Stage D heart failure in the United States: A study from a national database

Abstract

Study objective

To study the clinical characteristics and outcomes of hospitalized Stage D HF as there is paucity of data regarding this class of HF patients.

Design

This is a non-interventional retrospective cohort study from a National database. Statistical analyses were conducted using STATA/MP, version 18 (StataCorp, TX).

Setting and Participants

We identified all the principal hospitalizations of Stage D or End Stage HF in adults ≥18 years of old for the calendar year 2019.

Main outcome measure

Characteristics, in-hospital mortality, 30-day readmissions, and causes of readmissions.

Results

We identified 28,356 Stage D or End Stage HF patients, with a mean age of 66 years. 31% were females. 56% had Acute kidney injury, 5% were on hemodialysis, 51% had atrial fibrillation, 26% developed cardiogenic shock, 9% required mechanical ventilator support, and 4% sustained cardiac arrest on index hospitalization. All-cause mortality was 14% during hospitalization. The 30-day readmission rate was 7.98%, with the most common cause of readmissions being heart failure (49.78%), cardiac and vascular device-related infections (4.86%), sepsis (2.65%), acute renal failure (2.49%), and ventricular tachycardia (2.09%). 8.16% died during readmission. The mean length of stay (LOS) for the index hospitalization was 11.59 days.

Conclusion

Stage D or End Stage HF poses a significant public health challenge in the United States and is associated with increased mortality, readmissions, and financial burden.

Highlights

• •Stage D HF was more common in younger males (18–65 yrs).
• •Stage D HF was associated with a high prevalence of sleep-disordered breathing, depressive and anxiety disorders.
• •Fourteen percent died on index hospitalization and the 30-day readmission rate was 8%.
• •Stage D HF imposed a significant public health and healthcare burden.

1. Introduction

Congestive Heart Failure (HF) is a global public health burden that constitutes significant morbidity and mortality. It is the leading cause of hospitalizations among patients aged 65 years or older and the global prevalence of HF study has shown that it has doubled since 1990 [1]. HF is defined as a complex clinical syndrome with symptoms and/or signs due to structural and/or functional cardiac abnormality and with evidence of congestion (pulmonary or systemic) and/or elevated natriuretic peptide level [2]. The prevalence of HF is expected to rise to 8.5 million in the United States by 2030 [3]. The prevalence is projected to rise further due to availability of various evidence-based treatment modalities that prolong survival in HF patients. This poses a significant financial health burden. It is projected that total cost of HF to rise to 69.8 billion dollars by 2030 [4].

The definitions and classification of HF by stages continue to evolve. Advanced HF includes the subset of HF patients with severe signs and symptoms of at rest, those with recurrent hospitalizations despite guideline-directed medical therapy or refractory or intolerant to the therapy, and those requiring advanced therapies such as consideration for transplant, mechanical circulatory support, or palliative care [2]. Advanced or End stage HF patients correspond to Stage D HF patients within the ABCD classification system of the ACC/AHA classification. There is limited evidence regarding prevalence of Stage D HF. The pooled analysis of multiple cohort studies (Multi-Ethnic Study of Atherosclerosis, Cardiovascular Heart Study, and Atherosclerosis Risk in Communities) estimated prevalence of Stage D HF to be 2.7% [3]. The progression to stage D HF leads to poor quality of life due to increasing symptoms burden and recurrent hospitalizations. The median life expectancy of Stage D HF is expected to be 6–12 months [5].

Advanced heart failure interventions in these end stage heart failure patients can prolong survival in selected patients. Registry data have shown that one year survival following orthotopic heart transplantation (OHT) in transplant eligible patients exceed 85%. Similarly, in transplant ineligible patients, survival among left ventricular assist devices (LVAD) recipients as a bridge to transplant as high as OHT and in those as destination therapy up to 80% [6], [7], [8]. Heart transplantation and LVAD implantation among patients with ionotrope-dependent Stage D HF showed significant improvement in survival and quality of life in another prospective study [9]. Here, we aim to study the characteristics, burden, and readmissions outcomes of Stage D HF patients using national database.

2. Materials and methods

2.1. Study population and outcomes

We used the National Readmission Database (NRD) 2021 to capture all the discharges with primary or secondary diagnosis of Stage D HF using International Classification of Disease Tenth revision Clinical Modification/Procedure Coding System (ICD 10 CM/PCS) code I50.84 [10]. Current universal definition and classification of HF do not clearly define end stage HF, however, for the coding purpose I50.84 for end stage HF is used to code patients with HF at Stage D or at the end of life [11]. Prior studies have also categorized End stage HF into Stage D of ACC/AHA classification system [12], [13]. These terms were used interchangeably in a study by Cascino et al. [14]. Based on these, we also have used stage D and End stage HF interchangeably for our study cohort. We identified all patients age ≥ 18 years utilizing this ICD code during the year 2021 to obtain a cohort of our study population with Stage D or End Stage HF. We excluded patients admitted in the month of December because of the design if the database such that it does not track re-admissions across the calendar year. Patients were excluded if admissions were following traumatic incidents, pre-planned elective admissions, and if younger than 18 years. The primary outcomes of interest were index mortality, clinical characteristics and comorbidities, and 30-day readmissions. Secondary outcomes were causes of readmissions, resource utilization (length of stay and hospital charge). All the comorbidities and procedures of interest on index admission for our cohort were obtained using ICD 10 CM/PCS. We included obstructive sleep apnea (OSA), central sleep apnea (CSA), sleep-related hypoventilation (SRH), and Cheyne-Stokes breathing to define sleep-disordered breathing (SDB) for our study. We followed the World Health Organization classification of Mental and Behavioral Disorders Clinical descriptions and diagnostic guidelines to define depressive disorders, anxiety disorders and post-traumatic stress disorders and appropriate ICD 10 codes were used to capture all the sub-groups within these disorders [15]. We included chronic obstructive pulmonary disease and bronchiectasis to define chronic pulmonary disease for our study. NRD 2021 variables were used to obtain patient's age, gender, insurance provider (Medicare, Medicaid, Private, uninsured), median income in patient zip code ($1–$38,999, $39,000–$47,999, $48,000–$62,999, ≥$63,000), Hospital urban-rural designation (Large metropolitan areas with at least 1 million residents, Small metropolitan areas with less than 1 million residents, Micropolitan areas, or non-urban residual). The patients were identified using a unique verified patient linkage number available in the Health Care Utilization Report (HCUP) NRD 2021 which identifies each discharge within the state for the calendar year 2021. The in-hospital and mortality on readmissions were obtained using the “died” variable available in the NRD 2021. Total hospital charge was defined as the amount that the hospitals billed. The NRD contains de-identified patient information and approval from IRB was not necessary. The study was conducted using the HCUP reporting guidelines and in accordance with the Declaration of Helsinki.

2.2. Statistical analysis

STATA MP, version 18 (StataCorp, TX) was used for all the statistical analyses. Pearson's chi-squared test was applied for categorical variables to examine group differences. For continuous variables, the Mann–Whitney U test or the student's t-test was used, as appropriate, to examine group differences in the stratified cohort. Chalrlson's Comorbidity Index (CCI) score was obtained directly from STATA to assess and account for comorbidities which are stratified as 1, 2, and ≥ 3. We used appropriate discharge weights (DISCWT) using the svyset command in STATA to account for the complex survey design of NRD that included stratification, clustering, and weighting to produce nationally representative estimates.

3. Results

Among the 28,356 patients hospitalized with Stage D or End Stage heart failure, the mean age was 66 years. Forty-three percent were between 18 and 65 years of age, 35% were 65–80 years, and 22% were older than 80 years. Most patients were male (69%). Fifty-six percent had acute kidney injury and 5% were on dialysis on index hospitalization. In terms of comorbidities, 45% had diabetes, 50% had dyslipidemia and 66% had chronic kidney disease, 51% had atrial fibrillation. During index hospitalization, 26% developed cardiogenic shock, 9% needed mechanical ventilation, and 4% sustained cardiac arrest (Table 1). Among hospitalized patients, 609 patients received Impella support, 1355 received a ventricular assist device (VAD), 1181 underwent intra-aortic balloon pump (IABP) placement, and 308 received extracorporeal membrane oxygenation (ECMO). Additionally, 0.54% had documented heart transplant status on admission, 2.65% were awaiting organ transplant, and 3.63% underwent heart or total artificial heart transplantation. Major depression and other depressive disorders were present in 14%, anxiety disorders in 16%, and 16% had sleep-disordered breathing. 28% received a palliative care encounter. Sixty-eight percent of patients had Medicare and 67% were from large metropolitan area. (See Fig. 1.)

Table 1.

Baseline characteristics of patients with Stage D Heart Failure.

Characteristics of Stage D HF	Total n (n%), (N = 28,356)
Age category
18–65 years	12,261 (43.24%)
65–80 years	9811 (34.60%)
>80 years	6281 (22.15%)
Age median [Inter Quartile Range]	66 [78,57]
Female	8844 (31.19%)
Obesity	5490 (19.36%)
Prior Myocardial Infarction	4996 (17.62%)
Prior Percutaneous Coronary Intervention	3440 (12.13%)
Prior Coronary artery bypass grafting	3051 (10.76%)
Hypertension	1024 (3.61%)
Diabetes	12,789 (45.10%)
Dyslipidemia	14,223 (50.16%)
Smoking	3449 (12.16%)
Acute Kidney injury	15,770 (55.62%)
Chronic kidney disease	18,785 (66.25%)
Currently on dialysis	1466 (5.17%)
Major depression and other depressive disorders	3983 (14.06%)
Anxiety disorder	4548 (16.04%)
Post-traumatic stress disorder	264 (0.93%)
Central sleep apnea	77 (0.27%)
Sleep disordered breathing	4611 (16.27%)
Obstructive Sleep Apnea	4548 (16.04%)
Mechanical ventilation	2576 (9.09%)
Cardiac arrest	1120 (3.95%)
Cardiogenic shock	7381 (26.04%)
Ventricular Fibrillation	839 (2.96%)
Ventricular Tachycardia	7244 (25.59%)
Peripheral vascular disease	2139 (7.54%)
Chronic liver disease	4203 (14.82%)
Presence of Automatic Implantable Cardioverter Defibrillator	7163 (25.26%)
Debility	360 (1.27%)
Charlson's comorbidity index (CCI) score 1	1521 (5.36%)
Charlson's comorbidity index (CCI) score 2	3040 (10.72%)
Charlson's comorbidity index (CCI) score 3 or more	23,808 (83.92%)
Rehab transfer	715 (2.52%)
Insurance provider
Medicare	19,313 (68.11%)
Medicaid	4631 (16.33%)
Private	3885 (13.70%)
Uninsured	527 (1.86%)
Median income in patient zip code
$1–$38,999	9349 (32.97%)
$39,000–$47,999	7571(26.70%)
$48,000–$62,999	6289 (22.18%)
≥$63,000	5147 (18.15%)
Hospital urban-rural designation
Large metropolitan areas with at least 1 million residents	19,132 (67.47%)
Small metropolitan areas with less than 1 million residents	7537 (26.58%)
Micropolitan areas	1219 (4.30%)
Not metropolitan or micropolitan (non-urban residual)	468 (1.65%)
Mechanical circulatory support	2226 (7.85%)
Ventricular Assist Device use (VAD)	1355 (4.78%)
Impella	609 (2.15%)
Extra-corporeal membrane oxygenation (ECMO)	308 (1.09%)
Intra-aortic Balloon Pump (IABP)	1181 (4.16%)
Prior cancer	2124 (7.49%)
Iron deficiency anemia	3045 (10.74%)
Chronic pulmonary disease	7546 (26.61%)
Pacemaker	1437 (5.07%)
Atrial fibrillation	14,578 (51.41%)
Delirium	763 (2.69%)
Dementia	2081 (7.34%)
Heart transplant status	153 (0.54%)
Heart transplantation	1.029 (3.63%)
Awaiting Organ transplant	751 (2.65%)
Mortality on index	14%
Mean LOS (days)	11.59

Fig. 1.

Burden of Stage D/End Stage HF in the United States.

Overall, 14% (n = 3972) died during the index hospitalization. The mean length of stay (LOS) was 11.59 days, and the mean total hospital charge was $187,697 during the index hospitalization. The 30-day readmission rate was 7.98%, with the most common causes being heart failure (49.78%), cardiac or vascular device–related infections (4.86%), sepsis (2.65%), acute renal failure (2.49%), and ventricular tachycardia (2.09%). 8.16% died during readmission (Table 2). Readmitted patients accounted for a combined 19,327 hospital days, with total hospitalization charges reaching $288 million.

Table 2.

Outcomes on readmissions.

Outcomes on readmissions	Total (n%), (N = 1955)
30-day readmissions	7.98%
Mortality on readmissions	8.16%
Causes of Readmissions
1.Heart failure	49.78%
2.Cardiac or vascular device–related infections	4.86%
3.Sepsis	2.65%
4.Acute renal failure	2.49%
5.Ventricular tachycardia	2.09%

4. Discussion

Our study showed that Stage D HF imposes a significant healthcare burden. Most of the patients were primarily males, with a median age of 66 years. It was associated with a substantial comorbidity burden with high prevalence of sleep-disordered breathing, mental health disorders, renal disease, diabetes and dyslipidemia as shown in Table 1. The mortality on index was 14 %, and 28 % received a palliative care encounter during index hospitalization. Eight percent were readmitted in 30 days with heart failure being the most common cause of readmission. A total of 7.85% of patients required mechanical circulatory support, including 4.7% VAD implantation.

There is limited data regarding the characteristics and outcomes of patients with End stage or Stage D HF. Stage D HF often results in end-organ dysfunction, hepatic and renal insufficiency, cachexia, fatigue, and reduction in functional capacity and quality of life with variable clinical courses. It requires a multidisciplinary team of heart failure specialists, cardiac surgeons, and palliative or hospice care clinicians to make individualized decisions regarding advanced interventions [16].

In a multi-registry study of chronic Stage D HF (n = 1433) from the Acute Decompensated Heart Failure National Registry Longitudinal Module (ADHERE LM), the mean age was 69.6 ± 13.2 years, with the majority being male (65%). Comorbidities included diabetes (51%), dyslipidemia (65%), coronary artery disease (73%), chronic pulmonary disease (38%), hypertension (75%), chronic renal insufficiency (52%), and arrhythmias (62%). Twenty-one percent received VAD therapy [5]. They also found that stage D HF patients tended to be more younger males with dyslipidemia, coronary artery disease and chronic kidney disease compared to heart failure patients in the ADHERE CM or OPTIMIZE-HF registries [5].

Similarly, in the study of advanced HF patients from Olmsted County, Minnesota (n = 936), mean age was 76.9 ± 14.6 years, and 55.5% were males. Hypertension was present in 88.6%, hyperlipidemia in 73.1%, coronary artery disease in 70.6%, diabetes in 44.3%, COPD in 57.4%, Obesity in 43%, and stage 3 or higher chronic kidney disease in 72%. Additionally, 6.9% had dementia. During follow-up, 3.3% received LVAD and 2.9% underwent heart transplantation [17].

Sleep-disordered breathing, in particular, Obstructive (OSA) and central sleep apnea are highly prevalent in HF patients and negatively affect the outcomes due to heightened sympathetic activation [18]. In our study, 16% of hospitalized patients had sleep-disordered breathing with majority having OSA. Similarly, depression and anxiety disorders lead to poor heart failure outcomes secondary to elevated inflammatory biomarkers and poor dietary and medication adherence in these vulnerable patient population [19]. We observed that 14% had depressive disorders and 16% had anxiety disorders in our cohort. In the Minnesota cohort, 27.5% had depression.

Our study showed that mortality during the index hospitalization for Stage D HF was 14%, with mean length of stay of 11.59 days and 30-day readmission rate was 7.98%, and the most common reason for readmission was heart failure. In the ADHERE LM study, the mean follow-up was 364 days (IQR 147–575), during which 59% of patients were hospitalized and 31% died [5]. In the Minnesota cohort (2007–2017), one year mortality was 49.9% and there were 3138 all-cause hospitalizations during the follow up period. Of these, 32.1% were readmitted for HF [17]. Other common causes of hospitalization included pneumonia, respiratory failure, sepsis and gastrointestinal illnesses.

With the increasing prevalence of HF, the associated healthcare costs and resource utilization will pose considerable challenges, especially for those approaching end-stage HF who experience repeated hospitalizations. We observed that the mean total hospital charge for Stage D HF was $187,697 during the index hospitalization. Readmitted patients accounted for a combined 19,327 hospital days, with total hospitalization charges reaching $288 million. Early Palliative care referrals during the index hospitalization are important to establish goals of care for patients with end-stage HF before considering advanced health care interventions that can be costly and may affect quality of life. Palliative care has been shown to improve patient and family satisfaction, enhance symptom management, and reduce healthcare utilizations and costs [20]. In our study, 28% of the hospitalized patients received palliative care consultations.

Being an administrative database study, our study has several limitations. Discharge diagnoses in the NRD are based on ICD 10 codes, which are subject to coding inaccuracies. Patients transferred or readmitted in other states could not be captured due to inherent database limitations. We were also unable to obtain clinical symptoms, physical examination, echocardiographic, and laboratory data, as these variables are not available in the NRD.

5. Conclusion

Stage D HF poses a significant public health challenge and is associated with substantial in-hospital mortality, 30-day remissions and financial burden. Additional studies are needed to further characterize this unique and vulnerable HF population.

CRediT authorship contribution statement

Govinda Adhikari: Writing – review & editing, Writing – original draft, Validation, Supervision, Formal analysis, Data curation, Conceptualization. Nischit Baral: Writing – review & editing, Writing – original draft, Validation, Formal analysis, Data curation, Conceptualization. Dipendra Dahal: Writing – review & editing, Writing – original draft, Validation, Data curation, Conceptualization.

Declaration of competing interest

Govinda Adhikari: None.

Nischit Bral: None.

Dipendra Dahal: None.

During the preparation of the figure, the authors used [Microsoft Copilot] in order to [generate a journal infographic]. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.