Clinical Characteristics, Comorbidities, and Prognosis in Patients with Heart Failure with Unknown Ejection Fraction

2.1. Echocardiography

All images were digitally obtained from 3-5 (5 if atrial fibrillation) consecutive cardiac cycles using a Phillips iE 33 (Cleveland, Ohio) with a 2.5 mHz broadband transducer (X5) at end-expiration. Doppler echocardiography was obtained from multiple ultrasonic windows to provide views of all cardiac chambers and valves. Transmitral pulsed Doppler was obtained from 1-2 x 1-2 mm sample volume placed at the tips of the mitral leaflets. Using a 1-2 x 1-2 mm sample volume, LV outflow tract pulsed Doppler was obtained from the apical 3 or 5 chambers approximately 3-5 mm from the aortic valve annulus. Spectral tissue Doppler was obtained from the septal and lateral mitral annuli using a 3x3 mm pulsed Doppler sample volume. All measurements are the average of 3-5 (5 if atrial fibrillation) consecutive cardiac cycles at end-expiration and were performed according to the recommendation of the American Society of Echocardiography guidelines [12]. We calculated LV end-diastolic and end-systolic volumes (using Simpson’s rule), LV mass, and left atrial volumes and indexed these measures to body surface area. The ejection fraction was calculated in the usual manner. End diastolic basal right ventricular basal dimension and tricuspid annular plane systolic excursion were measured as described by the American Society of Echocardiography guidelines [12]. From the LV outflow tract pulsed Doppler recordings, we determined the peak velocity and velocity-time integral. The LV outflow tract area was calculated as the product of 0.785*LV outflow tract diameter^2, assuming circular geometry. The stroke volume index was calculated as the product of the velocity-time integral and LV outflow tract area and indexed to body surface area. From transmitral Doppler, E, A, E/A were obtained based on the American Society of Echocardiography guidelines [13]. Spectral Doppler of the septal and lateral mitral annulus e’ was analyzed and averaged. The E/e’ was calculated [13]. Tricuspid regurgitation velocity was obtained from a continuous wave Doppler transducer using multiple windows. Using the modified Bernoulli equation, right ventricular systolic pressure was estimated with the addition of an estimate of right atrial pressure based on respiratory variations of the inferior vena cava dimension using the American Society of Echocardiography recommendations [13].

2.2. Statistics

Data were expressed as mean+standard deviation for continuous normally distributed data as determined by the Kolmogorov-Smirnov test. For data that were not normally distributed, median and interquartile ranges were computed. Categorical data was expressed as a percentage of the group having that attribute. Differences between the groups were determined using 1-way analysis of variance or 1-way analysis of variance on ranks. If the F value was <0.05, the differences between individual groups were determined by Dunn’s test. Differences in percentages among groups were determined using chi-square. If p<0.05, then a multi-comparison technique was utilized to determine where the significant differences existed. (COMPROP-SAS, Cary, NC). P values <0.05 were considered significant. Multiple logistic regression was performed to determine the independent correlates of ACM and readmission using univariate statistics with p<0.10 for each phenotype. Cox proportional analysis was utilized to assess the relation of HF phenotype to ACM and readmission for HF taking into account the variables derived above. Adjusted Kaplan-Meier curves were subsequently constructed to determine HR’s for ACM and readmission for HF. Landmark analysis was performed on all patients already admitted once to determine HR’s for ACM and re-admission a 2^nd time for HF. Log-rank testing was performed to determine significance. Statistics were calculated with Sigma Stat 4 (SYSTAT, San Jose, CA) and SAS (Cary, NC).

4.1. Previous Literature

HF is a chronic problem associated with high morbidity and mortality with higher prevalence and incidence in the elderly population due to increased survival from myocardial infarction and increased prevalence of co-morbidities with aging. Diagnosis can be easily missed due to non-specific symptoms that simulate respiratory conditions and lack of access to or use of echocardiography. Diagnosis of HF can be further substantiated and phenotyped with echocardiographic assessment characterized by either HFrEF or HFpEF. According to the European Society of Cardiology, HF is clinically defined as a syndrome in patients with typical symptoms (shortness of breath, fatigue and ankle swelling) and signs (elevated jugular venous pressure, pulmonary crackles, and displaced heartbeat) resulting from an abnormality in cardiac structure and function [14]. The diagnosis of HFrEF requires 3 criteria: symptoms, signs as stated above, and a reduced LV ejection fraction. The diagnosis of HFpEF needs 4 criteria: typical symptoms, signs, LV ejection fraction >50%, and structural heart disease including LV hypertrophy, left atrial enlargement and/or diastolic dysfunction. However, there are patients who are not “phenotyped” in administrative or even patient-level databases because an assessment of LV ejection fraction was not performed. This study addresses the characteristics and some of their outcomes from a single rural database in a community medical school without a specific HF service.

Explanations as to why patients without an ejection fraction determination should demonstrate increased ACM are at best speculative. Differences in clinical characteristics include a lower prevalence of coronary disease, hyperlipidemia, mitral and aortic regurgitation as compared to HFrEF. These differences, except for coronary disease, are small and unrelated to ACM by multiple logistic regression. The lower prevalence would not have an obvious mechanistic reason to increase ACM. Additionally, HF medication (angiotensin-converting enzyme inhibitors/ angiotensin receptor blockers, beta-blockers, and aldosterone antagonists) that is used to promote survival, was limited in this group and similar to the other groups. Of note, readmission for HF was lower and consequently, attention to HF care may have been reduced in this group. We were unable to characterize the cause of ACM, whether it is non-cardiac or cardiac and the prevalence of sudden death.

Previous studies have reported an increased prevalence of unrecognized HF in elder populations referred to a cardiologist or to the hospital for echocardiography to confirm the diagnosis. Riet et al. [15], studied 585 patients >65 years old presenting to the primary care physician with dyspnea on exertion in the last 12 months but no confirmatory echocardiogram to determine the prevalence of HF in the elderly population. These patients underwent history examination, physical exam, electrocardiography and serum N-terminal pro-B-type natriuretic peptide testing with echocardiography performed only in patients with abnormal electrocardiography or serum N-terminal pro B-type natriuretic peptide >15 pmol/ml. Using European Society of Cardiology Guidelines, 92 patients had HF, 17/585 (2.9%) had HFrEF and 70/585 (12%) had HEpEF, while 5 (0.9%) had isolated right-sided HF. They also noted that HF was confused with chronic obstructive pulmonary disease and other respiratory problems in the primary care setting, which could lead to delay in diagnosis resulting in increased HF hospitalization, morbidity and mortality. Winter et al. [16], demonstrated the prevalence of unknown HF and LV dysfunction in patients >60 years of age with type 2 diabetes mellitus. Patients were diagnosed with HF on the basis of the European Society of Cardiology Guidelines. Of the 581 patients studied; 161 were diagnosed with HF: 28 (4.8%) had HFrEF and 133 (22.9%) had HFpEF. There was increased prevalence with age and a female sex preponderance. Barents et al. [17], reported the prevalence, misdiagnosis of chronic HF and the role of natriuretic peptides in 150 nursing home residents using the New York Heart Association classification and natriuretic peptides. HF diagnosis was confirmed in 24 residents, out of which, 15 had de-novo HF. Another important feature is that 13 out of 22, who were thought to have HF before screening, were rejected after the screening. Rutten et al. [18], reported the prevalence of HF in an elderly population with stable chronic obstructive pulmonary disease in whom patients had a detailed history, physical exam, chest radiography, ECG, echocardiography and pulmonary function testing. The diagnosis of HF was based on the European Society of Cardiology guidelines. Unrecognized HF was noted in 83 of 405 patients. Of the 83 patients, 42 had HFrEF and 41 had HFpEF. Most males who were younger had HFrEF, while HFpEF was common in elder women. Finally, Mosterd et al., included 5450 participants aged 55-95 years in the Rotterdam study to determine the prevalence of HF and symptomatic or asymptomatic LV systolic dysfunction. The overall prevalence of HF was 3.9%, with no difference between men and women, with an increased prevalence with age. The prevalence of LV systolic dysfunction was 3.7% and 2.5 times more common in men and most were asymptomatic. Most of the previous literature commented on the characteristics of populations screened for LV dysfunction or HF in variably symptomatic populations. There is little data characterizing patients with HFunEFand their outcomes, most likely due to national programs and core hospital measures advocating and requiring an assessment of LV ejection fraction. Patient data and outcomes are often not evaluated in this subset of patients.

4.2. Limitations

As this was a retrospective evaluation of a community-based outpatient practice, it is limited by numbers and population diversity and the inherent nature of a retrospective study. Second, data abstraction was based on a chart review of both inpatient and outpatient recorded symptoms and medications. ACM was based on chart review, national death index, and obituary review and may not have reflected all mortality. Cardiovascular mortality could not be discerned nor medication compliance. Furthermore, optimal dosing was not evaluated in this study with regard to outcomes because of changing dosages in many patients for unclear reasons despite chart abstraction. Third, readmissions for HF may have occurred outside the East Tennessee area. The additional hospital system’s records in the area were available and reviewed for admissions as patients may use both systems. Fourth, laboratory values were not available in all patients for BNP (7% missing), with non-significant differences among HF phenotypes ranging from 5-10%. Fifth, as there is inherent variability in the measurement of ejection fraction, this may lead to assigning patients to the wrong phenotype. As this was a retrospective study, there may be additional unknown differences among HF phenotype groups that were not accounted for.

4.3. Clinical Implications

Despite the requirement for hospital HF core measures for ejection fraction determination for all inpatients being observed in >90% cases, there are still patients with HF diagnosis who have do not have an assessment of ejection fraction. Furthermore, there appears to be an underutilization of guideline-based medications. The majority of such patients were seen by non-cardiovascular providers in this study, but is likely true in many populations. Unfortunately, these patients are readmitted less but have greater ACM. Demographic variables and laboratory variables, including brain natriuretic peptide, show little difference. Although this study portrays them as a separate phenotype, they clearly are not. More importantly, the lack of structural information afforded by echocardiography represents a lost opportunity and may have a role in their outcomes. It is also not inconceivable that non-cardiac causes may predominate their causes of death. The degree of detail and attention paid to ambulatory HF programs will need to be focused on all providers to afford patients’ optimal care and outcomes.