Specialty-Related Differences in the Acute-Phase Treatment and Prognosis in Patients With Acute Heart Failure　― Insights From REALITY-AHF ―

Abstract

Background: The aim of this study was to assess specialty-related differences in the treatment for patients with acute heart failure (AHF) in the acute phase and subsequent prognostic differences.

Methods and Results: We analyzed hospitalizations for AHF in REALITY-AHF, a multicenter prospective registry focused on very early presentation and treatment in patients with AHF. All patients were classified according to the medical specialty of the physicians responsible for contributed most to decisions regarding the initial diagnosis and treatment after the emergency department (ED) arrival. Patients initially managed by emergency physicians (n=614) or cardiologists (n=911) were analyzed. After propensity-score matching, vasodilators were used less often by emergency physicians than by cardiologists at 90 min after ED arrival (29.8% vs. 46.1%, P<0.001); this difference was also observed at 6, 24, and 48 h. Cardiologists administered furosemide earlier than emergency physicians (67 vs. 102 min, P<0.001). However, the use of inotropes, noninvasive ventilation, and endotracheal intubation were similar between groups. In-hospital mortality did not differ between patients managed by emergency physicians and those managed by cardiologists (4.1% vs. 3.8%, odds ratio 1.12; 95% confidence interval 0.58–2.14).

Conclusions: Despite differences in initial management, no prognostic difference was observed between emergency physicians and cardiologists who performed the initial management of patients with AHF.

Hospitalization of patients with acute heart failure (AHF) is a growing and major public health issue associated with high morbidity, mortality, and rehospitalization rates.^1–4 Approximately 80% of patients with AHF treated in the emergency department (ED) are ultimately admitted to the hospital.³ Therefore, the ED serves as the principal portal of entry for hospitalized patients with AHF and functions as the initial point at which definitive healthcare contact and primary stabilization occur. Rapid and accurate diagnosis of AHF is vital, and delayed therapy leads to increased mortality,^5,6 which suggests the importance of early management for AHF in the ED.

AHF is a syndrome in which emergency physicians, intensivists, cardiologists, internists, and hospitalists must cooperate to provide beneficial management that improves the patient’s clinical outcome. Despite the dissemination of clinical guidelines on the treatment of HF,¹ there could be specialty-related variations in clinical practice patterns for patients with AHF because the management of AHF remains largely opinion-based with little plausible evidence.^1,2,4 Additionally, previous reports have suggested that differences in the specialty-related process of care during the index hospitalization of patients with AHF are associated with differences in specialty-related outcomes.^7–10 However, most of those reports focused on management after admission, and data on acute-phase management remain sparse. Therefore, the purpose of this study was to clarify the specialty-related differences in the treatment for patients with AHF in the acute phase and subsequent prognostic differences.

Methods

Study Design and Selection of Participants

We used data from the Registry Focused on Very Early Presentation and Treatment in Emergency Department of Acute Heart Failure Syndrome (REALITY-AHF), the design of which has been published previously.⁶ Briefly, the REALITY-AHF was a prospective, multicenter, observational registry designed to evaluate the prognostic effect of treatment of AHF in the acute phase. All patients with AHF hospitalized through the ED at participating hospitals were consecutively enrolled upon initial hospital admission. Patients were eligible for inclusion in the REALITY-AHF if they met the modified Framingham criteria¹¹ and had been diagnosed with AHF within 3 h of the first evaluation by medical staffs in the ED. The exclusion criteria were: (1) age <20 years, (2) treatment with an intravenous (IV) drug prior to ED arrival, (3) previous cardiac transplantation, (4) treatment with either chronic peritoneal dialysis or hemodialysis, (5) acute myocarditis, and (6) acute coronary syndrome requiring emergency or urgent revascularization. Patients with missing data for B-type natriuretic peptide (BNP) or N-terminal-proBNP levels and patients with a BNP level <100 pg/mL or N-terminal-proBNP level <300 pg/mL at admission were also excluded. Eligibility was not contingent on the use of any particular therapeutic agent or treatment.

The participating institutes included 9 university hospitals and 11 non-university teaching hospitals. Study enrollment was performed from August 2014 to December 2015 with a follow-up of 1 year after discharge for each patient. The study protocol conformed to the Declaration of Helsinki and Japanese Ethical Guidelines for Medical and Health Research Involving Human Subjects. The study protocol was approved by the human (ethics) committee at each participating hospital; however, informed consent from individuals was not required for registry entry, and an opt-out method for participant recruitment was used because obtaining written informed consent in the ED may have delayed treatment and subsequently biased the results. Study information, which included the study objectives, inclusion and exclusion criteria, and names of participating institutes, was publicized in the University Hospital Information Network (UMIN-CTR, unique identifier: UMIN000014105) before enrollment of the first patient.

Data Collection, Definitions, and Outcomes

The data in REALITY-AHF include the patients’ demographic information, medical history, baseline clinical characteristics in the ED phase, treatment performed until 48 h after ED arrival, and clinical outcomes. The first 48 h after ED arrival were divided into 4 time windows: from ED arrival to 90 min after ED arrival, from 90 min to 6 h, from 6 to 24 h, and from 24 to 48 h. In each time window, treatment including noninvasive ventilation (NIV), endotracheal intubation (ETI), and IV drug administration were recorded. The frequency of ETI, intra-aortic balloon pumping, extracorporeal membrane oxygenation, and renal replacement therapy during hospitalization were also checked. In this registry, the time from the patient’s arrival in the ED to the first IV furosemide administration was recorded for all patients as the door-to-furosemide (D2F) time. Because a D2F <60 min is reportedly associated with a better in-hospital prognosis,⁶ the proportion of patients treated with furosemide within 60 min of arrival was also evaluated.

We classified all patients according to the medical specialty of the physician most responsible: emergency physician, cardiologist and other physicians. The physician most responsible was defined as the one who contributed most to decisions regarding the initial diagnosis and treatment in the ED. Physician specialties were assigned by each institute and reflected both the level of training of the physicians and the type of treatment provided by them.

The primary outcome of this study was all-cause in-hospital death. The secondary outcomes were length of hospital stay, 1-year all-cause death, and specialty-related differences in medical treatment until 48 h after ED arrival and during hospitalization.

Statistical Analysis

Continuous variables are expressed as mean±standard deviation or median with lower and upper quartiles as appropriate. Continuous variables were compared by Student’s t-test or the Mann-Whitney U test as appropriate. Categorical variables are expressed as number (percentage) and were compared by the chi-square test or Fisher’s exact test. The Get With The Guidelines-Heart Failure (GWTG-HF) risk score was calculated for each patient as previously described to evaluate the risk of in-hospital death of patients with AHF.¹² The GWTG-HF risk score has been successfully applied in Japanese patients with AHF with good discrimination and calibration.¹³

We performed propensity-score matching to eliminate bias caused by differences in patients’ backgrounds at baseline. The propensity score for each patient was generated using a logistic model constructed with the following variables: age; sex; arrival by ambulance; time of symptom onset; systolic blood pressure (BP); diastolic BP; heart rate; left ventricular ejection fraction; baseline ECG findings; history of hypertension, diabetes, HF, coronary artery disease, and chronic obstructive pulmonary disease; presence of jugular venous distension; rales; pulmonary edema and orthopnea in the ED; prescription of loop diuretics, angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, β-blockers, and aldosterone blockers at baseline; white blood cell count; hemoglobin; serum sodium; serum creatinine; blood urea nitrogen; and BNP. The Hosmer-Lemeshow goodness-of-fit test was used to check the calibration of the propensity-score model. Propensity-score matching was performed for patients who underwent management by an emergency physician or a cardiologist. These patients underwent one-to-one caliper matching using a caliper width of 20% of the standard deviation of the logit of the calculated propensity score.¹⁴ Standardized mean differences were calculated for all baseline variables to evaluate the balance of the matching, and a difference <0.1 was considered negligible.¹⁵ Kaplan-Meier survival analysis with a log-rank test was used to compare the all-cause death rate between groups. Odds ratio (OR) for dichotomous outcomes and hazard ratio (HR) for survival were calculated with 95% confidence interval (CI). For all analyses, patients managed by a cardiologist were the reference group. To account for intra-institutional correlations among patients, a generalized estimating equation (GEE) model was used before and after propensity-score matching for the outcome of in-hospital death. The GEE model was constructed assuming an exchangeable correlation structure where subjects’ outcomes from the same center had equal correlation with each other but were uncorrelated with outcomes from subjects in other centers. Additionally, we used robust sandwich variance.

All statistical analyses were performed using R version 3.1.2 (R Foundation for Statistical Computing, Vienna, Austria; https://www.r-project.org), and a two-tailed P-value <0.05 was considered statistically significant.

Results

Study Population and Patients’ Characteristics

During the study period, 1,762 patients met the inclusion criteria and were registered in REALITY-AHF. Of these, 80 who met the exclusion criteria were excluded, so 1,682 patients (850 male; mean age, 77 years) remained and comprised the REALITY-AHF cohort. Among these patients, 911 (54.2%) were managed initially by cardiologists as the physician most responsible in the ED, 614 (36.5%) were managed by emergency physicians, and 157 (9.3%) were managed by other physicians. After we had excluded 157 patients managed by other physicians because of the small sample size, 1,525 patients managed by cardiologists or emergency physicians were analyzed.

The patients’ baseline characteristics are listed in Table 1. Patients cared for by emergency physicians (emergency physician group) were likely to arrive by ambulance and show acute onset of symptoms, and they had a higher systolic BP and heart rate, and lower prevalence of atrial fibrillation, and higher prevalence of hypertension than patients cared for by cardiologists (cardiologist group). Signs of congestion were less frequently present and prescriptions of loop diuretics, angiotensin II receptor blockers, and β-blockers were less prevalent in the emergency physician group than in the cardiologist group. Of the laboratory data, the white blood cell count, alanine transaminase level, and glucose level were significantly higher and the sodium level significantly lower in the emergency physician group. In total, 932 patients were matched based on propensity scores (Table 1). The standardized mean differences were <0.1 for all variables (Supplementary Figure), and the Hosmer-Lemeshow test showed good calibration of the propensity score (P=0.38).

Table 1. AHF Patients’ Characteristics

Variables	Before match			After match
	Emergency physicians (n=614)	Cardiologists (n=911)	P value	Emergency physicians (n=466)	Cardiologists (n=466)
Age, mean (SD), years	77 (12)	77 (13)	0.99	78 (12)	78 (12)
Male sex, n (%)	346 (56.4)	504 (55.3)	0.69	265 (56.9)	269 (57.7)
Arrival by ambulance, n (%)	442 (72.0)	466 (51.2)	<0.001	306 (65.7)	304 (65.2)
Systolic BP, mean (SD), mmHg	151 (39)	147 (35)	<0.001	152 (38)	149 (36)
Diastolic BP, mean (SD), mmHg	84 (27)	83 (24)	0.02	85 (27)	83 (24)
Heart rate, mean (SD), beats/min	101 (27)	96 (30)	0.27	98 (26)	99 (30)
Symptom onset time, n (%)			<0.001
≤6 h	195 (31.8)	172 (18.9)		130 (27.9)	127 (27.3)
6 h–2 days	130 (21.2)	193 (21.2)		101 (21.7)	114 (24.5)
>2 days	288 (47.0)	545 (59.9)		235 (50.4)	225 (48.3)
ECG rhythm, n (%)			<0.001
Sinus	372 (60.6)	441 (48.7)		269 (57.7)	257 (55.2)
AF	180 (29.3)	366 (40.4)		148 (31.8)	161 (34.5)
Other	62 (10.1)	99 (10.9)		49 (10.5)	48 (10.3)
LVEF in ED, n (%)			0.54
<35%	232 (39.1)	314 (37.3)		176 (37.8)	189 (40.6)
35–50%	176 (29.7)	242 (28.7)		144 (30.9)	137 (29.4)
>50%	185 (31.2)	286 (34.0)		146 (31.3)	140 (30.0)
Medical history, n (%)
History of HF	300 (48.9)	481 (52.8)	0.14	234 (50.2)	243 (52.1)
Hypertension	433 (70.5)	590 (64.9)	0.02	317 (68.0)	315 (67.6)
Diabetes mellitus	240 (39.1)	321 (35.3)	0.13	175 (37.6)	161 (34.5)
Coronary artery disease	181 (29.5)	278 (30.6)	0.65	144 (30.9)	135 (29.0)
COPD	55 (9.0)	81 (8.9)	0.99	45 (9.7)	46 (9.9)
Physical findings in ED, n (%)
Peripheral edema	393 (64.1)	639 (70.3)	0.01	308 (66.1)	310 (66.5)
JVD	286 (47.1)	559 (62.1)	<0.001	248 (53.2)	256 (54.9)
Orthopnea	364 (59.5)	519 (57.2)	0.38	294 (63.1)	294 (63.1)
Pulmonary edema	441 (71.8)	665 (73.2)	0.57	343 (73.6)	332 (71.2)
Rales	405 (66.1)	608 (67.0)	0.72	308 (66.1)	322 (69.1)
Medications on admission, n (%)
Loop diuretics	279 (45.5)	494 (54.8)	<0.001	230 (49.4)	239 (51.3)
ACEIs	98 (16.0)	168 (18.5)	0.20	80 (17.2)	79 (17.0)
ARBs	168 (27.4)	294 (32.5)	0.03	130 (27.9)	134 (28.8)
β-blockers	245 (40.0)	414 (45.9)	0.02	190 (40.8)	208 (44.6)
Aldosterone blockers	125 (20.4)	210 (23.1)	0.20	101 (21.7)	107 (23.0)
Laboratory data
WBC, median (IQR), /μL	8,300 (6,200–11,200)	7,120 (5,600–9,200)	<0.001	8,000 (6,000–10,575)	7,900 (6,000–10,300)
Hb, median (IQR), g/dL	11.8 (10.3–13.3)	11.5 (10.1–13.2)	0.12	11.7 (10.3–13.3)	11.8 (10.0–13.4)
AST, median (IQR), IU/L	32 (23–49)	30 (22–44)	0.09	32 (23–47)	32 (23–47)
ALT, median (IQR), IU/L	23 (14–39)	21 (14–34)	0.05	22 (14–37)	20 (13–36)
Creatinine, median (IQR), mg/dL	1.07 (0.85–1.52)	1.14 (0.83–1.65)	0.21	1.08 (0.85–1.54)	1.13 (0.83–1.64)
BUN, median (IQR), mg/dL	25 (18–34)	25 (18–35)	0.44	25 (18–34)	25 (18–35)
Glucose, median (IQR), mg/dL	151 (117–226)	135 (111–186)	<0.001	145 (116–211)	141 (112–203)
Sodium, mean (SD), mEq/L	138 (5)	139 (4)	0.004	139 (5)	139 (5)
CRP, median (IQR), mg/dL	0.71 (0.21–2.06)	0.58 (0.19–2.13)	0.26	0.68 (0.22–2.04)	0.69 (0.21–2.58)
BNP, median (IQR), pg/mL	778 (482–1,394)	719 (431–1,330)	0.08	767 (474–1,397)	807 (470–1,443)
GWTG-HF risk score, mean (SD)	38 (8)	38 (8)	0.88	38 (8)	38 (8)

Data are presented as number (percentage) of patients, mean±SD, or median with lower and upper quartiles. ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; AHF, acute heart failure; ALT, alanine transaminase; ARB, angiotensin II receptor blocker; AST, aspartate transaminase; BNP, B-type natriuretic peptide; BP, blood pressure; BUN, blood urea nitrogen; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; ECG, electrocardiogram; ED, emergency department; GWTG-HF, Get With The Guidelines-Heart Failure; Hb, hemoglobin; IQR, interquartile range; JVD, jugular vein distention; LVEF, left ventricular ejection fraction; SD, standard deviation; WBC, white blood cell.

Differences in Treatment for AHF

The selection and timing of treatment for AHF are shown in Table 2. Although the emergency physician group was less likely to use vasodilators and more likely to use NIV and ETI in the first 48 h than was the cardiologist group before matching, these differences did not retain significance after propensity-score matching except for vasodilator use in the first 48 h. In the matched cohort, the rate of vasodilator use was 29.8% for emergency physicians and 46.1% for cardiologists within 90 min of ED arrival. The cumulative usage rate of vasodilators increased to 47.9% in the emergency physician group and 61.7% in the cardiologist group at 6 h after ED arrival, to 52.6% and 64.8% at 24 h after ED arrival, and to 53.6% and 66.5% at 48 h after ED arrival, respectively. Although the usage rates by both types of physicians revealed a generally constant trend, vasodilators were adopted more often by cardiologists than by emergency physicians at all time points (P<0.001).

Table 2. Treatment of AHF After ED Arrival

Treatment	Before match			After match
	Emergency physicians (n=614)	Cardiologists (n=911)	P value	Emergency physicians (n=466)	Cardiologists (n=466)	P value
Vasodilators
In 90 min, n (%)	191 (31.1)	362 (39.7)	0.001	139 (29.8)	215 (46.1)	<0.001
In 6 h, n (%)	313 (51.0)	556 (61.0)	<0.001	223 (47.9)	287 (61.6)	<0.001
In 24 h, n (%)	341 (55.5)	583 (64.0)	0.001	245 (52.6)	302 (64.8)	<0.001
In 48 h, n (%)	349 (56.8)	594 (65.2)	0.001	250 (53.6)	310 (66.5)	<0.001
Inotropes
In 90 min, n (%)	35 (5.7)	78 (8.6)	0.04	27 (5.8)	43 (9.2)	0.06
In 6 h, n (%)	65 (10.6)	112 (12.3)	0.33	49 (10.5)	62 (13.3)	0.23
In 24 h, n (%)	82 (13.4)	128 (14.1)	0.71	59 (12.7)	68 (14.6)	0.44
In 48 h, n (%)	91 (14.8)	137 (15.1)	0.94	66 (14.2)	73 (15.7)	0.57
NIV
In 90 min, n (%)	115 (18.7)	117 (12.8)	0.002	76 (16.3)	85 (18.2)	0.49
In 6 h, n (%)	132 (21.5)	129 (14.2)	<0.001	88 (18.9)	90 (19.3)	0.93
In 24 h, n (%)	135 (22.0)	136 (14.9)	<0.001	90 (19.3)	95 (20.4)	0.81
In 48 h, n (%)	136 (22.1)	139 (15.3)	0.001	91 (19.5)	97 (20.9)	0.61
ETI
In 90 min, n (%)	27 (4.4)	23 (2.5)	0.06	15 (3.2)	15 (3.2)	>0.99
In 6 h, n (%)	30 (4.9)	24 (2.6)	0.02	17 (3.6)	16 (3.4)	>0.99
In 24 h, n (%)	35 (5.7)	26 (2.9)	0.01	20 (4.3)	17 (3.6)	0.74
In 48 h, n (%)	39 (6.4)	27 (3.0)	0.002	23 (4.9)	17 (3.7)	0.42

Data are presented as number (percentage) of patients. AHF, acute heart failure; ED, emergency department; ETI, endotracheal intubation; NIV, noninvasive ventilation.

With regard to treatment with diuretics, the D2F time was shorter in the cardiologist group than in the emergency physician group (104 min vs. 80 min for the emergency physician vs. cardiologist group, respectively; P<0.001). This difference was retained even after propensity-score matching. The rate of patients with a D2F time of <60 min was numerically but not statistically higher in the cardiologist group than in the emergency physician group before propensity-score matching (35.6% vs. 39.6% for the emergency physician vs. cardiologist group, respectively; P=0.162). However, in the matched cohort, significantly more patients were treated with furosemide within 60 min of ED arrival in the cardiologist group than in the emergency physician group (36.2% vs. 45.7%, respectively; P=0.009).

The usage rates of ETI (7.3% vs. 6.0%; P=0.51), intra-aortic balloon pumping (1.5% vs. 2.1%; P=0.62), extracorporeal membrane oxygenation (0.4% vs. 0.2%; P=0.38), and renal replacement therapy (5.8% vs. 6.0%; P=0.99) during hospitalization were comparable between groups after propensity-score matching (Table 3).

Table 3. Treatment of AHF During Hospitalization

Treatment/procedure	Before match			After match
	Emergency physicians (n=614)	Cardiologists (n=911)	P value	Emergency physicians (n=466)	Cardiologists (n=466)	P value
ETI, n (%)	54 (8.8)	48 (5.3)	0.009	34 (7.3)	28 (6.0)	0.51
IABP, n (%)	13 (2.1)	11 (1.2)	0.24	7 (1.5)	10 (2.1)	0.62
ECMO, n (%)	4 (0.7)	1 (0.1)	0.18	2 (0.4)	1 (0.2)	0.38
RRT, n (%)	25 (4.1)	29 (3.2)	0.44	20 (5.8)	28 (6.0)	0.99

AHF, acute heart failure; ECMO, extracorporeal membrane oxygenation; ETI, endotracheal intubation; IABP, intra-aortic balloon pumping; RRT, renal replacement therapy. Data are presented as number (percentage) of patients.

Clinical Outcomes

During the index hospitalization, 30 patients (4.9%) initially managed by emergency physicians and 39 patients (4.3%) managed by cardiologists died; no significant difference was found before propensity-score matching (Table 4). Likewise, all-cause in-hospital mortality was not significantly different between groups after propensity-score matching (4.3% vs. 3.9%, OR 1.12; 95% CI 0.58–2.14, P=0.741). We obtained consistent results in the GEE model, in which in-hospital deaths were similar in the 2 groups both before (OR 1.21; 95% CI 0.78–1.87, P=0.400) and after (OR 1.22; 95% CI: 0.54–2.66, P=0.626) propensity-score matching. However, hospital stay was significantly shorter in the emergency physician group than in the cardiologist group after propensity-score matching (14 days vs. 18 days, P<0.001).

Table 4. AHF Patients’ Outcomes During Hospitalization

Variables	Before match				After match
	Emergency physicians (n=614)	Cardiologists (n=911)	P value	OR (95% CI)	Emergency physicians (n=466)	Cardiologists (n=466)	P value	OR (95% CI)
In-hospital death, n (%)	30 (4.9)	39 (4.3)	0.58	1.15 (0.71–1.87)	20 (4.3)	18 (3.9)	0.74	1.12 (0.58–2.14)
Length of hospital stay, (IQR), days	15 [9–25]	17 [12–25]	<0.001		14 [9–25]	18 [12–26]	<0.001

Reference group is cardiologists’ patients. Data are presented as number (percentage) of patients or median with lower and upper quartiles. AHF, acute heart failure; CI, confidence interval; IQR, interquartile range; OR, odds ratio.

Kaplan-Meier curves for all-cause death are shown in the Figure. Both before and after matching, no significant group difference in 1-year mortality was found (log-rank: P=0.24 and P=0.77 before and after propensity-score matching, respectively; HR 0.96, 95% CI 0.73–1.27, P=0.775 after propensity-score match).

Figure.

Kaplan-Meier survival analysis according to specialties within (A) the total cohort of 1,525 patients and (B) the matched cohort of 936 patients.

Discussion

In the present study, the outcome of patients with AHF remained unchanged regardless of whether the initial management after ED arrival was performed by an emergency physician or a cardiologist. Vasodilators were administered more frequently and earlier and furosemide was administered earlier by cardiologists, but no significant difference was observed in the timing and frequency of use of the other interventions, including inotropes, NIV, and ETI. This is the first study to demonstrate specialty-related variations in the initial AHF management strategies and prognosis of AHF.

The results of this study provide novel and clinically important information regarding the future outcomes of patients with AHF. In our cohort, although there were some differences in the treatment of AHF in the acute phase, no specialty-related differences in short- or long-term mortality were observed. In previous reports of patients with HF, cardiologists adhered more closely to evidence-based care with respect to implementation of hemodynamics or coronary artery evaluation during hospitalization and prescription of cardioprotective drugs at the time of discharge.^7–10,16 As a result, cardiologist-delivered AHF care is associated with significantly lower rates of adverse outcomes compared with non-cardiologist-delivered care.^7–10 However, those reports assessed management of inpatients with AHF ≥24 h after ED presentation during the index hospitalization. We hypothesized that initial management provided by cardiologists may lead to a better prognosis because previous studies have suggested that the prognosis of patients with acute myocardial infarction may improve if management during the first 24 h is provided by cardiologists.¹⁷ However, our study results implied that initial AHF management performed by different specialties yields comparable prognostic outcomes. We might be able to address this question because of the fact that all patients enrolled in REALITY-AHF were finally managed by cardiologists after admission even when the initial management was performed by non-cardiologists. Because we were interested in the outcome differences attributable to differences in initial management, this uniformity might have allowed us to achieve less biased results. Given that proper consultation reportedly improves the prognosis of AHF,⁹ our results may support the idea that it is important for cardiologists and emergency physicians to collaborate well but that it is not necessary to decide which specialty should first care for patients with AHF. In addition, this report provides important information in establishing the proper arrangement of doctors in the emergency field.

Interestingly, we found that cardiologists tended to use furosemide and vasodilators earlier than emergency physicians, which implied that earlier management provided by cardiologists instead of emergency physicians translates into a better prognosis. In addition, our previous study demonstrated the association between lower in-hospital mortality and early furosemide administration, or the D2F time. The result that there was no prognostic difference between groups seems contrary to expectation, but the findings of this study do not necessarily contradict our previous findings for the following reasons. On a continuous scale, we observed a 35-min group difference in the D2F time; however, the proportion of patients treated within 60 min differed only by 9.5% between groups. Given the reduction in mortality expected by early treatment (i.e., D2F time <60 min) and low in-hospital mortality in patients with AHF, our analysis was obviously underpowered to detect a difference in in-hospital mortality as a consequence of a group difference in the D2F time. Furthermore, recent reports propose that early intervention with an intravenous vasodilator does not always lead to a reduction of cardiovascular mortality,^18,19 which supports our finding that early vasodilator administration was not associated with prognosis. In addition, although a prognostic difference was not observed despite differences in the initial drug treatment, this result should not deny the importance of initial management, because the initial drug treatment might not differ enough to result in prognostic difference. Next, there was no difference in the use of NIV and ETI between emergency physicians and cardiologists at any time point in our study. Because delayed decisions regarding invasive respiratory support worsen the prognosis,²⁰ it is noteworthy that decisions regarding the use of NIV and ETI in both groups were made equally early (at ≤90 min), which is also one of the important findings that explain the lack of difference in prognosis. Additionally, we found no significant difference in the introduction rate of ETI, intra-aortic balloon pumping, extracorporeal membrane oxygenation, or renal replacement therapy during the subsequent phase of hospitalization, which suggests that AHF management by cardiologists during the acute phase does not lead to avoidance of invasive treatment.

Surprisingly, patients with AHF initially managed by cardiologists stayed in hospital longer than those managed by emergency physicians. We have no clear explanation for this finding. However, a previous study showed that cardiologists tend to use IV drugs for congestive HF more aggressively than do doctors of other specialties.¹⁰ In the present study, cardiologists used IV drugs more frequently than did emergency physicians; thus, greater use of IV drugs by cardiologists might lead to longer hospital stays because exposure to such drugs after symptoms have improved would easily prolong the treatment period. However, this hypothesis should be evaluated in future research.

Study Limitations

There are several limitations that need to be addressed. First, all treatments in hospital after initial care in the ED were performed by cardiologists. Although smooth and prompt consultation is considered important for improving clinical outcomes, especially in severe cases when the initial management was carried out by non-cardiologist physicians, we could not fully analyze the effect of the timing of consultation with cardiologists on prognosis. Second, this subanalysis was performed using a prospective observational dataset in Japan. Therefore, the generalizability of our findings might be limited because the medical care system for emergency outpatients may differ among countries. Third, although 20 hospitals, including university hospitals and non-university teaching hospitals, participated this registry in relatively equal proportions, they may not well represent Japanese hospitals that treat patients with AHF, which limits extrapolation of our study results to the whole spectrum of patients with AHF. Finally, interinstitutional differences in management quality might have confounded our study results. However, we tested this bias with a sensitivity analysis using a GEE model and observed consistent results. We believe that the consistency of our analysis results supports our conclusions.

Conclusions

Initial management of AHF in the acute phase differed to some extent between emergency physicians and cardiologists, but these treatment differences did not translate into differences in short- and long-term mortality.

Acknowledgement

None.

Funding Sources

REALITY-AHF was funded by The Cardiovascular Research Fund, Tokyo, Japan. The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Declaration of Interest

T.O. has received research grants from Ono Yakuhin, Bayer, Daiichi-Sankyo, Amgen Astellas, and Bristol-Myers outside the submitted work. Y.M. is supported by the Japan Society for the Promotion of Science Overseas Research Fellowships. Y.M. is supported by the Japan Society for the Promotion of Science Overseas Research Fellowships. Y.M. is affiliated with a department endowed by Philips Respironics, ResMed, Teijin Home Healthcare, and Fukuda Denshi, and received an honorarium from Otsuka Pharmaceutical Co., outside the submitted work. K. Yoshida received partial tuition support from Pfizer, Takeda, Bayer, and PhRMA. T. Murohara received lecture fees and unrestricted research grant for the Department of Cardiology, Nagoya University Graduate School of Medicine from Bayer, Daiichi-Sankyo, Dainippon Sumitomo, Kowa, MSD, Mitsubishi Tanabe, Boehringer Ingelheim, Novartis, Pfizer, Sanofi-Aventis, Takeda, Astellas, Otsuka, and Teijin.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-18-0724

References

• 1.   Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr, Drazner MH, et al. 2013 ACCF/AHA guideline for the management of heart failure: A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2013; 62: e147–e239.
• 2.   Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 2016; 37: 2129–2200.
• 3.   Weintraub NL, Collins SP, Pang PS, Levy PD, Anderson AS, Arslanian-Engoren C, et al. Acute heart failure syndromes: Emergency department presentation, treatment, and disposition: Current approaches and future aims: A Scientific Statement from the American Heart Association. Circulation 2010; 122: 1975–1996.
• 4.   Tsuchihashi-Makaya M, Hamaguchi S, Kinugawa S, Yokota T, Goto D, Yokoshiki H, et al. Characteristics and outcomes of hospitalized patients with heart failure and reduced vs. preserved ejection fraction: Report from the Japanese Cardiac Registry of Heart Failure in Cardiology (JCARE-CARD). Circ J 2009; 73: 1893–1900.
• 5.   Peacock WF, Emerman C, Costanzo MR, Diercks DB, Lopatin M, Fonarow GC. Early vasoactive drugs improve heart failure outcomes. Congest Heart Fail 2009; 15: 256–264.
• 6.   Matsue Y, Damman K, Voors AA, Kagiyama N, Yamaguchi T, Kuroda S, et al. Time-to-furosemide treatment and mortality in patients hospitalized with acute heart failure. J Am Coll Cardiol 2017; 69: 3042–3051.
• 7.   Jong P, Gong Y, Liu PP, Austin PC, Lee DS, Tu JV. Care and outcomes of patients newly hospitalized for heart failure in the community treated by cardiologists compared with other specialists. Circulation 2003; 108: 184–191.
• 8.   Uthamalingam S, Kandala J, Selvaraj V, Martin W, Daley M, Patvardhan E, et al. Outcomes of patients with acute decompensated heart failure managed by cardiologists versus noncardiologists. Am J Cardiol 2015; 115: 466–471.
• 9.   Philbin EF, Weil HF, Erb TA, Jenkins PL. Cardiology or primary care for heart failure in the community setting: Process of care and clinical outcomes. Chest 1999; 116: 346–354.
• 10.   Reis SE, Holubkov R, Edmundowicz D, McNamara DM, Zell KA, Detre KM, et al. Treatment of patients admitted to the hospital with congestive heart failure: Specialty-related disparities in practice patterns and outcomes. J Am Coll Cardiol 1997; 30: 733–738.
• 11.   McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural history of congestive heart failure: The Framingham study. N Engl J Med 1971; 285: 1441–1446.
• 12.   Peterson PN, Rumsfeld JS, Liang L, Albert NM, Hernandez AF, Peterson ED, et al. A validated risk score for in-hospital mortality in patients with heart failure from the American Heart Association Get with The Guidelines Program. Circ Cardiovasc Qual Outcomes 2010; 3: 25–32.
• 13.   Shiraishi Y, Kohsaka S, Abe T, Mizuno A, Goda A, Izumi Y, et al. Validation of the Get With The Guideline-Heart Failure risk score in Japanese patients and the potential improvement of its discrimination ability by the inclusion of B-type natriuretic peptide level. Am Heart J 2016; 171: 33–39.
• 14.   Austin PC. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat 2011; 10: 150–161.
• 15.   Haukoos JS, Lewis RJ. The propensity score. JAMA 2015; 314: 1637–1638.
• 16.   Rao A, Walsh J. Impact of specialist care in patients with newly diagnosed heart failure: A randomised controlled study. Int J Cardiol 2007; 115: 196–202.
• 17.   Birkhead JS, Weston C, Lowe D. Impact of specialty of admitting physician and type of hospital on care and outcome for myocardial infarction in England and Wales during 2004–5: Observational study. BMJ 2006; 332: 1306–1311.
• 18.   Packer M, O’Connor C, McMurray JJV, Wittes J, Abraham WT, Anker SD, et al. Effect of ularitide on cardiovascular mortality in acute heart failure. N Engl J Med 2017; 376: 1956–1964.
• 19.   Yan B, Peng L, Zhao X, Chung H, Li L, Zeng L, et al. Nesiritide fails to reduce the mortality of patients with acute decompensated heart failure: An updated systematic review and cumulative meta-analysis. Int J Cardiol 2014; 177: 505–509.
• 20.   Kang BJ, Koh Y, Lim CM, Huh JW, Baek S, Han M, et al. Failure of high-flow nasal cannula therapy may delay intubation and increase mortality. Intensive Care Med 2015; 41: 623–632.