Methods

This was a single-center cohort study of all consecutive adult patients (age >18 years) prospectively enrolled in Cleveland Clinic’s Heart Care at Home Transitional Care Program after index-hospitalization for a primary diagnosis of HF between January 2010 and April 2012 and had a comprehensive transthoracic echo at admission. All echo parameters from the index hospitalization were retrospectively collected. Heart Care at Home is a structured hospital-to-home transitional care program that helps older adults with HF for up to 40 days after discharge. Components of the program include (1) automated patient identification, (2) predischarge transition coaching, (3) at-home visits by nurses, dieticians, therapists, and social workers, (4) vital signs telemonitoring, (5) routine telephone calls, and (6) and prospective hospital readmission outcomes tracking. Inclusion of patients from this program ensured that all patients were provided comparable postacute care. Exclusion criteria were (1) concomitant acute coronary syndrome during admission defined as chest pain, ischemic electrocardiogram changes, and troponin elevation, (2) terminal noncardiac illness that could influence short-term prognosis, (3) potentially reversible causes of left ventricular (LV) dysfunction (e.g., takotsubo cardiomyopathy), (4) cardiac surgery, biventricular pacemaker, LV assist device or heart transplant during admission, and (5) echo images unsuitable for analysis. The study protocol was approved by the Institutional Review Board at the Cleveland Clinic. The primary hospital diagnosis of HF was confirmed by chart review using the Framingham Heart Failure Criteria and information including hospital course and results of additional cardiac tests. Patients were cared for by full-time Cleveland Clinic Physicians with support of the HF team as necessary. All patients received standard HF therapy during admission per routine Cleveland Clinic protocols.

Clinical data including demographics, cause of cardiomyopathy, co-morbidities, vital signs, pathology results, weight and discharge medications were entered prospectively into templates in the electronic medical record extracted using a relational database management system (Clarity, Epic Systems, Verona, Wisconsin). The records of 3.5% of the patients were manually reviewed to confirm the accuracy of the automated data extraction. Charlson Comorbidity Index scores were calculated for each patient using an SAS macro based on the International Classification of Diseases, Ninth Revision , codes. The Yale Center for Outcome Research and Evaluation (CORE) 30-day HF Readmission Score was calculated for each patient by entering patient data into a validated online risk calculator. This score uses a combination of 20 clinical, hemodynamic, and pathology parameters along with LVEF to provide a validated risk score. Quartiles of HF readmission scores were calculated, and patients with scores above the 75th percentile were classified as being at high risk for readmission.

All echo studies were performed by clinical sonographers as clinically indicated using commercially available systems and transducers. Standard techniques were used to obtain 2-dimensional and Doppler measurements in accordance with the American Society of Echocardiography (ASE) guidelines. LV end-diastolic and systolic volumes and ejection fraction were calculated using the biplane Simpson’s method. Left atrial volume was measured using the biplane method of disks (modified Simpson’s rule), whereas right atrial volume was measured from a 4-chamber view using the area length method. All ventricular and atrial volumes were indexed to body surface area.

Pulsed-wave Doppler derived mitral inflow early (E) and late (A) velocities, pulmonary vein flows, and medial and lateral tissue Doppler-derived early (e′) velocities were obtained from 3 to 5 heartbeats. The mean E/e′ was used in the analysis. LV filling pressure was determined for each patient in accordance with ASE guidelines and cut-off values. In patients with preserved and reduced LVEF in the intermediate filling pressure category, ≥2 of the remaining criteria were used to differentiate normal versus elevated filling pressures. Patients were grouped into normal or indeterminate and elevated filling pressures. The indeterminate group consisted of patients with prosthetic valves and rings, severe mitral annular calcification, severe arrhythmia precluding Doppler analysis, and those in whom diastolic parameters were not obtained. In patients with atrial fibrillation, deceleration time <150 milliseconds or 2 of isovolemic relaxation time ≤65 milliseconds, septal E/e′ ≥11, and deceleration time of pulmonary venous diastolic velocity ≤220 milliseconds were used to define elevated filling pressures. In patients with more than moderate mitral regurgitation, Ar–A ≥30 milliseconds and isovolemic relaxation time <60 milliseconds in those with preserved ejection fraction (EF) and an average E/e′ >15 in patients with reduced EF was used to define elevated filling pressures. Patients in whom an average E/e′ value was available were divided into tertiles. Patients without an average E/e′ available were categorized into the lower-risk group for analysis.

Valvular disease severity was obtained from the echo report and confirmed by 2 reviewers. Right ventricular systolic pressure was estimated using the tricuspid regurgitation peak velocity. Inferior vena cava (IVC) diameter and respiratory collapse and tricuspid annular plane systolic excursion were measured as routinely performed. Right atrial pressures were estimated using a combination of Inferior vena cava (IVC) diameter and respiratory collapse per ASE right heart guidelines. Patients were grouped into normal, intermediate, or indeterminate right atrial pressures and elevated pressures. The indeterminate category consisted of those without an IVC image or inability to determine respiratory variation. All echo analyses were performed blinded to all other clinical data.

All patients were followed prospectively for up to 40 days after discharge, and readmission outcomes were collected. The primary outcome of interest was HF-specific hospital readmission to any hospital within 30 days of discharge, as assessed by prospective telephone calls by Heart Care at Home telehealth nurses. No patients were lost to follow-up.

Data are presented as mean ± SD for continuous variables and proportions for categorical variables. Event-free survival between groups was examined using the Kaplan-Meier analysis and compared using the log-rank test. Cox proportional hazards regression modeling was used to examine the relation of clinical variables and echo parameters with 30-day readmission. Significant univariable predictors and clinically relevant parameters were used in the multivariable model. Additional exploratory models were created using various left and right heart parameters and other clinically important variables ( Appendix, Tables A through D ). For incremental benefit of echo parameters, nested Cox models were used with calculation of overall difference in −2 log likelihood chi-square. Net reclassification improvement was used to determine whether an E/e′ threshold would improve prediction of 30-day readmission over the Yale-CORE readmission score. Patients with indeterminate right atrial pressures and LV filling pressures, as well as patients without an E/e′ value, were included in the lower-risk category for analysis. A 2-tailed p value <0.05 was considered statistically significant. Statistical analysis was performed using SPSS version 19.0 (SPSS, Chicago, Illinois).

Results

Among the 309 patients who met inclusion and exclusion criteria, 26 had inadequate quality echos resulting in 283 patients in our final analysis. Among the included patients 30-day HF-specific readmission occurred in 46 (16.3%). The clinical and echo parameters of the included patients are summarized in Table 1 . Mean (SD) time between admission and echo was 1.9 (1.5) days. The cause of cardiomyopathy was ischemic in 154 patients (54%). The LVEF (mean ± SD) of all patients was 35% ± 17% and 75 patients (26.5%) had an EF ≥50%. At admission, 80% had New York Heart Association class III or IV symptoms. Evidence-based medical therapy at discharge with use up to 30 days after discharge is summarized in Table 2 . The use of angiotensin-converting enzyme inhibitors (ACE) or angiotensin II receptor blockers (ARB) was seen in 64% of the patients ( Table 2 ); however, in an additional 28%, reason for not using ACE or ARB was available and included HF with preserved EF, acute kidney injury, hyperkalemia, and chronic renal failure resulting in the use of nitrates and hydralazine.

Univariable associations of readmission included Yale-CORE HF readmission score, diastolic blood pressure at admission, and echo variables (elevated right atrial and LV filling pressure and the highest tertile of E/e′ ratio [>21.5]; Table 3 ). In multivariable analysis ( Table 4 ), right atrial pressures (hazard ratio [HR] 3.70 [1.82 to 7.52], p <0.001) and elevated LV filling pressure (HR 7.46 [2.31 to 24.14], p = 0.001) remained independently associated with readmission along with weight change during admission (HR 0.93 [0.87 to 0.99], p = 0.02). When E/e′ >21.5 was substituted for elevated LV filling pressure in the model ( Table 5 ), this parameter remained significant (HR 5.36 [2.89 to 9.95], p <0.001) with improvement in the model chi-square. We also tested 6 additional models with left and right heart parameters ( Appendix Tables A and B ), using morphologic and functional variables that have prognostic significance in patients with HF. Elevated right atrial pressure and LV filling pressures remained independently associated with the outcome in all tested models. None of the additional parameters tested were significant in the multivariable model, and the model chi-square values remained lower than that in the final model in Table 5 . We tested 2 other clinical parameters (hypertension at admission and ACE or ARB at discharge) in multivariable models ( Appendix Table C ) and did not find either to be an independent predictor of the outcome. Finally, we also tested a model that used a combined parameter of E/e′ >21.5 and elevated right atrial pressures ( Appendix Table D ) and demonstrated that this parameter remained independently associated with readmission (HR 9.63 [5.09 to 18.21]), with a model chi-square value of 82.8. To find the best cut-off value of E/e′ to predict outcome, we created several models with different cut-off values for E/e′ and the variables from our multivariable model ( Table 4 ). The E/e′ of 23.0 (mean of medial and lateral annulus) had the best performance in predicting 30-day HF specific readmission ( Figure 1 ).

Table 3

Univariable analysis of 30-day heart failure-specific hospital readmission

Variables	HR	(95% CI)	p Value
Age	0.99	0.97–1.01	0.23
Male gender	0.90	0.50–1.60	0.72
Non–Caucasian	0.87	0.49–1.57	0.65
Hypertension	1.77	0.97–3.25	0.07
Diabetes mellitus	1.26	0.78–2.26	0.43
Hypercholesterolemia	0.94	0.52–1.69	0.84
History of coronary artery disease	1.02	0.56–1.84	0.96
History of congestive heart failure	1.37	0.73–2.57	0.32
Length of admission (d)	1.03	0.99–1.07	0.21
Weight change during admission	0.95	0.89–1.01	0.08
Admission log BNP (pg/ml) (n = 213 only) ∗	1.70	0.79–3.68	0.18
Charlson Comorbidity Index	1.04	0.93–1.16	0.53
Yale-CORE heart failure readmission score	1.07	1.01–1.13	0.02
β blocker at discharge	0.89	0.32–2.47	0.82
ACE or ARB at discharge	0.58	0.33–1.04	0.07
Aldosterone antagonist at discharge	0.50	0.21–1.19	0.12
Diuretic at discharge	0.70	0.10–5.08	0.73
Preserved EF ≥50%	0.86	0.44–1.70	0.67
Admission systolic blood pressure (mm Hg)	0.99	0.98–1.00	0.14
Admission diastolic blood pressure (mm Hg)	0.98	0.96–1.00	0.03
Admission heart rate	0.99	0.98–1.01	0.40
Echocardiographic variables
LVEF (%)	1.00	0.98–1.02	0.96
Left ventricular end diastolic volume index (ml/m 2 )	1.00	0.99–1.01	0.88
Left ventricular end systolic volume index (ml/m 2 )	1.00	0.99–1.01	0.92
Left atrial volume index (ml/m 2 )	1.00	0.98–1.01	0.71
Mitral regurgitation ≥ moderate	1.36	0.76–2.42	0.30
Aortic regurgitation ≥ moderate	0.77	0.24–2.47	0.66
Aortic stenosis ≥ moderate	0.52	0.15–2.55	0.51
Elevated LV filling pressures	7.98	2.47–25.71	0.001
Tertiles of E/e′
First tertile (0–15.8)	1
Second tertile (15.9–21.5)	2.25	0.96–5.29	0.06
Third tertile >21.5	5.69	2.73–11.86	<0.001
Right ventricular dysfunction ≥ moderate	0.94	0.52–1.69	0.83
Tricuspid annular plane systolic excursion	1.02	0.96–1.07	0.61
Right ventricular systolic pressure (mm Hg)	1.00	0.98–1.01	0.59
Right atrial volume index (ml/m 2 )	1.00	0.99–1.02	0.53
Elevated right atrial pressures	4.00	1.99–8.08	<0.001
Tricuspid regurgitation ≥ moderate	1.17	0.66–2.09	0.60

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Prognostic Utility of the SYNTAX Score in Patients With Single Versus Multivessel Disease Undergoing Percutaneous Coronary Intervention (from the Acute Catheterization and Urgent Intervention Triage StrategY [ACUITY] Trial)
2. Fatty Acids and Recovery During First Hours of Acute Myocardial Ischemia
3. Outcome of Transcatheter Aortic Valve Implantation in Patients With Low-Gradient Severe Aortic Stenosis and Preserved Left Ventricular Ejection Fraction
4. Recognizing Pregnancy-Associated Cardiovascular Risk Factors

Stay updated, free articles. Join our Telegram channel

Join