellas, Cytokinetics, CorThera Inc., Pericor Therapeutics, GlaxoSmithKline, Johnson & Johnson, Abbott, Errekappa Terapeutici, Protein Design Laboratories, AstraZenica, Protein Design Laboratories, sanofi-aventis. L.A.A. had an unrestricted grant from CHF Solutions that was terminated in 2008. P.J.H. has served as a consultant for Otsuka, BioControl Medical, Merck, Cardiokine, and ARCA and was on the speaker's bureau for GSK. A.P.M. received honoraria from Otsuka. K.S. reported receiving research grants from AstraZeneca, Servier, and Amgen; being a consultant for Cytokinetics, Servier, and Novartis; and receiving honoraria from AstraZeneca, Otzuka, Amgen, and Servier. F.Z. reports receiving consultant honoraria from Servier; AstraZeneca, Pfizer, Boehringer Ingelheim, Novartis, Abbott, Relypsa, Resmed, Merck, Daiichi Sankyo, Takeda, Boston Scientific; Medtronic and Otsuka. M.A.K. has served as a consultant or had other financial relationships with Otsuka, Merck, Cardiokine, Biogen, Orqis Medical, Boston Scientific, sanofi, Cytokinetcis, and Novartis. J.A.S. has served as a consultant for United Healthcare and St. Jude Medical; received research grants and contracts from NHLBI, ACCF, Amgen, Lilly, BMS/sanofi, and JnJ; and holds intellectual property rights for the SAQ, KCCQ, and PAQ (Prism Software).
References
- 1.Fonarow GC, Stough WG, Abraham WT, Albert NM, Gheorghiade M, Greenberg BH, O'Connor CM, Sun JL, Yancy CW, Young JB. Characteristics, treatments, and outcomes of patients with preserved systolic function hospitalized for heart failure: a report from the OPTIMIZE-HF Registry. J Am Coll Cardiol. 2007;50:768–777. doi: 10.1016/j.jacc.2007.04.064. [DOI] [PubMed] [Google Scholar]
- 2.Gheorghiade M, Pang PS. Acute heart failure syndromes. J Am Coll Cardiol. 2009;53:557–573. doi: 10.1016/j.jacc.2008.10.041. [DOI] [PubMed] [Google Scholar]
- 3.Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360:1418–1428. doi: 10.1056/NEJMsa0803563. [DOI] [PubMed] [Google Scholar]
- 4.McAlister FA, Stewart S, Ferrua S, McMurray JJ. Multidisciplinary strategies for the management of heart failure patients at high risk for admission: a systematic review of randomized trials. J Am Coll Cardiol. 2004;44:810–819. doi: 10.1016/j.jacc.2004.05.055. [DOI] [PubMed] [Google Scholar]
- 5.Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S, Rubin HR. Comprehensive discharge planning with postdischarge support for older patients with congestive heart failure: a meta-analysis. J Am Med Assoc. 2004;291:1358–1367. doi: 10.1001/jama.291.11.1358. [DOI] [PubMed] [Google Scholar]
- 6.Epstein AM. Revisiting readmissions–changing the incentives for shared accountability. N Engl J Med. 2009;360:1457–1459. doi: 10.1056/NEJMe0901006. [DOI] [PubMed] [Google Scholar]
- 7.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. doi: 10.1161/CIRCULATIONAHA.105.167586. [DOI] [PubMed] [Google Scholar]
- 8.Adams KF, Lindenfeld J, Arnold JM, Massie BM, Baker DW, Mehra MR, Barnard DH, Miller AB. Executive summary: HFSA 2006 comprehensive heart failure practice guideline. J Card Fail. 2006;12:e1–e122. [Google Scholar]
- 9.Kripalani S, LeFevre F, Phillips CO, Williams MV, Basaviah P, Baker DW. Deficits in communication and information transfer between hospital-based and primary care physicians: implications for patient safety and continuity of care. J Am Med Assoc. 2007;297:831–841. doi: 10.1001/jama.297.8.831. [DOI] [PubMed] [Google Scholar]
- 10.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. doi: 10.1016/j.jacc.2005.11.021. [DOI] [PubMed] [Google Scholar]
- 11.Soto GE, Jones P, Weintraub WS, Krumholz HM, Spertus JA. Prognostic value of health status in patients with heart failure after acute myocardial infarction. Circulation. 2004;110:546–551. doi: 10.1161/01.CIR.0000136991.85540.A9. [DOI] [PubMed] [Google Scholar]
- 12.Bonow RO, Bennett S, Casey DE, Jr, Ganiats TG, Hlatky MA, Konstam MA, Lambrew 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. 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. doi: 10.1161/CIRCULATIONAHA.105.170072. [DOI] [PubMed] [Google Scholar]
- 13.Gheorghiade M, Orlandi C, Burnett JC, Demets D, Grinfeld L, Maggioni A, Swedberg K, Udelson JE, Zannad F, Zimmer C, Konstam MA. Rationale and design of the multicenter, randomized, double-blind, placebo-controlled study to evaluate the efficacy of vasopressin antagonism in heart failure: outcome study with tolvaptan (EVEREST) J Card Fail. 2005;11:260–269. doi: 10.1016/j.cardfail.2005.03.009. [DOI] [PubMed] [Google Scholar]
- 14.Gheorghiade M, Konstam MA, Burnett JC, Jr, Grinfeld L, Maggioni AP, Swedberg K, Udelson JE, Zannad F, Cook T, Ouyang J, Zimmer C, Orlandi C. Short-term clinical effects of tolvaptan, an oral vasopressin antagonist, in patients hospitalized for heart failure: the EVEREST clinical status trials. J Am Med Assoc. 2007;297:1332–1343. doi: 10.1001/jama.297.12.1332. [DOI] [PubMed] [Google Scholar]
- 15.Konstam MA, Gheorghiade M, Burnett JC, Jr, Grinfeld L, Maggioni AP, Swedberg K, Udelson JE, Zannad F, Cook T, Ouyang J, Zimmer C, Orlandi C. Effects of oral tolvaptan in patients hospitalized for worsening heart failure: the EVEREST outcome trial. Jama. 2007;297:1319–1331. doi: 10.1001/jama.297.12.1319. [DOI] [PubMed] [Google Scholar]
- 16.Beutler E, Waalen J. The definition of anemia: what is the lower limit of normal of the blood hemoglobin concentration? Blood. 2006;107:1747–1750. doi: 10.1182/blood-2005-07-3046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.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. doi: 10.1016/s0735-1097(00)00531-3. [DOI] [PubMed] [Google Scholar]
- 18.Chan PS, Soto G, Jones PG, Nallamothu BK, Zhang Z, Weintraub WS, Spertus JA. Patient health status and costs in heart failure: insights from the eplerenone post-acute myocardial infarction heart failure efficacy and survival study (EPHESUS) Circulation. 2009;119:398–407. doi: 10.1161/CIRCULATIONAHA.108.820472. [DOI] [PubMed] [Google Scholar]
- 19.Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology. 1983;148:839–843. doi: 10.1148/radiology.148.3.6878708. [DOI] [PubMed] [Google Scholar]
- 20.Pencina MJ, D'Agostino RB, Sr, D'Agostino RB, Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157–172. doi: 10.1002/sim.2929. discussion 207-12. [DOI] [PubMed] [Google Scholar]
- 21.Raghunathan TE, Solenberger PW, van Hoewyk J. Ann Arbor, MI: Survey Research Center, Institute for Social Research; 2002. IVEWARE: imputation and variance estimation software—user guide. [Google Scholar]
- 22.Dickstein K, Cohen-Solal A, Filippatos G, McMurray JJ, Ponikowski P, Poole-Wilson PA, Stromberg A, van Veldhuisen DJ, Atar D, Hoes AW, Keren A, Mebazaa A, Nieminen M, Priori SG, Swedberg K, Vahanian A, Camm J, de Caterina R, Dean V, Funck-Brentano C, Hellemans I, Kristensen SD, McGregor K, Sechtem U, Silber S, Tendera M, Widimsky P, Zamorano JL, Auricchio A, Bax J, Bohm M, Corra U, della Bella P, Elliott PM, Follath F, Gheorghiade M, Hasin Y, Hernborg A, Jaarsma T, Komajda M, Kornowski R, Piepoli M, Prendergast B, Tavazzi L, Vachiery JL, Verheugt FW, Zannad F. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: the task force for the diagnosis and treatment of acute and chronic heart failure 2008 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM) Eur J Heart Fail. 2008;10:933–989. doi: 10.1016/j.ejheart.2008.08.005. [DOI] [PubMed] [Google Scholar]
- 23.Abraham WT, Fonarow GC, Albert NM, Stough WG, Gheorghiade M, Greenberg BH, O'Connor CM, Sun JL, Yancy CW, Young JB. Predictors 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;52:347–356. doi: 10.1016/j.jacc.2008.04.028. [DOI] [PubMed] [Google Scholar]
- 24.Gheorghiade M, Abraham WT, Albert NM, Greenberg BH, O'Connor CM, She L, Stough WG, Yancy CW, Young JB, Fonarow GC. Systolic blood pressure at admission, clinical characteristics, and outcomes in patients hospitalized with acute heart failure. J Am Med Assoc. 2006;296:2217–2226. doi: 10.1001/jama.296.18.2217. [DOI] [PubMed] [Google Scholar]
- 25.Mueller C, Frana B, Rodriguez D, Laule-Kilian K, Perruchoud AP. Emergency diagnosis of congestive heart failure: impact of signs and symptoms. Can J Cardiol. 2005;21:921–924. [PubMed] [Google Scholar]
- 26.Braunwald E. Biomarkers in heart failure. N Engl J Med. 2008;358:2148–2159. doi: 10.1056/NEJMra0800239. [DOI] [PubMed] [Google Scholar]
- 27.Fonarow GC, Peacock WF, Phillips CO, Givertz MM, Lopatin M. Admission B-type natriuretic peptide levels and in-hospital mortality in acute decompensated heart failure. J Am Coll Cardiol. 2007;49:1943–1950. doi: 10.1016/j.jacc.2007.02.037. [DOI] [PubMed] [Google Scholar]
- 28.Dunlay SM, Weston SA, Redfield MM, Killian JM, Roger VL. Anemia and heart failure: a community study. Am J Med. 2008;121:726–732. doi: 10.1016/j.amjmed.2008.03.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hillege HL, Nitsch D, Pfeffer MA, Swedberg K, McMurray JJ, Yusuf S, Granger CB, Michelson EL, Ostergren J, Cornel JH, de Zeeuw D, Pocock S, van Veldhuisen DJ. Renal function as a predictor of outcome in a broad spectrum of patients with heart failure. Circulation. 2006;113:671–678. doi: 10.1161/CIRCULATIONAHA.105.580506. [DOI] [PubMed] [Google Scholar]
- 30.Klein L, O'Connor CM, Leimberger JD, Gattis-Stough W, Pina IL, Felker GM, Adams KF, Jr, Califf RM, Gheorghiade M. Lower serum sodium is associated with increased short-term mortality in hospitalized patients with worsening heart failure: results from the outcomes of a prospective trial of intravenous Milrinone for exacerbations of chronic heart failure (OPTIME-CHF) study. Circulation. 2005;111:2454–2460. doi: 10.1161/01.CIR.0000165065.82609.3D. [DOI] [PubMed] [Google Scholar]
- 31.Raphael C, Briscoe C, Davies J, Ian Whinnett Z, Manisty C, Sutton R, Mayet J, Francis DP. Limitations of the New York Heart Association functional classification system and self-reported walking distances in chronic heart failure. Heart. 2007;93:476–482. doi: 10.1136/hrt.2006.089656. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Colucci WS, Packer M, Bristow MR, Gilbert EM, Cohn JN, Fowler MB, Krueger SK, Hershberger R, Uretsky BF, Bowers JA, Sackner-Bernstein JD, Young ST, Holcslaw TL, Lukas MA. Carvedilol inhibits clinical progression in patients with mild symptoms of heart failure. US Carvedilol Heart Failure Study Group. Circulation. 1996;94:2800–2806. doi: 10.1161/01.cir.94.11.2800. [DOI] [PubMed] [Google Scholar]
- 33.Hobbs FD, Kenkre JE, Roalfe AK, Davis RC, Hare R, Davies MK. Impact of heart failure and left ventricular systolic dysfunction on quality of life: a cross-sectional study comparing common chronic cardiac and medical disorders and a representative adult population. Eur Heart J. 2002;23:1867–1876. doi: 10.1053/euhj.2002.3255. [DOI] [PubMed] [Google Scholar]
- 34.Packer M, Colucci WS, Sackner-Bernstein JD, Liang CS, Goldscher DA, Freeman I, Kukin ML, Kinhal V, Udelson JE, Klapholz M, Gottlieb SS, Pearle D, Cody RJ, Gregory JJ, Kantrowitz NE, LeJemtel TH, Young ST, Lukas MA, Shusterman NH. Double-blind, placebo-controlled study of the effects of carvedilol in patients with moderate to severe heart failure. The PRECISE trial: prospective randomized evaluation of Carvedilol on symptoms and exercise. Circulation. 1996;94:2793–2799. doi: 10.1161/01.cir.94.11.2793. [DOI] [PubMed] [Google Scholar]