
==== Front
JACC Adv
JACC Adv
JACC: Advances
2772-963X
Elsevier

S2772-963X(24)00430-7
10.1016/j.jacadv.2024.101199
101199
Original Research
Outcomes for Mechanically Ventilated Patients With Acute Myocardial Infarction Admitted to Medical vs Cardiac Intensive Care Units
Shahu Andi MD, MHS a
Namburar Sathvik MD b
Banna Soumya MD b
Harris Alyssa MPH c
Schenck Christopher MD b
Trejo-Paredes Camila MD a
Thomas Alexander MD a
Ali Tariq MD, MBA a
Carnicelli Anthony P. MD d
Barnett Christopher F. MD, MPH e
Solomon Michael A. MD, MBA f
Miller P. Elliott MD, MHS Elliott.miller@yale.edu
a∗
a Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut, USA
b Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut, USA
c Center for Advanced Analytics and Informatics, Vizient, Inc, Irving, Texas, USA
d Division of Cardiology, Department of Medicine, Medical University of South Carolina, Charleston, South Carolina, USA
e Division of Cardiology, Department of Medicine, University of California, San Francisco, California, USA
f Critical Care Medicine Department, National Institutes of Health Clinical Center and Cardiovascular Branch, National Heart, Lung, and Blood Institute of the National Institutes of Health, Bethesda, Maryland, USA
∗ Address for correspondence: Dr P. Elliott Miller, Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut 06517, USA. Elliott.miller@yale.edu
14 8 2024
9 2024
14 8 2024
3 9 10119922 3 2024
28 5 2024
1 6 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Background

Acute myocardial infarction (AMI) remains a common reason for admission to the intensive care unit (ICU). However, there is limited data comparing outcomes for patients with AMI admitted to specific ICUs.

Objectives

The purpose of this study was to assess clinical outcomes between patients with AMI requiring invasive mechanical ventilation admitted to the medical ICU (MICU) compared to cardiac (CICU).

Methods

We utilized the Vizient Clinical Data Base to identify patients with a primary diagnosis of AMI between October 2015 and December 2019 and requiring invasive mechanical ventilation. Using multivariable logistic regression, we compared clinical outcomes for patients admitted to the MICU vs CICU.

Results

We identified 12,639 patients, 25.2% (n = 3,185) of which were admitted to a MICU and 74.8% (n = 9,454) to a CICU. Patients admitted to a CICU were more likely to present with STEMI (57.0% vs 42.8%), cardiogenic shock (46.0% vs 37.4%), and require mechanical circulatory support and vasoactive medications (all, P < 0.001). Median ventilator days were 4 days in both ICUs and not statistically different after multivariable adjustment (P = 0.81). In-hospital mortality was 42.7% compared to 41.3% for MICU vs CICU admissions, respectively (P = 0.15). After multivariable adjustment, CICU admission was associated with lower in-hospital mortality (OR: 0.85, 95% CI: 0.78-0.93, P = 0.001), which persisted when stratified by cardiogenic shock, cardiac arrest, STEMI, largest hospital size (>750 beds), and teaching hospitals (all, P < 0.05).

Conclusions

Admission to the CICU, as compared to MICU, was associated with lower in-hospital mortality for patients with AMI. These findings may support optimal triage of critically ill patients with AMI.

Central Illustration

Key words

acute myocardial infarction
intensive care unit
invasive mechanical ventilation
Abbreviations and Acronyms

AAMC Association of American Medical Colleges

AMI acute myocardial infarction

CABG coronary artery bypass grafting

CICU cardiac intensive care unit

DNR do-not-resuscitate

ESRD end stage renal disease

ICD-10-CM International Classification of Diseases-10th Revision-Clinical Modification

ICU intensive care unit

IMV invasive mechanical ventilation

IPTW inverse probability treatment weighting

IRR incident rate ratio

MICU medical intensive care unit

PCI percutaneous coronary intervention
==== Body
pmcOver the last 20 years, cardiac intensive care units (CICU) have evolved to manage critically ill patients with noncardiac comorbidities and deliver a multitude of critical care therapies.1, 2, 3, 4 Although the capabilities of these specialized units have advanced, patients with primary and concomitant cardiovascular disease are still commonly admitted to noncardiac ICUs. A recently published multicenter study found that nearly half of the patients with a primary cardiac diagnosis are admitted to noncardiac ICUs.5 However, many questions remain regarding the optimal triage of patients with acute cardiovascular disease requiring ICU level care.

Acute myocardial infarction (AMI) remains a common diagnosis with considerable morbidity and mortality, as well as a frequent indication for ICU admission.6, 7, 8 Although highly variable between centers, contemporary registry data suggests that up to one-third of all admissions to the CICU are still for AMI.2 However, a large proportion of patients with AMI are still frequently admitted to noncardiac ICUs, such as the medical ICU (MICU).9 North American registry data suggested that approximately 18% of patients with AMI admitted to the CICU require invasive mechanical ventilation (IMV),10 and approximately 2 to 4% of all AMI admissions to the hospital require IMV.11,12 Healthcare systems typically have institutional-specific factors that determine whether critically ill patients with AMI are triaged to the CICU versus MICU, resulting in unique ICU admission patterns.9,13 One specific factor, the requirement of IMV is associated with poor outcomes in patients with cardiovascular disease and has been increasing in frequency in the CICU.8,10,14, 15, 16 However, there is limited data describing whether clinical outcomes for critically ill patients with AMI differ based on the type of ICU patients are admitted to.

Therefore, using data from a large, contemporary database, we aimed to assess critical care therapies and outcomes for patients with a primary diagnosis of AMI requiring IMV based on admission to a MICU or CICU.

Methods

Data source and study population

We utilized the Vizient Clinical Data Base, which includes patient-level details from 97% of United States (US) academic medical centers and their affiliates, including over 300 community hospitals.17 It includes administrative, financial, and pharmacy-related inpatient information. We queried the database for all adults aged ≥18 years admitted to continuously enrolling hospitals between October 2015 and December 2019, with a primary diagnosis of AMI who required IMV.18 Only patients that were admitted to a MICU or CICU were included in the study. More specifically, patients admitted to mixed units (medical/cardiac) were not included in our analyses. ICU type is designated by the hospital based on a list of predefined ICU types. All data obtained from Vizient were deidentified and exempt from the Yale University Institutional Review Board review. Data from the Vizient Clinical Data Base used with the permission of Vizient, Inc, all rights reserved.

Covariates of interest

Demographics included age, gender, race and ethnicity, primary payer, smoking status, and income quartile by patient home zip code. Comorbidities present on admission included coronary artery disease, prior percutaneous coronary intervention (PCI), prior coronary artery bypass grafting (CABG), prior MI, diabetes mellitus, peripheral vascular disease, congestive heart failure, dyslipidemia, hypertension, valvular disease, chronic pulmonary disease, chronic liver disease, end stage renal disease (ESRD), obesity, and dementia. Discharge diagnoses, coded as present on admission, included cardiogenic shock and cardiac arrest. Hospital characteristics included Association of American Medical Colleges (AAMC) teaching hospital status (yes/no), bed range (<350, 350-499, 500-750, >750 beds), rural vs urban location, and United States census regions (Midwest, Northeast, South, West).

The database includes time stamps, reported as the day since admission, for procedures and medications. We extracted occurrences of left and right heart catheterization/pulmonary artery catheterization, PCI, CABG, intra-aortic balloon pump, Impella, extracorporeal membrane oxygenation, heart transplantation, durable left ventricular assist device, which we defined as cardiac critical care therapies. We also identified noninvasive ventilation, renal replacement therapy, bronchoscopy, blood transfusions, and tracheostomy, which we defined as noncardiac procedures and therapies. Supplemental Table 1 lists the International Classification of Diseases-10th Revision-Clinical Modification (ICD-10-CM) codes used to identify procedures. Cost data is calculated using center-specific Medicare Cost Report cost-to-charge ratios with line-item hospital charges.19

Outcomes

The primary outcome of interest was in-hospital mortality. Secondary outcomes included discharge status (home, home with services, skilled nursing/rehab, against medical advice, hospice, expired, and other), total hospital cost, ventilator days, and hospital and ICU length of stay.

Statistical analysis

Baseline characteristics were compared between patients admitted to the MICU vs CICU. Continuous variables were described as mean ± SD or median (IQR), and categorical variables were described as frequencies and percentages. The t-test or Wilcoxon rank sum test was used to compare continuous variables and chi-squared test for categorical variables. Only patients with complete data were included in our analyses. Using multivariable logistic regression, we assessed for an association between MICU vs CICU admission and in-hospital mortality. Covariates used in the multivariable model were included to account for demographics, comorbidities, hospital characteristics, and markers of acuity. The model included demographics (age and gender), diagnoses present on admission (cardiogenic shock, cardiac arrest, AMI type, and acute renal failure), comorbidities (heart failure, ESRD, coronary artery disease, and chronic pulmonary disease), hospital characteristics (bed size, AAMC teaching status, rural vs urban location, and U.S. region), and procedures before intubation (any mechanical circulatory support, vasoactive medications, and renal replacement therapy). Poisson (expressed as an incident rate ratio [IRR]) was used to assess for the association between length of stay and ventilator days. A gamma regression model was used to evaluate adjusted total hospital costs.

To confirm the robustness of our findings, we performed several subgroup and sensitivity analyses. First, since larger hospitals and those designated as teaching hospitals are more likely to have both a MICU and CICU, we completed 2 analyses with the first including only hospitals with the largest bed size (>750 beds) and then only teaching hospitals. Next, we repeated analyses for key diagnoses present on admission, including cardiogenic shock, cardiac arrest, and STEMI. Since some patients have unique considerations or may potentially not undergo certain therapies (eg, left heart catheterization) due to life limiting comorbidities or goals of care,20 we repeated analyses in similar patients without a cancer diagnosis (which served as a falsification hypothesis) and again in similar patients without a do-not-resuscitate status (DNR) on admission.21 Finally, we repeated our analyses in hospitals with both a MICU and CICU as well as those with only a MICU or only a CICU.

In sensitivity analysis, to minimize confounding, we used inverse probability treatment weighting (IPTW) to assess the impact of ICU admission on the primary outcome of in-hospital mortality. Weighted standardized differences were assessed to ensure covariate balance (all <±0.10). All analyses were performed on STATA 16.0 (Stata Corp) with statistical significance considered at a 2-tailed P < 0.05.

Results

Patient characteristics

We identified a total of 12,639 patients who presented with AMI requiring IMV, 25.2% (n = 3,185) of which were admitted to an MICU and 74.8% (n = 9,454) were admitted to a CICU (Figure 1). The mean age was 66.7 ± 12.3 years and similar between ICU groups (Table 1). Patients admitted to the CICU were more likely to be male (67.6% vs 64.5%, P = 0.001), admitted to a teaching hospital and/or hospital with >750 beds (both, P < 0.001), and present with STEMI (57.0% vs 42.8%, P < 0.001) or cardiogenic shock (46.0% vs 37.4%, P < 0.001). Patients admitted to the MICU were more likely to have NSTEMI, cardiac arrest, and renal failure on admission (all, P < 0.001). Patients with complete heart block and any mechanical complications were more commonly admitted to the CICU (both, P < 0.05) (Table 2).Figure 1 Consort Diagram

AMI = acute myocardial infarction; CICU = cardiac intensive care unit; ICU = intensive care unit; IMV = invasive mechanical ventilation; MICU = medical intensive care unit.

Table 1 Characteristics of Patients With AMI Requiring IMV Admitted to the MICU vs CICU

	MICU Admission (n = 3,185)	CICU Admission (n = 9,454)	P value	Weighted Standardized Differences	
Demographics				
 Age, y	67.0 ± 12.6	66.6 ± 12.2	0.13	−0.001	
 Men	2,053 (64.5%)	6,391 (67.6%)	0.001	0.010	
 Race/ethnicity		0.009	−0.010	
 Black	586 (18.4%)	1,667 (17.6%)			
 Hispanic	240 (7.5%)	566 (6.0%)			
 White	1,939 (60.9%)	5,908 (62.5%)			
 Insurance type		0.02	−0.010	
 Commercial	691 (21.7%)	2,238 (23.7%)			
 Medicaid	435 (13.7%)	1,141 (12.1%)			
 Medicare	1,847 (58.0%)	5,397 (57.1%)			
 Other insurance	212 (6.7%)	678 (7.2%)			
Hospital characteristics		
 Teaching hospital	2,550 (80.0%)	8,520 (90.1%)	<0.001	0.011	
 Location				0.026	
 Urban	3,083 (96.8%)	9,300 (98.4%)	<0.001		
 Rural	103 (3.2%)	154 (1.6%)			
 Region			<0.001	0.028	
 Midwest	643 (20.2%)	2,573 (27.2%)			
 Northeast	1,374 (43.1%)	2,516 (26.6%)			
 South	793 (24.9%)	3,370 (35.6%)			
 West	375 (11.8%)	995 (10.5%)			
 Hospital size		<0.001	0.016	
 <350 beds	606 (19.1%)	783 (8.3%)			
 350-499 beds	484 (15.2%)	869 (9.2%)			
 500-750 beds	1,207 (37.9%)	3,599 (38.1%)			
 >750 beds	888 (27.9%)	4,203 (44.5%)			
Diagnoses present on admission		
 AMI type				0.007	
 NSTEMI	1,822 (57.2%)	4,069 (43.0%)	<0.001		
 STEMI	1,363 (42.8%)	5,385 (57.0%)	<0.001		
 Cardiac arrest	628 (19.7%)	1,464 (15.5%)	<0.001	0.001	
 Cardiogenic shock	1,190 (37.4%)	4,353 (46.0%)	<0.001	−0.005	
 Renal failure	1,338 (42.0%)	3,738 (39.5%)	0.01	<0.001	
 GI bleed	153 (4.8%)	413 (4.4%)	0.30	0.015	
 Sepsis	298 (9.4%)	698 (7.4%)	<0.001	0.005	
 DNR	305 (9.6%)	651 (6.9%)	<0.001	0.004	
Early hemodynamic supporta	
 Any vasoactive medication	1,666 (52.3%)	6,228 (65.9%)	<0.001	−0.025	
 Any MCS	785 (24.6%)	3,521 (37.2%)	<0.001	−0.002	
Values are mean ± SD or n (%).

AMI = acute myocardial infarction; CICU = cardiac intensive care unit; DNR = do-not-resuscitate; GI = gastrointestinal; MCS = mechanical circulatory support; MICU = medical intensive care unit; NSTEMI = non-ST-segment elevation myocardial infarction; STEMI = ST-segment elevation MI.

a Before or same day of intubation.

Table 2 Mechanical and Electrical Complications of Acute Myocardial Infarction Stratified by MICU vs CICU

	MICU Admission (n = 3,185)	CICU Admission (n = 9,454)	P Value	
Complete heart block	171 (5.4%)	690 (7.3%)	<0.001	
Any mechanical AMI complication	16 (0.5%)	110 (1.2%)	0.001	
 Papillary muscle rupture	4 (0.1%)	29 (0.3%)	0.08	
 Ventricular septal defect	9 (0.3%)	58 (0.6%)	0.03	
 Free wall rupture	3 (0.1%)	25 (0.3%)	0.08	
Postprocedural hematoma	28 (0.9%)	160 (1.7%)	0.001	
Values are n (%).

AMI = Acute myocardial infarction; CICU = Cardiac ICU; MICU = Medical intensive care unit.

Cardiovascular comorbidities, such as pre-existing coronary artery disease, prior MI and prior revascularization (PCI and CABG), and heart failure, were more common in patients admitted to the CICU while those admitted to the MICU were more likely to have noncardiac comorbidities, such as ESRD, chronic pulmonary disease, and cancer (all, P < 0.05) (Table 3). There were no differences in the proportion of patients with a history of stroke, valvular disease, diabetes, obesity, or liver disease admitted to either ICU (all, P > 0.05).Table 3 Comorbidities for Patients With AMI Requiring IMV Admitted the MICU vs CICU

	MICU Admission (n = 3,185)	CICU Admission (n = 9,454)	P Value	Weighted Standardized Differences	
Charlson comorbidity index	4.2 (2.4)	4.0 (2.3)	<0.001	0.007	
Coronary artery disease	2,569 (80.7%)	8,355 (88.4%)	<0.001	−0.013	
Hypertension	1,415 (44.4%)	4,391 (46.4%)	0.05	−0.005	
Dyslipidemia	1,672 (52.5%)	5,334 (56.4%)	<0.001	0.003	
Tobacco use	1,580 (49.6%)	4,820 (51.0%)	0.18	<0.001	
Prior AMI	489 (15.3%)	1,603 (17.0%)	0.04	0.010	
Prior PCI	629 (19.7%)	2,131 (22.5%)	<0.001	0.005	
Prior CABG	349 (11.0%)	871 (9.2%)	0.004	0.011	
Heart failure	1,431 (44.9%)	4,594 (48.6%)	<0.001	−0.026	
Valvular disease	430 (13.5%)	1,368 (14.5%)	0.18	−0.002	
Stroke	162 (5.1%)	525 (5.6%)	0.31	0.002	
PVD	325 (10.2%)	1,083 (11.5%)	0.06	<0.001	
End-stage renal disease	373 (11.7%)	861 (9.1%)	<0.001	−<0.001	
Chronic pulmonary disease	841 (26.4%)	2,222 (23.5%)	<0.001	0.012	
Cancer	117 (3.7%)	243 (2.6%)	0.001	0.004	
Obesity	693 (21.8%)	2,171 (23.0%)	0.16	−0.003	
Depression	318 (10.0%)	970 (10.3%)	0.66	−0.007	
Dementia	169 (5.3%)	339 (3.6%)	<0.001	0.004	
Diabetes	1,130 (35.5%)	3,262 (34.5%)	0.32	0.008	
Liver disease	162 (5.1%)	467 (4.9%)	0.74	−0.002	
Values are n (%).

AMI = acute myocardial infarction; CABG = coronary artery bypass grafting; CICU = cardiac intensive care unit; MICU = medical intensive care unit; PCI = percutaneous coronary intervention; PVD = peripheral vascular disease.

Critical care therapies

Procedure and medication use stratified by ICU type are shown in Figure 2 and Supplemental Table 2. Compared to patients with AMI admitted to the MICU, those admitted to the CICU more often underwent left and right heart catheterization, PCI, CABG, temporary pacemaker placement, any mechanical circulatory support, and left ventricular assist device implantation. Patients admitted to the MICU more commonly received noninvasive ventilation before IMV (P < 0.001). However, we did not find significant differences between ICU types for other noncardiac procedures and therapies, including renal replacement treatment, blood transfusions, colonoscopy, esophagogastroduodenoscopy, thoracentesis, and tracheostomy (P > 0.05). Utilization of critical care therapies was similar when stratified by AMI type (STEMI or NSTEMI) as the primary diagnosis (Supplemental Table 3).Figure 2 Procedures and Therapies for Patients Admitted to the MICU vs CICU

#Before Intubation. ∗Statistically significant, P > 0.05. CICU = cardiac intensive care unit; ECMO = extracorporeal membrane oxygenation; MCS = mechanical circulatory support; MICU = medical intensive care unit; PAC = pulmonary artery catheterization; RHC = right heart catheterization.

Clinical outcomes

The unadjusted in-hospital mortality of patients with AMI requiring IMV was not statistically different between those admitted to a CICU compared to an MICU (41.3% vs 42.7%, P = 0.15, respectively) (Table 4, Central Illustration). After multivariable adjustment for demographics, comorbidities, hospital characteristics, and markers of acuity, we found that admission to the CICU for AMI with IMV was associated with a lower in-hospital mortality (OR: 0.85, 95% CI: 0.78-0.93, P = 0.001) (Figure 3). In subgroup analysis, when stratified by the largest hospital bed size (>750 beds) (OR: 0.75, 95% CI: 0.64-0.88, P < 0.001) or teaching status (OR: 0.83, 95% CI: 0.75-0.92, P < 0.001), CICU admission remained associated with a lower adjusted in-hospital mortality. Finally, results were similar when stratified by patients presenting with cardiogenic shock, cardiac arrest, or STEMI (all, P > 0.05), but was not statistically different for those presenting with NSTEMI (P = 0.11).Table 4 Unadjusted Outcomes of Patients With AMI Requiring IMV Admitted to the MICU vs CICU

	MICU Admission (n = 3,185)	CICU Admission (n = 9,454)	P Value	
In-hospital death	1,360 (42.7%)	3,900 (41.3%)	0.15	
Ventilator days	4 (2-8)	4 (2-9)	<0.001	
Hospital LOS	8 (3-17)	10 (4-18)	<0.001	
ICU LOS	4 (2-9)	6 (2-11)	<0.001	
Renal replacement therapya	210 (6.6%)	656 (6.9%)	0.50	
Strokeb	156 (4.9%)	512 (5.4%)	0.26	
Total hospital costs, $	36,603 (18,948-70,056)	47,546 (24,442-85,857)	<0.001	
Disposition		<0.001	
 Home	476 (14.9%)	1,760 (18.6%)		
 Home with services	344 (10.8%)	1,022 (10.8%)		
 Skilled nursing/rehab	833 (26.2%)	2,392 (25.3%)		
 AMA	21 (0.7%)	51 (0.5%)		
 Hospice	123 (3.9%)	245 (2.6%)		
 Other	28 (0.9%)	79 (0.8%)		
Values are n (%) or median (IQR).

AMA = against medical advice; CICU = cardiac intensive care unit; LOS = length of stay; MICU = medical intensive care unit.

a Occurring after intubation.

b Coded as not present on admission.

Central Illustration In-Hospital Mortality Stratified by Intensive Care Unit Admission

CI = confidence interval; ICU = intensive care unit.

Figure 3 Forest Plot of Odds Ratios for In-Hospital Mortality Stratified by Intensive Care Unit Admission

CICU = cardiac intensive care unit; CS = cardiogenic shock; OHCA = out-of-hospital cardiac arrest; MICU = medical intensive care unit; NSTEMI = non-ST-segment elevation myocardial infarction; STEMI = ST-segment elevation myocardial infarction.

Both the median hospital (10 vs 8 days, P < 0.001) and ICU (6 vs 4 days, P < 0.001) lengths of stay were longer for patients admitted to the CICU compared to MICU (Table 4). Among survivors, ICU length of stay and total costs were not significantly different (both, P > 0.05). Median ventilator days were 4 in both ICUs. Total hospitalization costs were significantly higher in the CICU group ($47,545 vs $36,589, P < 0.001). After multivariable adjustment, hospital (IRR 1.06; 95% CI: 1.05-1.07, P < 0.001) and ICU length of stay (IRR 1.11; 95% CI: 1.09-1.13, P < 0.001) and total cost (+$10,081, 95% CI: $7,146-$13,015, P < 0.001) remained significantly higher in those admitted to the CICU. However, total ventilator days were no longer statistically different (P = 0.26).

Sensitivity analysis

Amongst patients without a cancer diagnosis, the in-hospital mortality was statistically similar (OR: 0.86, 95% CI: 0.78-0.94, P = 0.001). Similarly, excluding those with a DNR status on admission, in-hospital mortality remained lower in those admitted to a CICU (OR: 0.85, 95% CI: 0.78-0.94, P = 0.001). After IPTW including variables from Tables 1 and 3, admission to a CICU remained associated with lower in-hospital mortality (weighted mean −2.9%; 95% CI: −5.0% to −0.9%, P = 0.006). Weighted standardized differences were well-balanced (all <0.10) and are shown in Tables 1 and 3. IPTW results were unchanged in those with and without cancer and patients with and without a DNR status on admission.

Finally, including hospitals with both a MICU and CICU, in-hospital mortality was not statistically significant (OR: 0.95; 95% CI: 0.83-1.09, P = 0.47). For patients at hospitals with only a MICU or only a CICU, admission to a CICU was associated with a lower in-hospital mortality (OR: 0.75; 95% CI: 0.64-0.87, P < 0.003).

Discussion

In this multicenter, national analysis of ICU admission for mechanically ventilated patients with AMI, we found that admission to the CICU was associated with lower in-hospital mortality than patients admitted to the MICU. Our results persisted in key subgroups, including when stratified by hospital bed size, teaching status, and those presenting with cardiogenic shock and/or cardiac arrest. We also found that patients admitted to the CICU were more likely to undergo cardiac critical care therapies (eg, revascularization, temporary pacemakers, and mechanical circulatory support) while other noncardiac critical care therapies (eg, renal replacement therapy) were largely similar between ICUs. Notably, in subgroup analysis, there was no significant difference in mortality for hospitals with both ICUs. However, among hospitals with only a MICU or only a CICU, CICU admission remained associated with a lower mortality. Our findings highlight the importance of further investigation into the optimal triage of patients with acute cardiovascular disease.

A unique feature of this analysis is the focus on a specific disease process (eg, those with AMI requiring IMV) in a contemporary patient population. Utilizing a different administrative database, Brusca et al. previously assessed outcomes between adults with primary cardiac diagnoses admitted to a noncardiac ICUs vs CICU. Including over 16,000 patients from 14 hospitals between 2009 to 2014, they found that patients with primary cardiac diagnoses had a lower in-hospitality mortality when admitted to a CICU. These findings however were no longer statistically significant after controlling for patient- and hospital-level variables. Importantly, outcomes favored noncardiac ICUs after accounting for concurrent noncardiac ICU-level diagnoses.5 In our analysis, we directly compared CICUs with MICUs, as opposed to any noncardiac ICU, which are likely the 2 most commonly utilized ICUs for AMI. Our study was conducted in a more contemporary period, included substantially more hospitals, and had several additive findings to their study, including total hospital cost, ventilator days, and a detailed comparison of critical care therapies between each ICU.

In our multicenter cohort, patients admitted to the CICU more commonly underwent revascularization, hemodynamic monitoring with right heart or pulmonary artery catheterization, and hemodynamic support, both medically and mechanically, compared with patients admitted to the MICU. Procedures that have traditionally been described as noncardiac critical care therapies were statistically similar between ICUs.22 This pattern was largely unchanged when stratified by AMI type (STEMI and NSTEMI), which is particularly notable since NSTEMI includes a broad group of pathologies from demand-mediated injury from sepsis to coronary plaque rupture (the former more commonly seen in the MICU). These findings are in contrast to previous research, in which patients admitted to the MICU were more likely to receive general critical care therapies, such as vasoactive medication administration, renal replacement therapy, blood transfusions, and mechanical ventilation.23 However, their study is older, single-center, and provided limited data on other critical care procedures. Overall, our findings suggest that a traditional primary diagnosis, such as AMI, benefit from CICU care even when receiving a therapy not traditionally considered a cardiac therapy, such as IMV. Taken together, these studies may indicate the continued development and the need for CICU staffing models that incorporate critical care medicine.

Taken together, these findings highlight the growing need for further high-quality evidence to guide the ideal staffing, organizational structure, and triage of increasingly complicated patients in the CICU.6 Currently available data suggests that outcomes are improved in units staffed with dual trained critical care cardiologists as compared to general cardiologists7 and in units staffed by a single (often termed “closed” ICU models) as compared to multiple physicians (“open” ICU models).8 However, this data is limited as the former may not entirely represent care models in the United States and both were single-center studies. Most importantly, the work force necessary to accomplish these staffing models are far from feasible.9 Comanagement models with intensivists or advanced heart failure physicians with additional critical care training have been proposed as ways to fill these gaps.10,11 Regardless, it is critical that the evidence base and care models in the CICU improve to match the increasingly complicated patient population.

Study Limitations

In addition to being a retrospective study, we utilized administrative coding, which does not include details on vital signs, hemodynamics, coronary anatomy, severity of illness scores, or the reason why IMV was required. However, the Vizient Clinical Data Base includes detailed temporal data of procedures and pharmacy data, such as mechanical circulatory support and vasoactive use, which was utilized in our analyses and has previously been validated to estimate cardiogenic shock severity.24 In addition, we do not have details on ICU staffing models (eg, critical care managed, consultations, etc) or previously defined CICU levels.25 We are also unable to differentiate true plaque rupture from demand mediated ischemia, which may be particularly important for patients triaged to the MICU with other critical care diagnoses (eg, GI bleeding). Finally, given the nature of retrospective analyses, there is almost certainly residual confounding despite multivariable adjustment, multiple subgroup analyses, and sensitivity analysis using IPTW despite excellent covariate balance.

Conclusions

In this multicenter, national analysis of ICU admission for critically ill patients with AMI, we found that admission to the CICU was associated with lower in-hospital mortality than patients admitted to the MICU. These findings persisted when analyzed by hospital type and critical diagnoses (eg, cardiogenic shock and cardiac arrest). Finally, patients admitted to the CICU more commonly underwent cardiac critical care therapies and a similar proportion of historically described noncardiac critical care therapies. Future research is needed to explore other specific disease states that may benefit from CICU triage as well as to identify patient-specific factors driving these findings.Perspectives COMPETENCY IN MEDICAL KNOWLEDGE: Patients with acute myocardial infarction admitted to the cardiac ICU as compared to the medical ICU was associated with lower in-hospital mortality. Results were similar when stratified by the most critical ill patients, such as those presenting with cardiogenic shock and cardiac arrest.

TRANSLATIONAL OUTLOOK: Future studies are needed to identify the optimal triage of acute cardiovascular patients to the most suitable ICU.

Funding support and author disclosures

Dr Solomon has received research support from the 10.13039/100000002 National Institutes of Health Clinical Center intramural research funds. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Supplementary Data

Supplemental material

The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.

Appendix

For supplemental tables, please see the online version of this paper.
==== Refs
References

1 Quien M. Thomas A. Ludmir J. Miller P.E. Staffing models in the cardiac intensive care unit Curr Opin Crit Care 28 2022 453 459 35757936
2 Bohula E.A. Katz J.N. van Diepen S. Demographics, care patterns, and outcomes of patients admitted to cardiac intensive care units: the critical care cardiology trials network prospective North American multicenter registry of cardiac critical illness JAMA Cardiol 4 2019 928 935 31339509
3 Miller P.E. Chouairi F. Thomas A. Transition from an open to closed staffing model in the cardiac intensive care unit improves clinical outcomes J Am Heart Assoc 10 2021 e018182
4 Miller P.E. Thomas A. Breen T.J. Prevalence of noncardiac multimorbidity in patients admitted to two cardiac intensive care units and their association with mortality Am J Med 134 2021 653 661.e5 33129785
5 Brusca S.B. Galiatsatos P. Warner S. Outcomes of patients with primary cardiac diagnoses admitted to cardiac vs noncardiac intensive care units JACC Adv 1 2022 100114
6 Jentzer J.C. Ahmed A.M. Vallabhajosyula S. Shock in the cardiac intensive care unit: changes in epidemiology and prognosis over time Am Heart J 232 2021 94 104 33257304
7 Tsao C.W. Aday A.W. Almarzooq Z.I. Heart disease and stroke statistics-2022 update: a Report from the American heart association Circulation 145 2022 e153 e639 35078371
8 Rubini G.M. Miller P.E. Alviar C.L. Outcomes associated with respiratory failure for patients with cardiogenic shock and acute myocardial infarction: a substudy of the CULPRIT-SHOCK trial J Clin Med 9 2020 860 32245139
9 Chen R. Strait K.M. Dharmarajan K. Hospital variation in admission to intensive care units for patients with acute myocardial infarction Am Heart J 170 2015 1161 1169 26678638
10 Metkus T.S. Miller P.E. Alviar C.L. Advanced respiratory support in the contemporary cardiac ICU Crit Care Explor 2 2020 e0182
11 Metkus T.S. Miller P.E. Stephens R.S. Schulman S.P. Eid S.M. Incidence, predictors, and outcomes of failure of noninvasive ventilation in acute heart failure hospitalization Respir Care 65 2020 1527 1533 32398254
12 Metkus T.S. Albaeni A. Chandra-Strobos N. Eid S.M. Incidence and prognostic impact of respiratory support in patients with ST-segment elevation myocardial infarction Am J Cardiol 119 2017 171 177 27956004
13 Silverman M.G. Morrow D.A. Hospital triage of acute myocardial infarction: is admission to the coronary care unit still necessary? Am Heart J 175 2016 172 174 27179737
14 Jentzer J.C. Alviar C.L. Miller P.E. Trends in therapy and outcomes associated with respiratory failure in patients admitted to the cardiac intensive care unit J Intensive Care Med 37 2022 543 554 33759608
15 Miller P.E. Guha A. Khera R. National trends in healthcare-associated infections for five common cardiovascular conditions Am J Cardiol 124 2019 1140 1148 31371062
16 Miller P.E. Van Diepen S. Metkus T.S. Association between respiratory failure and clinical outcomes in patients with acute heart failure: analysis of 5 pooled clinical trials J Card Fail 27 2021 602 606 33556546
17 Vizient Clinical Data Base https://www.vizientinc.com/what-we-do/operations-and-quality/clinical-data-base
18 Schenck C. Banna S. Heck C. Ali T. Miller P.E. Rocuronium versus succinylcholine in patients with acute myocardial infarction requiring mechanical ventilation J Am Heart Assoc 12 2023 e8468
19 Axelrod D.A. Schwantes I.R. Harris A.H. The need for integrated clinical and administrative data models for risk adjustment in assessment of the cost transplant care Clin Transplant 36 2022 e14817
20 Iliescu C. Grines C.L. Herrmann J. SCAI expert consensus statement: Evaluation, management, and special considerations of cardio-oncology patients in the cardiac catheterization laboratory (Endorsed by the Cardiological Society of India, and Sociedad Latino Americana de Cardiologia Intervencionista) Cathet Cardiovasc Interv 87 2016 895 899
21 Wimmer N.J. Resnic F.S. Mauri L. Matheny M.E. Yeh R.W. Comparison of transradial versus transfemoral percutaneous coronary intervention in routine practice: evidence for the importance of "falsification hypotheses" in observational studies of comparative effectiveness J Am Coll Cardiol 62 2013 2147 2148 23954334
22 Lilly C.M. Zuckerman I.H. Badawi O. Riker R.R. Benchmark data from more than 240,000 adults that reflect the current practice of critical care in the United States Chest 140 2011 1232 1242 21868469
23 Goldfarb M. van Diepen S. Liszkowski M. Jentzer J.C. Pedraza I. Cercek B. Noncardiovascular disease and critical care Delivery in a contemporary cardiac and medical intensive care unit J Intensive Care Med 34 2019 537 543 29187011
24 Thayer K.L. Zweck E. Ayouty M. Invasive hemodynamic assessment and classification of in-hospital mortality risk among patients with cardiogenic shock Circ Heart Fail 13 2020 e007099
25 Morrow D.A. Fang J.C. Fintel D.J. Evolution of critical care cardiology: transformation of the cardiovascular intensive care unit and the emerging need for new medical staffing and training models: a scientific statement from the American Heart Association Circulation 126 2012 1408 1428 22893607
