Readmission Risk Factors and Heart Failure With Preserved Ejection Fraction

Methods

This study was approved by the Institutional Review Board at the Ascension Macomb Hospital in Warren, Michigan.

Electronic health records were retrospectively reviewed in October 2018 by several authors (J.R., F.R., H.G., and I.P.) for patients admitted between August 1, 2017 and March 1, 2018 to Ascension Macomb-Oakland Hospital with a diagnosis of HFrEF greater than 40% who were screened but did not qualify for the ongoing, nationwide CONNECT-HF trial, for which our hospital was enrolling patients at the time of this report and which was being conducted by Duke Clinical Research. CONNECT-HF is a quality improvement initiative based on clinical outcomes, medication use, quality-of life assessments for HF patients with reduced left ventricular EF.⁶

Patients 18 years or older who fell into 1 of 2 classifications (New York Heart Association Class [NYHA] II-IV or American College of Cardiology/American Heart Association [ACC/AHA] Stage B-D) and who had a life expectancy greater than 6 months were included. NYHA classification separates patients with heart failure based on symptom and ACC/AHA separates patients with heart failure based on structural abnormality.¹ Patients were excluded if they had terminal illness other than HF, a prior heart transplant or were on a transplant list, a current or planned placement of a left ventricular assist device, chronic kidney disease requiring hemodialysis, inability to use mobile applications, or inability to participate in longitudinal follow up.

Results

Of 492 patients treated for HFrEF during the retrospective study period, 250 patients had left ventricular ejection fraction >40, and 212 patients met our inclusion and exclusion criteria (Figure 1). Demographic distributions are shown in Table 1. Our patients were 59.4% (126) women, 40.5% (86) men, had a mean age of 72 years (standard deviation [SD], +/- 13.3), and had an average BMI of 33.4 kg/m². Of the 212 patients included, 123 had a BMI greater than 30 kg/m². Detailed chart review found a total of 45 (21.23%) patients readmitted for HFpEF within 90 days of initial discharge. Of the 45 readmitted, 32 (71.11%) patients were readmitted within 30 days of initial discharge.

Figure 1.

Inclusion and exclusion criteria for heart failure patients during the retrospective study period.*Terminal illness other than HF, prior heart transplant or were on a transplant list, current or planned placement of a left ventricular assist device, chronic kidney disease requiring hemodialysis, inability to use mobile applications, and unable to participate in longitudinal follow up.

Clinical findings of pertinent underlying CV diseases included: 22 (10.33%) patients with a history of previous myocardial infarction (MI); 101 (47.41%) with a history of arrhythmia (atrial fibrillation/atrial flutter); 32 (15.02%) with a history of stroke (CVA); 27 (12.68%) with known peripheral vascular disease (PVD); 140 (66.99%) with documented pulmonary hypertension; 94 (44.13%) with COPD; and 172 (80.75%) with a documented history of HF.

Clinical laboratory data analyzed in our study included hemoglobin, BNP, troponin on admission, and peak troponin levels (Table 2). The median BNP of HFpEF patients was 2727 pg/mL, while the median troponin on admission was 0.03ng/ml and the median peak troponin level was also 0.03 ng/ml. Interestingly, both baseline and peak mean values were found to be statistically significantly higher in the patients who were not readmitted within 90 days (baseline, P=.035; peak, P=.014). Significant anemia was defined as a hemoglobin less than 10 mg/dL, which was found in 146 (68.86%) of the reviewed charts and noted to be a statistically significant factor for readmissions (P=.029)

Echocardiographic findings were included to assess the significance of valvular stenosis or regurgitation and pulmonary hypertension on HFpEF patients. Echocardiographic information was reviewed on all 212 patients and all abnormalities were recorded based on the final signed report by a cardiologist. In our studied population, 38 patients were found to have moderate-to-severe pulmonary hypertension (defined as systolic pulmonary pressure greater than 50 mmHg) with concomitant anemia. Of those 38 patients, 16 (42%) were readmitted for HF within 90 days (P=.24).

Discharge medication review found that ACEI or angiotensin receptor blockers (ARB) were more frequently prescribed on discharge in the group of patients who were not readmitted for heart failure within 90 days of initial discharge (80 [48.8%] vs. 13 [28.9%]; P=.017).

The median LACE score was higher in patients who were readmitted within 90 days vs those were not readmitted (16 vs 13; P<.001). The median Charlson comorbidity index was also higher for those the readmission group compared to those who were not readmitted (6 vs 4; P<.002), reaffirming utility of LACE index score and Charlson comorbidity index to stratify patients at risk for readmission. Results are reported in Table 3.

Several factors proved to be statistically significant after performing a side-by-side comparison of the above variables in patients who were readmitted within 90 days for heart failure and those who were not. PVD was more common in those readmitted within 90 days (12 [26.7%] vs.15 [9%]; P<.002), as was any valvular heart disease on TTE (31 [68.9%] vs. 86 [51.1%]; P=.037), anemia (37 [82.2%] vs. 109 [65.3%]; P=.029), moderate-to-severe pulmonary hypertension (19 [44.2%] vs. 38 (23%); P=.049) than the group who was not readmitted. Review of patients’ most recent echocardiogram showed that severe tricuspid regurgitation was also more common in the readmission group (11 [24.4%] vs. 10 [6%]; P=.001).

Discussion

Our results provide several important findings that can further improve collective understanding of HFpEF pathophysiology and the impact of its various comorbidities. The majority of our study population had obesity, pulmonary hypertension, anemia, and arrhythmias. Importantly, obesity is among underlying factors of HFpEF, as it is a significant contributor to systemic inflammation leading to myocardial remodeling.⁹ Based on a 30-day readmission rate of 15% and a 90-day readmission rate of 21.3%, our main goal was to identify the predominant readmission risk factors in HFpEF patients in our community.

Our findings correlate well with existing knowledge about the impact of higher LACE and Charlson comorbidity indices on readmission rate.^7,8 These scores could be used as a guide to help providers identify patients at risk for readmission, effectively coordinate care, and provide adequate community resources to reduce readmission rates and hopefully improve patient longevity.

In our study, patients with PVD were more likely (P<.002) to be readmitted within 90 days of discharge, suggesting that comorbidities in patients with HFpEF lead to microvascular inflammation, which adversely affects the adjacent cardiomyocytes through decreased nitric oxide bioavailability, reduced cyclic guanosine monophosphate availability, and altered phosphorylation of titin. Microvascular ischemia, concentric left ventricular remodeling, and fibrosis from endothelial to mesenchymal transition contribute further to diastolic dysfunction in these patients.^2,9

After review of the laboratory data on each initial heart failure admission, brain natriuretic peptide (BNP) was not found to be significant as a readmission factor. This insignificance could be a result of the average BMI of patients in our study (33.4 kg/m²) being classified as obese. There is an inverse relationship between obesity and BNP; obesity suppresses plasma BNP levels in acute heart failure.^1,10 Obese HFpEF patients were shown to have a lower BNP level, but the range of values obtained was very large (49 to 70001), which could make it difficult to show statistical significance.¹² Troponin values were also assessed for baseline and peak levels on initial admission. Interestingly, both baseline and peak mean values were found to be statistically significantly higher in the patients who were not readmitted within 90 days. It is possible that patients who were not readmitted received a more comprehensive workup or medications that were more aggressively titrated, resulting in statistically-significant higher average levels of troponin found at both baseline and peak in our study.

Readmission rates in our study were also higher in patients found to have significant anemia (hemoglobin <10 mg/dL) on initial admission (P=.029). Symptomatic HFrEF patients with iron deficiency anemia who were treated accordingly have been associated with reduced hospitalizations, improved exercise capacity, and quality of life.¹³ It is unclear whether HFpEF patients will demonstrate similar association to iron deficiency anemia treatment in any of the prospective trials currently in progress.¹⁴ Although the goal of this study was not to evaluate the etiology of anemia, impaired myocardial energetics in HFpEF patients suggest that targeting cardiometabolic functional abnormalities may be an important approach in these patients in the future.²

Likewise, patients with TTE findings signifying severe tricuspid regurgitation with estimated moderate-to-severe pulmonary hypertension, defined as a systolic pulmonary pressure greater than 50 mmHg, were more likely to have a HF readmission within 90 days. Although there was no statistical significance in this finding, partly due to small sample size, this trend certainly has meaningful clinical implications.

The presence and severity of pulmonary hypertension is strongly associated with the development of RV dysfunction. Beyond increased pulmonary pressure, RV-arterial uncoupling plays an important role further contributing to worse outcomes in HFpEF. Therapy directly aimed at pulmonary hypertension to reduce RV afterload, thereby improving RV function in HFpEF, has not been shown to decrease morbidity and mortality in HFpEF patients.^2,15

Lastly, we compared all categories of cardiac medications given to patients in our study upon discharge after their initial HF admission. Patients discharged on an ACEI/ARB were noted to have a lower risk of readmission (80 [48.8%] vs. 13 [28.9%]; P<.017). Although this is statistically significant and supports previous evidence that ACEI/ARB may be beneficial in HFpEF patients from trials such as the CHARM-Preserved Trial,¹⁶ the I-PRESERVE Trial,¹⁷ and a substudy of the BACH trial,¹⁸ any clinical significance remains uncertain due to the multiple confounding factors of the patients in our retrospective study.^16-18 Based on these previous larger studies, we can suggest, but do not confidently conclude, that readmission rates at our community hospital could be reduced if patients admitted for CHF with an EF >40% were discharged on an ACEI/ARB when clinically indicated. Additionally, data from the CHAMPION trial¹⁹ suggests that addressing volume overload via diuresis may be a fundamental component of the optimal management of HFpEF.

Conclusion

HF exacerbations continue to be a large obstacle affecting not only patients' quality of life but causing significant financial burden on our healthcare systems. Although reducing left heart congestion along with left atrial hypertension carries prognostic benefit in HFpEF, it is increasingly recognized that various conditions, especially obesity, anemia, vascular disease, and pulmonary hypertension, are underlying contributors to cardiac remodeling. Early recognition and amelioration of these risk factors is of paramount importance. Our study findings affirm the need to continue to research into the pathophysiology and associated comorbidities for patients with HFpEF to improve their quality of life and lower healthcare system costs.

Author Contributions

All authors provided substantial contribution to conception and design, acquisition of data, or analysis and interpretation of data; all authors drafted the article or revised it critically for important intellectual content; all authors gave final approval of the version of the article to be published; and all authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.