Patient‐Reported Health Status Among Patients With Heart Failure With Improved Ejection Fraction

Abstract

Background

The prevalence of heart failure with improved ejection fraction (HFimpEF) is growing. The association of ejection fraction (EF) recovery and changes in health status has not been previously reported. We aimed to characterize patient‐reported health status among patients with HFimpEF, heart failure with reduced ejection fraction, and heart failure with midrange (HFmrEF) or preserved ejection fraction (HFpEF).

Methods

We identified patients with heart failure with 2 clinic visits, who completed a Kansas City Cardiomyopathy Questionnaire‐12 from August 2020 to October 2023. HFimpEF was defined as most recent EF >40% from a preceding echocardiogram ≥30 days prior with EF ≤40%. We analyzed Kansas City Cardiomyopathy Questionnaire‐12 Overall Summary Score across EF classifications with and without adjustment for patient characteristics via multivariable linear regression. We calculated the R ² to determine the impact of clinical characteristics on variation in health status among patients with HFimpEF.

Results

A total of 2519 patients were included, of which 18.7% had HFimpEF, 55.7% had HFmrEF/HFpEF, and 25.6% had HFrEF. Patients with HFimpEF were less likely to be women compared with patients with HFmrEF/HFpEF (41% versus 58%, P<0.01). Overall Summary Score was 5.2 points lower among patients with HFrEF versus HFimpEF (95% CI, −8.2 to −2.3); P=0.010) but patients with HFimpEF had similar scores as HFmrEF/HFpEF (−1.1 [95% CI, −3.7 to 1.6]; P=0.43). A multivariable regression model explained 16% of the variation in the Overall Summary Score in the HFimpEF subgroup.

Conclusions

Despite improvement in EF, patients with HFimpEF continue to have impaired health status similar to HFpEF. Further research is needed to understand and improve health status among patients with HFimpEF.

Nonstandard Abbreviations and Acronyms

CSS

Clinical Summary Score

DELIVER

Dapagliflozin Evaluation to Improve the Lives of Patients With Preserved Ejection Fraction Heart Failure

HFimpEF

heart failure with improved ejection fraction

HFmrEF

heart failure with midrange ejection fraction

HFpEF

heart failure with preserved ejection fraction

HFrEF

heart failure with reduced ejection fraction

KCCQ

Kansas City Cardiomyopathy Questionnaire

PRO‐HF

Patient‐Reported Outcome Measurement in Heart Failure Clinic

Research Perspective.

What Is New ?

• Heart failure (HF) with improved ejection fraction is a rapidly growing subgroup of patients living with HF in the modern era of guideline‐directed medical therapy; however, these patients are still at substantial risk for recurrent HF events.

• Our study shows that patients with HF with improved ejection fraction have impairment in their functional status similar to patients with HF with preserved ejection fraction that is not explained by sociodemographic characteristics, comorbidities, or medications.

What Question Should be Addressed Next?

• There is a need for better understanding of impaired health status to identify interventions to improve health status among patients with HF with improved ejection fraction.

In the context of continued advancements in medical and device therapies, patient with a history of heart failure with reduced ejection fraction (HFrEF) who experience recovery of left ventricular ejection fraction (LVEF) represent an increasingly prevalent cohort of patients along the heart failure (HF) spectrum. Heart failure with improved ejection fraction (HFimpEF) has been defined as a new HF subtype in the 2022 American Heart Association/American College of Cardiology Heart Failure guidelines ¹ to reflect this expanding population. With the availability and implementation of guideline‐directed medical therapy and devices for patients with HFrEF, the estimated proportion of patients with HFimpEF has been shown to be as high as 40% ² and is expected to continue to increase.

HFimpEF occurs in a heterogeneous population and historically has not been well‐defined, often reclassified among other subtypes of HF. ³ Although existing data suggest that patients with HFimpEF are more likely to have favorable clinical outcomes as compared with patients with persistent HFrEF, ⁴ whether recovery of ejection fraction (EF) signifies remission of the HF syndrome is not well known. In fact, persistent limitations in exercise capacity, contractile reserve and risk for later decline in LVEF and subsequent reclassification to HFrEF have been described in small cohorts of patients with HFimpEF. ⁵ , ⁶ , ⁷ As a result, maintaining guideline‐directed medical therapy in this growing subset of patients has remained a cornerstone of their management. ⁸ Patient‐reported health status is an important prognostic and patient‐centric factor in patients with HF, ⁹ , ¹⁰ and this has not been described in patients with HFimpEF treated with contemporary guideline‐directed medical therapy as prior studies were conducted in small patient cohorts. ¹¹ Thus, larger studies of quality of life and functional health status among patients with HFimpEF are needed.

This study aims to describe patient‐reported health status in ambulatory patients with HFimpEF across diverse patients seen in an academic HF outpatient practice. In this study, we evaluated and compared patient‐reported health status among patients with HFimpEF, HFrEF, and heart failure with preserved ejection fraction (HFpEF).

Methods

Data Source

This study used data derived from the electronic health record from a single academic health care system including patient demographics, diagnoses, laboratory values, medication orders, echocardiographic parameters, procedures, and patient‐completed health status questionnaires. This study was approved by the Stanford University Institutional Review Board; informed consent by participants was not required. Data for this study are available from the corresponding author upon reasonable request.

Study Population and Patient‐Reported Health Status Metrics

Patients aged at least 18 years with a diagnosis of HF based on International Classification of Diseases, Tenth Revision (ICD‐10) codes and at least 2 outpatient cardiology clinic visits between August 1, 2020, and October 31, 2023, were identified. For each patient, we identified the last clinic visit with a completed Kansas City Cardiomyopathy Questionnaire (KCCQ)‐12 questionnaire and excluded patients without a completed KCCQ‐12.

We used the KCCQ‐12 score, which is a 12‐question abbreviated version of the original 23‐question KCCQ tool. The KCCQ‐12 encompasses 4 domains of patient‐reported health status related to HF: (1) symptom frequency, (2) physical limitations, (3) social limitations, and (4) quality of life. The KCCQ‐12 includes 4 domain scores and an Overall Summary Score (OSS), which is an average of the 4 domain scores. Scores ranging from 0 to 25 represent the most severe limitations in quality of life and poor health status, 25 to 49 represent limited to fair health status, 50 to 74 represent fair to good health status, and 75 to 100 represents excellent health status with no limitations. The Clinical Summary Score (CSS) is calculated as an average of the physical limitation and symptom frequency scores. The KCCQ‐12 was collected as part of the PRO‐HF (Patient‐Reported Outcome Measurement in Heart Failure Clinic) trial, ¹² which was a pragmatic, open‐label, patient‐level randomized trial comparing routine assessment with KCCQ‐12 to usual care at a tertiary center. Patients were randomized to complete the KCCQ‐12 before each HF clinic visit or undergo usual care. The baseline characteristics of patients randomized to KCCQ‐12 during this trial were comparable to patients randomized to usual care. ¹³ After the completion of the PRO‐HF trial, all patients routinely receive a KCCQ‐12 form to complete, as this has become standard of practice.

LVEF Categories

The primary exposure of interest was LVEF category based on each patient’s most recent LVEF at the time of the KCCQ‐12 questionnaire. All patients were required to have at least 2 echocardiograms since August 2020 to ensure accurate classification of patients into the HFimpEF group on the basis of LVEF. Patients were categorized as follows ¹ : HFrEF if LVEF was ≤40%, heart failure with midrange ejection fraction (HFmrEF) if LVEF was >40% and <50%, HFpEF if LVEF ≥50%, and HFimpEF if a patient had a prior LVEF ≤40% at least 30 days before a subsequent echocardiogram with LVEF ≥40%. We compared patients with HFimpEF to patients with HFrEF and a combined HFmrEF/HFpEF category. We performed several sensitivity analyses. First, we stratified patients with HFmrEF and HFpEF to evaluate differences between these groups. To ensure that patients had a meaningful improvement in their LVEF (ie, experienced an improvement in LVEF of at least 10% to minimize the potential confounding effect of patients with minimal change in their LVEF), we performed sensitivity analyses were performed defining HFimpEF as prior LVEF ≤40% at least 30 days before a subsequent echocardiogram with LVEF ≥50%.

Statistical Analysis

We compared patient sociodemographic and clinical characteristics across LVEF categories (HFimpEF versus HFrEF versus HFmrEF/HFpEF). We described characteristics using means and standard deviations for normally distributed continuous variables, medians and interquartile ranges for skewed continuous variables, and counts and frequencies for categorical variables. Differences according to each LVEF category were evaluated by ANOVA for continuous, normal variables, the Kruskal–Wallis test for nonnormal variables, and χ² tests for binary variables.

We then evaluated the distribution of KCCQ‐12 scores (domain score, OSS, and CSS) across LVEF categories with and without adjustment for other patient characteristics. Nested models were generated as follows: (1) model 1 with adjustment for other sociodemographic characteristics, (2) model 2 with adjustment for medical comorbidities, and (3) model 3 with adjustment for HF medications. We then evaluated the associations between KCCQ‐12 scores and patient sociodemographic data and medical comorbidities among patients with HFimpEF. We measured the overall discrimination of patient characteristics on the KCCQ‐12 OSS via the R ². Statistical testing was performed with Stata SE 17 (StataCorp, College Station, TX). We used a 2‐sided α of 0.05 for statistical significance.

Covariates included sociodemographic characteristics including age, race, ethnicity, sex, and Centers for Disease Control Social Vulnerability Index for each patient on the basis of patient residential zip code. The Social Vulnerability Index indicates the relative vulnerability of every US Census tract. There are 4 categories that are defined on the basis of census measures for a given neighborhood: socioeconomic status, household, racial and ethnic minority status, and housing type/transport. Each category is synthesized into a measurement of the degree of vulnerability of an individual patient’s neighborhood, with 100% representing the most vulnerable and 0% representing the least vulnerable classification. Additional covariates included clinical comorbidities, Charlson comorbidity index, ¹⁴ and HF medications (β blockers, renin–angiotensin system inhibitors, mineralocorticoid receptor antagonists, sodium–glucose cotransporter‐2 inhibitors, and loop diuretics) that were ordered within the 15 months before the clinic visit of interest.

Results

A total of 13 453 patients had an ICD‐10 diagnosis of HF and at least 2 cardiology clinic visits from August 2020 through October 2023 (Figure 1). Of these patients, 2693 had a completed KCCQ‐12 questionnaire. We excluded 174 patients without at least 2 LVEF measurements by echocardiography within the study window. Of the remaining cohort, 471 (18.7%) patients were classified as HFimpEF, 645 (25.6%) as HFrEF, and 1403 (55.7%) as HFpEF/HFmrEF.

Figure 1. Consolidated Standards of Reporting Trials diagram.

Consolidated Standards of Reporting Trials diagram for the study consisted of patients who completed a KCCQ‐12 questionnaire and had at least 2 echocardiograms since August 2020.HFimpEF indicates heart failure with improved ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; and KCCQ, Kansas City Cardiomyopathy Questionnaire.

Patient Characteristics

Patients with HFmrEF/HFpEF were more likely to be women (n=807 [58%]) compared with patients with HFimpEF (n=191 [41%]) or HFrEF (n=186 [29%]) (P<0.01; Table 1). Among patients with HFimpEF, 17% identified as Asian, 5% identified as Black, 3% identified as Pacific Islander, and 59% identified as White. Neighborhood Social Vulnerability Index was similar across groups. Patients with HFimpEF and HFrEF were more likely to have a prior HF hospitalization compared with patients with HFmrEF/HFpEF (HFimpEF [47%] and HFrEF [45%] versus HFmrEF/HFpEF [19%]; P<0.01). The most common comorbidities among patients with HFimpEF included history of prior myocardial infarction (34%), peripheral artery disease (51%), and renal disease (39%).

Table 1.

Baseline Characteristics Stratified by HF Phenotype

	Patients with HFimpEF (n=471)	Patients with HFrEF (n=645)	Patients with HFmref/HFpEF (n=1403)	P value
Adjusted R 2	0.0123
Age, y	64.9	64.5	65.1
Female sex	191 (40.6)	186 (28.8)	807 (57.5)	0.000
Weight, lbs	180.8±55.9	188.9±60.6	180.6±53.4	0.045
Race				0.000
Asian	81 (17.4)	104 (16.1)	236 (16.8)
Black	23 (4.9)	64 (9.9)	46 (3.1)
Native American	2 (0.4)	4 (0.6)	7 (0.5)
Other	68 (14.4)	98 (15.2)	202 (14.4)
Pacific Islander	13 (2.8)	17 (2.6)	23 (1.6)
White	279 (59.2)	345 (53.5)	848 (60.4)
Unknown	4 (0.9)	13 (2.0)	41 (2.9)
Ethnicity				0.016
Non‐Hispanic	410 (87.1)	556 (86.2)	1196 (85.3)
Hispanic	57 (12.1)	76 (11.8)	156 (11.1)
CDC SVI categories
Below 150% poverty	0.32 (0.2)	0.4 (0.3)	0.3 (0.3)	0.000
Unemployed	0.5 (0.3)	0.5 (0.3)	0.4 (0.2)	0.314
CDC overall SVI
Socioeconomic status	0.3±0.3	0.4±0.3	0.3±0.3	0.000
Household	0.4±0.3	0.5±0.3	0.4±0.3	0.08
Racial and ethnic minority status	0.7±0.2	0.7±0.2	0.7±0.2	0.251
Housing type transport	0.6±0.3	0.6±0.3	0.6±0.3	0.978
Married	305 (64.9)	366 (56.7)	881 (62.8)	0.027
Comorbidities
History of any HF hospitalization	223 (47.4)	293 (45.4)	269 (19.2)	0.000
History of HF hospitalization in the past year	79 (16.8)	168 (26.1)	127 (9.1)	0.000
History of HF hospitalization in the past month	15 (3.2)	31 (4.8)	28 (2.0)	0.000
Charlson comorbidity index	3.4 (2.5)	3.0 (2.2)	3.0 (2.5)	0.000
Charlson index comorbidities
Acute myocardial infarction	161 (34.2)	278 (43.1)	358 (25.5)	0.000
COPD	120 (25.5)	132 (20.5)	341 (24.3)	0.090
Renal disease	184 (39.1)	273 (42.3)	453 (32.3)	0.000
Diabetes	178 (37.8)	229 (35.5)	356 (25.4)	0.000
Cancer	94 (20.0)	78 (12.1)	298 (21.2)	0.000
Dementia	19 (4.0)	21 (3.3)	38 (2.7)	0.343
Cardiovascular disease	75 (15.9)	122 (18.9)	218 (15.5)	0.150
Liver disease	73 (15.5)	88 (13.6)	198 (14.1)	0.665
Peripheral artery disease	239 (50.7)	414 (64.2)	567 (40.4)	0.010
History of smoking				0.280
Never smoker	312 (66.2)	398 (62.1)	941 (67.5)
Current smoker	8 (1.7)	26 (4.1)	29 (2.1)
Former smoker	318 (33.8)	498 (33.1)	742 (29.3)
Medications
β blocker	297 (65.4)	433 (68.0)	387 (35.2)	0.000
RAAS inhibitor	394 (86.8)	546 (85.7)	786 (71.5)	0.000
ARNI	239 (52.6)	374 (58.7)	139 (12.7)	0.000
MRA	304 (67.0)	456 (71.6)	502 (45.7)	0.000
SGLT2i	232 (51.1)	416 (65.3)	352 (32.0)	0.000
Loop diuretic	229 (50.4)	399 (62.6)	592 (53.9)	0.000

All values are presented as counts (percentage of total) or mean±SD. P values were obtained by χ² for categorical variables and ANOVA for continuous variables. R ² reflects the fully adjusted model 3.

ARNI indicates angiotensin receptor–neprilysin inhibitor; CDC, Centers for Disease Control and Prevention; COPD, chronic obstructive pulmonary disease; HF, heart failure; HFimpEF, heart failure with improved ejection fraction; HFmrEF, heart failure with midrange ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; MRA, mineralocorticoid receptor antagonist; RAAS, renin–angiotensin–aldosterone system; SGLT2i, sodium–glucose cotransporter‐2 inhibitor; and SVI, Social Vulnerability Index.

KCCQ‐12 Scores Across HF Subtypes

The median KCCQ‐12 OSS was 80.3 among patients with HFimpEF compared with 73.9 for patients with HFrEF and 80.8 among patients with HFmrEF/HFpEF (Figure S1). Among patients with HFimpEF, 275 (58.4%) had a KCCQ‐12 OSS score between 75 and 100, 101 (21.4%) between 50 and 74, 62 (13.2%) between 25 and 49, and 33 (7.0%) between 0 and 24. Without adjustment for patient characteristics, the KCCQ‐12 OSS was not significantly different between patients with HFimpEF and those with HFmrEF/HFpEF (Figure 2). The KCCQ‐12 OSS was 7.2 points higher for patients with HFimpEF than those with HFrEF (P=0.000). Figure 2 illustrates KCCQ‐12 across EF classifications. These findings persisted after adjustment for sociodemographic characteristics, medial comorbidities, and HF medical therapy. Using model 3, the poorest functional status by KCCQ‐12 was significantly associated with loop diuretics (−13.8 [95% CI, −16.0 to −11.6]; P=0.000), identifying as Black (−10.0 [95% CI, −25.2 to −2.8]; P=0.000) and patients with dementia (−8.5 [95% CI, −13.1 to −4.0]; P=0.000), whereas the best functional status was significantly associated with male sex (6.5 [95% CI, 4.4–8.5]; P=0.000) (Figure 3).

Figure 2. Differences in KCCQ‐12 Overall Summary Score by heart failure phenotype.

A, The kernel density plot demonstrates the density of histogram responses across a range of KCCQ‐12 scores from 0 (representing worst functional status) to 100 (representing best functional status). B, The horizontal line in the boxes represents the median, and the bottom and top of the boxes the 25th and 75th percentiles, respectively. I bars represent the maximum and minimum values. Unadjusted P values reflect differences between heart failure subtypes (HFimpEF serves as the reference group). There was a statistically significant difference between KCCQ‐12 Overall Summary Scores for patients with HFrEF compared with those with HFimpEF and HFpEF; the average scores for HFimpEF and HFpEF are similar. HFimpEF indicates heart failure with improved ejection fraction; HFpEF, heart failure preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; and KCCQ, Kansas City Cardiomyopathy Questionnaire.

Figure 3. Multivariable analysis of KCCQ‐12 Overall Summary Score.

This figure demonstrates a multivariable analysis of KCCQ‐12 Overall Summary Scores across LVEF phenotype (HFimpEF, HFrEF and HFpEF/HFmrEF) after adjusting for relevant sociodemographic characteristics, medical comorbidities, and cardiac medications (model 3). HFimpEF was used as a reference group. AMI indicates acute myocardial infarction; ARNI, angiotensin receptor–neprilysin inhibitor; COPD, chronic obstructive pulmonary disease; HFimpEF, heart failure with improved ejection fraction; HFmrEF, heart failure with midrange ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; MRA, mineralocorticoid receptor antagonist; RAASi, renin–angiotensin–aldosterone system; and SGLT2i, sodium–glucose cotransporter‐2 inhibitor.

The differences in the KCCQ‐12 CSS between groups were overall similar. In the unadjusted analysis, patients with HFimpEF had on average KCCQ‐12 OSS scores 5.2 points higher compared with patients with HFrEF (95% CI, −8.2 to −2.3; P=0.001), whereas there was no difference between HFimpEF and HFmrEF/HFpEF (−1.1 [95% CI, −3.7 to 1.6]; P=0.430).

We performed a sensitivity analysis further stratifying patients with HFmrEF and HFpEF. Median KCCQ‐12 OSS scores were 87.4 and 83.3 for HFmrEF and HFpEF subgroups, respectively. Using HFmrEF with LVEF between 40% and 50% as a reference group, there was no statistically significant difference between HFimpEF and HFmrEF (−3.2 [95% CI, −7.5 to 1.2]; P=0.157) and HFpEF (−1.6 [95% CI, −5.2 to 2.1]; P=0.402) after adjustment for the same covariates. We also evaluated patients with HFimpEF with LVEF normalization to >50%; there was no difference across race and ethnicity, medical comorbidities, or medications with the exception of a loop diuretic prescription. In comparison with other HF subgroups, HFimpEF patients with LVEF >50% demonstrated similar KCCQ‐12 OSS scores (0.27 [95% CI, −4.1 to 4.6]; P=0.903).

Association Between KCCQ‐12 Scores and Patient Characteristics Among Patients With Improved EF

Table 2 shows the association between patient characteristics and KCCQ‐12 scores among patients with HFimpEF. Men demonstrated higher KCCQ‐12 OSS scores (KCCQ‐12 OSS point difference, 5.9 [95% CI, 1.5–10.2]; P=0.000), whereas non‐Hispanic Black patients (KCCQ‐12 OSS point difference, −10.5 [95% CI, −20.7 to −0.2]; P=0.045), patients with a history of acute myocardial infarction (KCCQ‐12 OSS point differenc, −4.7 [95% CI, −9.5 to 0]; P=0.051) and dementia (KCCQ‐12 OSS point difference, −12.4 [95% CI, −23.1 to −1.5]; P=0.026) had poorer health status in adjusted models. Treatment with evidence‐based β blocker therapy demonstrated higher KCCQ‐12 OSS scores on average compared with those not on β blockers (KCCQ‐12 CSS point difference, 4.9 [95% CI, 0.4–9.5]; P=0.031). Loop diuretic use was associated with poorer health status (KCCQ‐12 CSS point difference, −9.1 [95% CI, −13.8 to −4.4]; P=0.000). Patient characteristics only modestly captured the variation in KCCQ‐12 OSS with an R ² of 0.20.

Table 2.

Association Between Patient Characteristics and KCCQ‐12 OSS Among Patients With HFimpEF

	Adjusted for sociodemographics (R 2=0.09), point difference (95% CI); P value	Adjusted for socio‐demographics+medical comorbidities (R 2=0.16), point difference (95% CI); P value	Adjusted for sociodemographics+medical comorbidities+medications (R 2=0.20), point difference (95% CI); P value
R 2	0.09	0.16	0.20
Age, per 10 y	−0.4 (−0.46 to −0.3); 0.000	−0.3 (−0.5 to −0.2); 0.000	−0.2 (−0.4 to −0.1); 0.000
Male sex	4.5 (0.1 to 8.8); 0.043	6.1 (1.8 to 10.3); 0.006	5.9 (1.5 to 10.2); 0.008
Race and ethnicity
Non‐Hispanic Asian	−0.5 (−6.3 to 5.3); 0.861	2.0 (−3.9 to 8.0); 0.503	1.9 (−4.0 to 7.9); 0.524
Non‐Hispanic Black	−13.5 (−23.5 to −3.5); 0.009	−8.1 (−18.2 to 2.0); 0.116	−10.5 (−20.7 to −0.2); 0.045
Native American	16.8 (−16.8 to 49.7); 0.314	20.3 (−12.1 to 52.7); 0.219	29.3 (−16.6 to 75.1); 0.210
Other	−4.3 (−12.5 to 3.8); 0.297	−3.3 (−11.5 to 4.8); 0.421	−1.9 (−10.2 to 6.4); 0.650
Pacific Islander	−15.2 (−28.5 to −2.0); 0.024	−11.9 (−25.1 to 1.2); 0.075	−9.5 (−22.6 to 3.7); 0.158
Unknown	2.8 (−31.9 to 37.5); 0.874	0.8 (−33.5 to 35.1); 0.965	−1.2 (−35.4 to 33.0); 0.946
Non‐Hispanic White (reference)		…	…
Ethnicity
Hispanic/Latino	−7.1 (−15.9 to 1.7); 0.114	−3.9 (−12.7 to 4.8); 0.381	−4.1 (−13.1 to 4.9); 0.374
Non‐Hispanic	…	…	…
Unknown	−1.7 (−36.5 to 33.1); 0.923	−0.3 (−34.6 to 34.0); 0.987	0.9 (−33.3 to 35.1); 0.959
Medical comorbidities
Acute myocardial infarction	…	−5.7 (−10.5 to −1.1); 0.017	−4.7 (−9.5 to 0); 0.051
Peripheral vascular disease	…	−0.8 (−5.1 to 3.5); 0.702	−0.4 (−4.7 to 4.0); 0.874
Cerebrovascular disease	…	−2.9 (−8.9 to 3.0); 0.333	−2.2 (−8.2 to 3.8); 0.473
Diabetes	…	−3.0 (−7.7 to 1.7); 0.214	−2.6 (−7.4 to −2.1); 0.272
Renal disease	…	−4.9 (−9.6 to −0.2); 0.042	−2.6 (−7.6 to 2.4); 0.316
Dementia		−12.4 (−23.2 to −1.7); 0.024	−12.3 (−23.1 to −1.5); 0.026
COPD		−3.2 (−8.3 to 1.9); 0.208	−2.5 (−7.7 to 2.7); 0.353
Liver disease		−4.6 (−10.6 to 1.5); 0.138	−2.8 (−8.9 to 3.3); 0.369
Cancer		2.2 (−3.3 to 7.8); 0.428	1.9 (−3.8 to 7.6); 0.505
Medications
β blocker	…	…	4.9 (0.4–9.5); 0.034
RAASi			−0.7 (−7.7 to 6.3); 0.852
ARNI	…	…	2.0 (−2.8 to 6.7); 0.42
MRA	…	…	0.4 (−4.4 to 5.3) 0.856
SGLT2i	…	…	1.2 (−3.4 to 5.7); 0.616
Loop diuretics	…	…	−9.1 (−13.8 to −4.4); 0.000

Multivariable analyses adjusted for sociodemographic characteristics, comorbidities and medications among patients with HFimpEF. Relevant covariates explained only 20% of variation in KCCQ‐12 score. R ² reflects the fully adjusted Model 3.

ARNI indicates angiotensin receptor–neprilysin inhibitor; COPD, chronic obstructive pulmonary disease; HFimpEF, heart failure with improved ejection fraction; KCCQ, Kansas City Cardiomyopathy Questionnaire; MRA, mineralocorticoid receptor antagonist; OSS, Overall Summary Score; RAASi, renin‐angiotensin–aldosterone system; and SGLT2i, sodium–glucose cotransporter‐2 inhibitor.

Tables S1 and S2 show analyses of KCCQ‐12 CSS scores; patients with HFimpEF reported similar limitations to patients with HFpEF, and both groups reported less severe symptoms compared with patients with HFrEF.

Discussion

Based on the validated KCCQ‐12 tool, we compared patient‐reported health status across patients with HFimpEF, HFrEF, and HFmrEF/HFpEF in a contemporary HF cohort. Despite improvement in EF, 42% of patients with HFimpEF still had moderately impaired health status with KCCQ‐12 OSS <75. Patients with HFimpEF reported higher patient‐reported health status than patients with HFrEF, but similar health status as patients with HFmrEF/HFpEF. We found that patient characteristics explained <9% of the variation in health status among patients with HFimpEF. Even after adjustment for multiple medical comorbidities, sociodemographic characteristics, and medical therapies, racial and ethnic/sex disparities in functional status persist among patients with HFimpEF.

Our study highlights this mismatch in patient‐reported outcomes and LVEF improvement in the HFimpEF population. Despite recovery of EF, patients with HFimpEF continue to have impaired health status across a spectrum of functional domains and perform similar to patients with HFpEF and HFmrEF. Although HF management in practice depends on the stratification of patients according to LVEF, LVEF does not capture the complete underlying disease processes affecting patients with HFimpEF. Our study offers a functional correlate that supports heterogeneity in the biochemical and proteomic processes underpinning myocardial remodeling in HFimpEF ¹⁵ and supports differences between patients in “myocardial remission” ¹⁶ with durable recovery of LVEF who still have significant symptoms and who may remain at risk for recurrent HF events despite improved LVEF. Consistent with our results, significant discrepancies in LVEF improvement and functional status have been seen in other subgroups of HF patients, including those with left ventricular assist devices, ¹⁶ cardiac resynchronization therapy, ¹⁷ and early revascularization in acute coronary syndromes. ¹⁸ , ¹⁹ , ²⁰ , ²¹ Factors such as timing of LVEF improvement after initial HF diagnosis, differences in HF pathogenesis and magnitude of LVEF improvement impact accurate HFimpEF prognostication and outcomes. ⁸ Our study solidifies that the improvement in LVEF does not correlate with improvement in the HF syndrome that is persistent among patients with HFimpEF.

There is a compelling need to improve the health status and functional outcomes in patients with HFimpEF. Although studies have supported the continuation of guideline‐directed medical therapy in patients with normalized LVEF, little is known about the nuances of therapeutic strategies for this subset of patients. In a prespecified analysis of the DELIVER (Dapagliflozin Evaluation to Improve the Lives of Patients With Preserved Ejection Fraction Heart Failure) trial, ²² patients randomized to the dapagliflozin arm did experience an increase in KCCQ‐12 summary scores regardless of HFimpEF status, although the degree of change in KCCQ‐12 was small. Similarly, more recent trials examining the effects of glucagon‐like receptor 1 antagonists have demonstrated remarkable benefit in quality‐of‐life measures among patients with HFpEF, suggesting that there is substantial need for improving HF symptoms regardless of LVEF. Alternative and patient‐centered approaches, such as cardiac rehabilitation, have been explored in other subcategories of HF ²³ and may be applicable to patients with HFimpEF. The additive impact of comorbidities on KCCQ‐12 score itself has been demonstrated previously, with a higher burden of comorbid medical conditions such as atrial fibrillation and chronic obstructive pulmonary disease resulting in lower KCCQ‐12 overall and individual domain scores in both HFrEF and HFpEF. ²⁴ Heightened attention on cardiovascular–kidney–metabolic health may offer insight into the complex interplay of these comorbidities in HFimpEF. Our findings highlight future directions in the management of HFimpEF.

Our analysis emphasizes the importance of quality‐of‐life measures in HFimpEF in the deep appreciation for the trajectory of patients with HFimpEF. Prior studies ²⁵ have demonstrated similar outcomes, noting a slight advantage for HFimpEF subgroups over HFrEF, but these studies have featured small sample sizes and heterogeneity in the quality measurement tool applied and cutoff values defining improved LVEF. Our relatively larger study population (considering existing patient‐reported outcome studies) allowed for the adjustment for a wide array of covariates that have well‐established intersections with HF, lending strength to the generalization of these results. Inclusion of medical therapies such as sodium–glucose cotransporter‐2 inhibitors and guideline‐directed definitions of HFimpEF and other subcategories reflects a contemporary assessment of the entity of HFimpEF. PRO‐HF, a randomized clinical trial spearheaded by our group, found that collection of KCCQ‐12 markedly improved the accuracy of clinician assessments of patients with HF in the ambulatory setting. ¹² , ²⁶ Our findings further support the application of KCCQ‐12 as a tool for assessment of functional status, especially among patients HFimpEF, where improved LVEF does not capture the full clinical HF syndrome. ²⁷ , ²⁸ , ²⁹ Optimal cutoff values to validate meaningful change in KCCQ‐12 have been shown in prior analyses that are consistent with our findings. ²⁸ , ³⁰

There are several limitations in our study. First, the HF cohort from our academic center may not be generalizable to other settings. The KCCQ‐12 tool was initially used as part of the PRO‐HF trial, where patients were randomized at a patient level to either measurement of patient‐reported outcomes or standard of care. Although our sample of patients who were exposed to the collection of KCCQ‐12 represents a small subset of the total population of patients with HF at this institution, (1) there were no significant differences in baseline characteristics between patients randomized to KCCQ‐12 intervention arm and those who received usual care (Table S3) and (2) KCCQ‐12 became integrated into the standard HF clinic flow. Similarly to the study population included in the PRO‐HF trial, ³¹ our study population reflects all patients being seen in a cardiology clinic with a diagnosis of HF to optimize generalizability. As a result, the baseline KCCQ‐12 scores of patients included in this cohort are higher and reflective of all comers with HF to the cardiology clinic than other cohort analyses of patient‐reported outcomes in HF. ²⁰ , ³¹ These studies demonstrate lower KCCQ‐12 scores across patients with evidence of clinical HF based on the judgment of the clinician, likely reflecting a sicker population of patients with HF. Second, adjunctive biomarker and echocardiographic data were not included in this study, which are foundational components of the evaluation of HF phenotypes and are relevant future directions for this work. Third, granular data on medication use and adherence were not collected. Although our initial cohort was based on the standard definition of HFimpEF as per the 2022 American Heart Association/American College of Cardiology Heart Failure Guidelines, where HFimpEF is defined as an improvement in LVEF from <40% to >40%, we recognize that this framework has some inherent limitations (eg, by this definition, patients with an increase in their LVEF from 38% to 41% could be included but arguably have not truly “improved” their LVEF). We found no significant difference in our results after stratification by patients who improved their LVEF to >50%.

In a contemporary real‐world cohort, a substantial proportion of patients with HFimpEF continue to have substantially impaired HF‐related health status. While their health status is higher than those with reduced EF, it remains comparable with patients with preserved or mildly reduced EF. These results support existing data that improvement in EF does not represent a cure to the HF phenotype. Further research is needed to better understand impaired health status and to identify interventions to improve health status among patients with HFimpEF. Finally, clinicians should continue to carefully assess health status among patients after their EF improves.

Sources of Funding

Funding to support the publication of this work has been provided by the American Heart Association and the National Institutes of Health awarded to Dr Sandhu.

Disclosures

Dr Sandhu has received research funding from the American Heart Association, National Institutes of Health, Astra Zeneca, Bayer Pharmaceuticals, Novo Nordisk, and Novartis; and for consulting from Reprieve Cardiovascular and Cleerly. Dr Ambrosy has received relevant research support through grants to his institution from the National Heart, Lung, and Blood Institute (K23HL150159); the American Heart Association (Second Century Early Faculty Independence Award); the Permanente Medical Group; Northern California Community Benefits Programs; Garfield Memorial Fund; Abbott Laboratories; Amarin Pharma, Inc.; Bayer; Cordio Medical; Edwards Lifesciences LLC; Esperion Therapeutics, Inc.; Merck; and Novartis Dr. DeJong’s spouse is employed by and owns stock in iRhythm Technologies.

Supporting information

Tables S1–S3

Figure S1