Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y V O L . 6 9 , N O . 1 6 , 2 0 1 7

ª 2 0 1 7 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N

P U B L I S H E D B Y E L S E V I E R

I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 7 . 0 2 . 0 3 3

ORIGINAL INVESTIGATIONS

Acute Noncardiovascular Illness in theCardiac Intensive Care Unit

Eric M. Holland, MD,a,b Travis J. Moss, MD, MSCa,b

ABSTRACT

Fro

Me

Vir

dis

Ma

BACKGROUND Fifty years after the inception of the cardiac intensive care unit (CICU), noncardiovascular illnesses have

become more prevalent and may contribute to morbidity and mortality.

OBJECTIVES The authors performed multivariate statistical analyses to determine the association of acute

noncardiovascular illnesses with outcomes, including length of stay (LOS), mortality, and hospital readmission.

METHODS We studied 1,042 admissions between October 12, 2013 and November 28, 2014 to the CICU at the

University of Virginia Health System, a tertiary-care academic medical center. Through systematic inspection of individual

charts, we identified primary and secondary diagnoses, vital sign measurements, length of stay (LOS), hospital

readmissions, and mortality.

RESULTS The most common primary diagnosis was acute coronary syndrome (25%), which consisted of both

non�ST-segment elevation acute coronary syndrome (14%) and ST-segment elevation myocardial infarction (11%). Sepsis

was the most frequent noncardiovascular primary diagnosis (5%), but it only occurred in 16% of all admissions. Acute

kidney injury and acute respiratory failure each occurred in 30% of admissions. One-half of all admissions (n ¼ 524; 50%)

were marked by acute respiratory failure, acute kidney injury, or sepsis. Median LOS in the CICU and the hospital were 2

days (interquartile range [IQR]: 1 to 5 days) and 6 days (IQR: 3 to 11 days). Mortality was 7% in the CICU and 12% in the

hospital. Of the 920 patients who survived to hospital discharge, 171 (19%) were readmitted within 30 days. Sepsis, acute

kidney injury, and acute respiratory failure were associated with mortality. Acute kidney injury, acute respiratory failure,

and new-onset subclinical atrial fibrillation, which occurred in 8% of admissions, were all associated with CICU LOS.

CONCLUSIONS Many patients in the modern CICU have acute noncardiovascular illnesses that are associated with

mortality and increased LOS. (J Am Coll Cardiol 2017;69:1999–2007) © 2017 by the American College of Cardiology

Foundation.

I n 1961, Julian (1) presented the first descriptionof the coronary care unit to the British ThoracicSociety. Recognizing the association of cardiac

arrest with acute myocardial ischemia, these early vi-sionaries established a unit equipped with an electro-cardiography (ECG) alarm system and specialized

m the aDepartment of Medicine, University of Virginia Health System, Ch

dical Analytics, University of Virginia Health System, Charlottesville, Virg

ginia Health System. Both authors have reported that they have no rela

close.

nuscript received September 23, 2016; revised manuscript received Janua

staff trained in prompt treatment of peri-infarctionarrhythmias (2). Initial coronary care unit reportsreceived mixed reviews, with challengers believingit was irresponsible to suggest that all myocardialinfarction patients should be admitted to a criticalcare unit (2). However, the coronary care unit concept

arlottesville, Virginia; and the bCenter for Advanced

inia. The study was sponsored by the University of

tionships relevant to the contents of this paper to

ry 12, 2017, accepted February 10, 2017.

ABBR EV I A T I ON S

AND ACRONYMS

AF = atrial fibrillation

CICU = cardiac intensive care

unit

ECG = electrocardiogram

ICU = intensive care unit

IQR = interquartile range

LOS = length of stay

NSTE-ACS = nonLST-

segment elevation acute

coronary syndrome

OASIS = Oxford Acute Severity

of Illness Score

STEMI = ST-segment elevation

myocardial infarction

Holland and Moss J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7

Outcomes in the Modern Cardiac Intensive Care Unit A P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7

2000

quickly spread after Killip and Kimbal (3) re-ported an approximate 20% mortality reduc-tion with this care model.

More than 50 years later, a very differentlandscape from the one proposed by Julianhas evolved. The field of medicine, and inparticular, cardiovascular medicine, has pro-gressed exponentially with improvements indiagnostic studies and treatments, which haslead to a population that is living longer, butone that has comorbidities. This once-predominant post–acute coronary syndromeobservation unit has evolved into a hetero-geneous population rich in both cardiovas-cular and noncardiovascular illnesses. Katzet al. (4) demonstrated a significant temporal

change in the characteristics of cardiac intensive careunit (CICU) patients, with substantial increases inillness severity, comorbid critical illness, andnoncardiac procedures.

SEE PAGE 2008

Yet, as the once-defined lines of the CICU continueto blur with those of the medical and surgical inten-sive care units (ICUs), we have noted a surprising lackof data on the contemporary CICU related to de-mographics, length of stay (LOS), mortality, andreadmissions. These data are imperative in under-standing how to improve patient outcomes, whiledecreasing overall costs as we transition to a morevalue-based health care model in which qualitymeasures and readmissions are linked toreimbursements.

We tested the hypothesis that acute non-cardiovascular conditions are independently associ-ated with patient outcomes, including LOS, mortality,and readmission.

METHODS

STUDY POPULATION. We studied consecutiveadmissions to the CICU at the University of VirginiaHealth System with available continuous ECG datafrom October 12, 2013 to November 28, 2014. Teams ofmedicine residents, cardiology fellows, and attendingcardiologists staff the 10-bed unit. Physician staffingat nights is reduced, but both medicine residents anda cardiology fellow provide care around the clockunder the supervision of attending cardiologists. ACICU or advanced heart failure attending cardiologistand corresponding housestaff assume primary re-sponsibility for all patients admitted or transferred tothe CICU. Critical care consultation is available, butnot mandated. We followed patients through July 21,

2015 to capture all subsequent hospital admissions tothe University of Virginia. Each patient in the CICUhad continuous ECG monitoring, and an institutionalelectronic data warehouse archived the completemedical record.

We queried the electronic medical record to obtaindemographics, vital sign measurements, comorbiddiagnosis codes, LOS, mortality, and all InternationalClassification of Diseases-Ninth Revision diagnosescodes associated with an admission, which weascribed as acute and chronic comorbid conditions.Rather than rely on administrative coding to selectthe primary diagnosis for each admission, we care-fully reviewed every history, physical examination,and discharge summary. Patients whose hospitalcourse was complicated by cardiogenic shock andcardiac arrest received a primary diagnosis consistentwith the principal etiology of the shock or arrest.Thus, a patient presenting with ST-segment elevationmyocardial infarction (STEMI) found to be in cardio-genic shock received the primary diagnosis of STEMI.We categorized admissions as medical if no proced-ures were performed in the 7 days before or after ICUadmission (5). The University of Virginia InstitutionalReview Board approved this study, with a waiver ofinformed consent.

SEVERITY OF ILLNESS. To assess severity of illness,we calculated the Oxford Acute Severity of IllnessScore (OASIS) for all CICU admissions. OASIS is anabbreviated acute physiology score that has equiva-lent discrimination and calibration of the AcutePhysiology, Age, and Chronic Health Evaluation IVsystem from which it was derived (6). OASIS scoresthe worst measurements from the first 24 h of ICUadmission and includes pre-ICU LOS, age, Glasgowcoma score, heart rate, mean arterial pressure, res-piratory rate, temperature, urine output, mechanicalventilation status, and admission type.

RHYTHM CLASSIFICATION. We applied a previouslyvalidated rhythm classification methodology thatdetects atrial fibrillation (AF) at a burden of $5% ($90s of a 30-min segment) (7) to the continuous ECG re-cord. We categorized admissions as having previousAF if there was evidence of pre-existing AF or atrialflutter, as determined by diagnosis code, 12-lead ECGreport, or as the first detected rhythm from the ICUbedside monitor. In those without previous AF, if wesubsequently detected AF or atrial flutter during ICUmonitoring, we categorized the admission as new AF.We categorized all other admissions as being withoutAF. We considered new AF to be clinical or recognizedif it was associated with a diagnosis code or confirmedby 12-lead ECG during the hospitalization.

TABLE 1 Characteristics of CICU Admissions (N ¼ 1,042)

Age, yrs 65.9 (55.6–75.9)

Male 647 (62)

BMI, kg/m2 27.4 (23.9–33.4)

Ethnicity

Caucasian 827 (79)

African American 187 (18)

Source of admission

Emergency department 455 (44)

Outside hospital transfer 369 (35)

Direct admission 218 (21)

Comorbidities

Hypertension 842 (81)

Coronary artery disease 762 (73)

Hyperlipidemia 702 (67)

Arrhythmia 674 (65)

Heart failure 621 (60)

Diabetes 452 (43)

Chronic kidney disease 398 (38)

Tobacco abuse 331 (32)

Conduction disease 296 (28)

Valvular heart disease 283 (27)

Chronic pulmonary disease 276 (27)

Cardiomyopathy 246 (24)

Obesity 245 (24)

Pulmonary hypertension 229 (22)

Obstructive sleep apnea 205 (20)

Ischemic stroke 186 (18)

Malignancy 154 (15)

Alcohol abuse 73 (7)

Hyperthyroidism 24 (2)

Type of admission

Medical 726 (70)

Scheduled procedure 267 (26)

Unscheduled procedure 49 (5)

NYHA functional classification

I 288 (30)

II 225 (23)

III 251 (26)

IV 208 (21)

Not obtained 70 (7)

Continued in the next column

TABLE 1 Continued

Before admission medications

Beta-blocker or calcium-channel blocker 600 (58)

Antiarrhythmic agent 118 (11)

Antiplatelet agent 622 (60)

Anticoagulation 272 (26)

Digoxin 63 (6)

Severity of illness

OASIS 25 (20.0–30.0)

AF

CHA2DS2-VASc score 4.0 (3.0–5.0)

CHA2DS2-VASc score $2 944 (91)

No AF 502 (48)

Known AF 443 (43)

New subclinical AF 79 (7.6)

New clinical AF 18 (1.7)

Median days in CICU 2.0 (1.0–4.0)

Median days in hospital 6.0 (3.0–11.0)

CICU mortality 69 (6.6)

Hospital mortality 122 (12)

30-day readmission to hospital 171 (19)

Values are median (range) or n (%).

AF ¼ atrial fibrillation; BMI ¼ body mass index; CHA2DS2-VASc ¼ congestiveheart failure, hypertension, age $75 years, diabetes mellitus, prior stroke, tran-sient ischemic attack, or thromboembolism, vascular disease, age 65–74 years, sexcategory (female); CICU ¼ coronary intensive care unit; NYHA ¼ New York HeartAssociation; OASIS ¼ Oxford Acute Severity of Illness Score.

TABLE 2 Descriptive Statistics Stratified by Type of Admission

MedicalScheduledProcedure

UnscheduledProcedure p Value

n 726 (70) 267 (26) 49 (5) –

Age, yrs 65 (54–74) 71 (61–80) 70 (53–77) <0.0001

OASIS 26 (23–32) 20 (16–25) 27 (22–31) <0.0001

Mortality 102 (14) 14 (5) 6 (12) 0.0007

CICU LOS 2.2 (1.2–4.8) 2.0 (1.0–3.1) 1.2 (0.8–2.1) <0.0001

30-day readmissions 126 (17) 34 (13) 11 (22) 0.1107

Values n (%) or median (interquartile range).

LOS ¼ length of stay; other abbreviations as in Table 1.

J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7 Holland and MossA P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7 Outcomes in the Modern Cardiac Intensive Care Unit

2001

STATISTICAL ANALYSES. For categorical variables,we calculated percentages and counts, and forcontinuous variables, we calculated median (inter-quartile range [IQR]). We tested for differences inadmission characteristics among the 3 types of ad-missions using a Kruskal-Wallis test. To assess theassociation of variables on CICU LOS, hospital mor-tality, and 30-day readmission, we developedregression models, adjusting for demographics,severity of illness, and common acute comorbidconditions. On the basis of previous work thatexamined the impact of subacute, potentially cata-strophic illness and AF in the medical and surgicalICUs, we included age, illness severity (OASIS), and

the diagnosis codes that indicated acute kidneyinjury, hemorrhage, acute respiratory failure, andsepsis as candidate predictors of hospital mortality(8,9). We also included cardiogenic shock and cardiacarrest on the basis of previously published findings(10,11). In addition, for CICU LOS, we included AF andpost-operative status as candidate predictors on thebasis of previous findings (9). For hospital read-missions, we included risk factors associated withreadmission, including diabetes mellitus, heart fail-ure, chronic kidney disease, and previous stroke(12,13). We also examined the impact of CICU LOS,and whether an admission was for an anticipated orunanticipated procedure. For patients with multiple

FIGURE 1 Primary Diagnoses in the CICU

Pericardial Disease2%

Arterial Aneurysm orDissection 3%

Bradyarrhythmia4%

Atrial Tachyarrhythmia4%

VentricularTachyarrhythmia

6%

STEMI44%

NSTE-ACS56%

Hypertensive Crisis4%

Sepsis5%

Other21%

Valvular Disease10%

Heart Failure15%

AcuteCoronary Syndrome

25%

Primary diagnoses associated with admissions arranged in order of frequency. CICU ¼ cardiac intensive care unit; NSTE-ACS ¼ non�ST-segment elevation acute

coronary syndrome; STEMI ¼ ST-segment elevation myocardial infarction.

Holland and Moss J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7

Outcomes in the Modern Cardiac Intensive Care Unit A P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7

2002

readmissions during the study period, we analyzedonly the first admission.

We allowed continuous variables, such as age, tohave nonmonotonic associations through use ofrestricted cubic splines (14). In all multivariableregression models, we always constrained our eventsper variable ratio to exceed 15 (15). We quantifiedpredictive accuracy using a concordance index(C-statistic) or R2, and validated these models inter-nally using bootstrap resampling to estimate theperformance on a new sample of observations fromthe same study patients (14). We performed all sta-tistical analyses in R 3.2.3 (16).

RESULTS

We analyzed 1,042 CICU admissions with availableECG data (84% of 1,238 admissions screened). Themedian age was 65.9 years (IQR: 55.6 to 75.9 years),and 62% of admissions were men. Although manyadmissions originated from the emergency depart-ment (44%), more than one-third were transferredfrom outside hospitals (35%). Severity of illness, asquantified by OASIS, was variable, with scoresranging from 6 to 69, with a median value of 25(IQR: 20 to 30) (Table 1).

TYPES OF ADMISSION. We categorized 70% ofadmissions as medical, 26% as scheduled procedures,and 4% as unscheduled procedures. Minimally inva-sive valvular procedures (e.g., transcatheter aorticvalve replacement, MitraClip, Abbott Vascular, MenloPark, California) accounted for 26% of procedure ad-missions; 11% were for electrophysiology procedures,and the remaining 63% were for surgical procedures inthe operating room. Compared with admissions foracute medical illness and unscheduled procedures,those for scheduled procedures had decreased acuityandmortality, and fewer 30-day hospital readmissions(Table 2).

PRIMARY DIAGNOSES AND COMORBID DISEASE. Themost common primary diagnosis was acute coronarysyndrome (ACS), which accounted for 25% of all ad-missions, and comprised non�ST-segment elevationacute coronary syndrome (NSTE-ACS) (14%) andSTEMI (11%). Acute heart failure exacerbation (15%)and valvular disease (10%) were also frequent causesof admission. Noncardiovascular primary diagnosesaccounted for 14% of admissions, with sepsis mostnotably the fifth most common primary diagnosis(5%). Figure 1 depicts the most common primary di-agnoses for all CICU admissions. Frequently,

FIGURE 2 Determinants of Hospital Mortality

0 1 2 3 4 5 6 7 8 9Odds Ratio

10

Age1.37 (0.97 - 1.94)

OASIS2.49 (1.86 - 3.33)

Acute Kidney Injury1.85 (1.15 - 2.98)

Hemorrhage2.45 (0.65 - 2.98)

Acute Respiratory Failure3.64 (2.17 - 6.11)

Sepsis2.09 (1.22 - 3.58)

Cardiogenic Shock4.87 (2.78 - 8.56)

Cardiac Arrest4.61 (2.09 - 10.19)

Independent predictors of mortality as assessed by logistic regression analysis. For continuous predictors, we estimated the odds ratio

between the interquartile range (75th percentile value vs. 25th percentile value). OASIS ¼ Oxford Acute Severity of Illness Score.

J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7 Holland and MossA P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7 Outcomes in the Modern Cardiac Intensive Care Unit

2003

admissions were complicated by cardiogenic shock(n ¼ 104; 10%) and cardiac arrest (n ¼ 52; 5%),including 1 of every 4 acute coronary syndrome ad-missions to the CICU. One-half of all admissions(n ¼ 524; 50%) were marked by either acute respira-tory failure, acute kidney injury, or sepsis. Mostadmissions for acute noncardiovascular illness hadsevere chronic cardiovascular disease (85%). Pre-existing AF was present in 43% of patients, whereasnew AF occurred in 9%. Of the 97 patients with newAF, 79 had subclinical AF. Nearly all patients (91%)had a CHA2DS2-VASc score $2, with a median score of4 (Table 1).

OUTCOMES. Median LOS in the CICU and hospitalwere 2 days (IQR: 1 to 5 days) and 6 days (IQR: 3 to11 days), respectively. Overall hospital mortality was12%, with 7% occurring within the CICU. Of the 920patients who survived to hospital discharge, 171 (19%)were readmitted to the University of Virginia MedicalCenter within 30 days, of whom 49 (29%) were read-mitted to the CICU. Most readmissions were for car-diovascular diagnoses (n ¼ 99; 58%), with heartfailure being the most common (n ¼ 30; 17%). Other

common principal diagnoses for readmissionsincluded sepsis (n ¼ 14; 8%), gastrointestinal bleeding(n ¼ 14; 8%), ACS (n ¼ 11; 6%), and atrial tachyar-rhythmia (n ¼ 8; 5%).

Figure 2 shows the association between multipleindependent predictors in relation to hospital mor-tality. The model for hospital mortality with only 8predictors had excellent discrimination (validatedC-statistic: 0.90; optimism <0.01). Sepsis, the mostfrequent noncardiovascular primary diagnosis (5%),but which was present in 16% of all admissions, wasassociated with an increased risk of mortality, with anodds ratio of 2.1 (95% confidence interval: 1.22 to 3.58;p ¼ 0.0076). Cardiogenic shock, cardiac arrest, acuterespiratory failure, and hemorrhage also had strongassociations. Regression analysis for CICU LOS(Table 3) demonstrated that acute kidney injury,acute respiratory failure, new subclinical AF, andpost-operative states were all significantly associatedwith longer LOS in the CICU (with marginal effectsizes of 5.5 � 1.5 days, 6.7 � 1.7 days, 6.3 � 2.6 days,and 3.6 � 1.7 days, respectively). Thirty-day hospitalreadmission analysis (Table 3) showed that comorbidconditions, such as chronic kidney disease, were

TABLE 3 Determinants of CICU LOS and 30-Day Hospital

Readmission

Predictors Coefficient SE p Value

CICU LOS

Age �0.2209 0.0453 <0.0001

OASIS �0.0700 0.0897 0.4353

Acute kidney injury 5.5497 1.4793 0.0002

Hemorrhage �2.9431 5.5823 0.5981

Acute respiratory failure 6.6954 1.6693 <0.0001

Sepsis 2.6333 1.952 0.1776

Cardiogenic shock �0.0572 2.3011 0.9802

Cardiac arrest �4.3523 3.3593 0.1954

New subclinical AF 6.28817 2.5656 0.0144

New clinical AF 3.5183 5.151 0.4947

Previous AF 2.4192 1.4544 0.0968

Post-operative 3.6355 1.7033 0.0330

30-day hospital readmission

Age �0.0100 0.0061 0.1031

OASIS 0.0022 0.0131 0.8688

Acute kidney injury 0.1154 0.2044 0.5722

Hemorrhage 0.0562 0.8394 0.9466

Acute respiratory failure 0.0957 0.2316 0.6797

Sepsis 0.1730 0.2654 0.5145

Diabetes mellitus 0.3444 0.1853 0.0631

Heart failure �0.0734 0.2062 0.7219

Chronic kidney disease 0.5502 0.2055 0.0074

Previous stroke 0.4064 0.2156 0.0594

CICU LOS 0.0129 0.0114 0.2571

Scheduled procedure �0.3606 0.2375 0.1290

Unscheduled procedure 0.4404 0.3777 0.2436

Abbreviations as in Tables 1 and 2.

Holland and Moss J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7

Outcomes in the Modern Cardiac Intensive Care Unit A P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7

2004

associated with increased risk of readmission(C-statistic: 0.61; optimism: 0.04).

DISCUSSION

In a modern tertiary-care academic medical centerCICU, we studied 1,042 patients admitted over thecourse of a year, and found that although cardiovas-cular conditions, including acute coronary syndrome,remained common, acute noncardiovascular condi-tions were strongly associated with outcomes,including mortality and LOS. One-half of all admis-sions (n ¼ 524; 50%) were marked by either acuterespiratory failure, acute kidney injury, or sepsis(Central Illustration). To our knowledge, this was thefirst study to report hospital readmission data on anunselected cohort of CICU patients.

The CICU and hospital mortality observed in >1year of consecutive admissions were 7% and 12%,respectively. These findings were consistent withthose reported in several previous studies in whichCICU mortality ranged from 6% to 9%, and hospitalmortality ranged from 11 to 33% (4,17–19). New-onset

subclinical AF, which occurred in 8%, was associatedwith increased CICU LOS, which was consistent withsimilar findings in medical and surgical ICU cohorts(9,20).

With continued advances in revascularization andpharmacotherapy, hospital mortality in acutemyocardial infarctions fell significantly in the 1960s,and has continued to decline from w12% in the 1990sto w6% in the mid-2000s (3,21). Despite advances inACS treatment, Katz et al. (4) noted that CICU mor-tality rates did not change significantly over the 18years observed due to a significant increase in acutecritical illnesses and complex comorbidities. Like-wise, the balance of CICU diagnoses shifted, with asignificant decline in STEMI and significant increasesin both NSTE-ACS and noncardiovascular disease(4,21,22). Another study of 7,869 patients found thatthere was no difference in clinical outcomes in stableNSTE-ACS among those initially admitted to a ward orto a CICU (23).

Our data echoed these findings, with NSTE-ACSand acute heart failure exacerbation being morecommon than STEMI. Similarly, noncardiovasculardiagnoses and comorbid conditions were both prev-alent. Cardiologists staff most CICUs and have rela-tively limited experience treating the most severeforms of hospital-acquired noncardiovascular condi-tions compared with intensivists who staff medicalICUs. ICUs (including CICUs) that are either staffed byintensivists or mandate critical care consultationshave improved patient outcomes compared with ICUswithout such staffing (24,25).

We demonstrated that the strongest predictors ofincreased CICU LOS were noncardiovascular condi-tions; patients with acute respiratory failure or acutekidney injury stayed an average of 6 to 7 days longer inthe CICU, independent of other risk factors. Thesefindings affirmed recent statements from the Amer-ican Heart Association calling for staffing by eitherdedicated cardiac intensivists or co-management withgeneral intensivists, in which the experienced inten-sivist takes primary responsibility for the care of eachcritically ill patient in the CICU (e.g., patientswith multiorgan dysfunction or those requiring me-chanical ventilation), with ongoing collaborative co-management by the cardiologist (26,27). Others haveadvocated for greater adoption of standard-of-careprotocols and practices aimed to prevent or mitigatethe risk of these acute noncardiovascular conditionsencountered in the CICU (28).

Approximately 1 of every 5 patients in our study wasreadmitted within 30 days of discharge. In recentyears, the U.S. Centers for Medicare & Medicaid Ser-vices began publicly reporting readmission rates and

CENTRAL ILLUSTRATION Impact of Noncardiovascular Illness in the CICU

Holland, E.M. et al. J Am Coll Cardiol. 2017;69(16):1999–2007.

Patients represented as blue and black are those with and without acute kidney injury, acute respiratory failure, or sepsis, and the patients represented in

red are those who died during admission. CICU ¼ cardiac intensive care unit.

J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7 Holland and MossA P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7 Outcomes in the Modern Cardiac Intensive Care Unit

2005

reducing reimbursements to hospitals for higher thanexpected readmission rates. As a result, hospitals havefocused on preventing readmissions from acutemyocardial infarction and heart failure. Our analysisunderscored the difficulty of predicting hospitalreadmission accurately. A systematic review byKansagra et al. (29) found that current models per-formed so poorly that using risk-standardized read-mission rates to compare and reimburse hospitals wasa potentially unsuitable practice. More recently, Katz

et al. (30) described the evolution of the CICU andenumerated several opportunities for further im-provements in care delivery, which will require newchannels of investigation to examine care models,training pathways, therapeutic strategies, and qualityimprovement metrics, such as ICU readmission, whichhas yet to be substantiated among CICU cohorts.

STUDY LIMITATIONS. This was a single-center retro-spective study in an unselected cohort from the CICU

PERSPECTIVES

COMPETENCY IN PATIENT CARE AND

PROCEDURAL SKILLS: Patients in the CICU often

have noncardiovascular illnesses that contribute to

mortality, lengthy hospital stays, and hospital read-

missions. One-half of all cases are marked by acute

respiratory failure, acute kidney injury, or sepsis. The

extent of comorbidities demands that clinicians caring

for patients in the CICU have the competencies

necessary to manage patients with both cardiovascu-

lar and noncardiovascular conditions.

TRANSLATIONAL OUTLOOK: Additional studies

are needed to define optimum models of team-based

care for critically ill patients in the CICU.

Holland and Moss J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7

Outcomes in the Modern Cardiac Intensive Care Unit A P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7

2006

of a rural, tertiary-care academic medical center, andalthough our findings were consistent with previouslyreported studies, other hospitals might treat patientswith different acuity or provide different specialtyservices that could affect the generalizability of theseresults. Continuous ECG data were unavailable for 16%of all CICU admissions due to technical complicationswith our network processing and archival system. Inaddition, we might have failed to account for admis-sions to other hospitals. LOS in clinical studies ischallenging to predict and is confounded by nonclin-ical factors, such as fluctuations in the availability ofboth floor beds and nurse staffing. We calculatedillness severity using OASIS, which failed to accountfor aggressive resuscitation or interventions that werepursued among themost critically ill patients. Age wasnonsignificantly associated with mortality, likely dueto our study being underpowered to detect a differ-ence after adjusting for other characteristics and to theheterogeneous age distribution among those admittedwith high-risk conditions, such as cardiac arrest andcardiogenic shock. Although we relied on standarddiagnosis codes (which have variable validity) tocharacterize comorbid conditions, we carefullyinspected themedical record, including progress notesand discharge summaries, to characterize the primarydiagnoses of admissions.

CONCLUSIONS

Noncardiovascular illnesses, in particular, acute res-piratory failure, acute kidney injury, and sepsis,complicate CICU admissions by contributing to mor-tality and LOS. Many patients in the modern CICUhave acute noncardiovascular illnesses that are

associated with mortality and increased LOS. Clini-cians caring for these patients increasingly needgreater competency in treating complex comorbidnoncardiovascular conditions.

ACKNOWLEDGMENTS The authors thank J. RandallMoorman and Douglas E. Lake for mentorship; GaryHuband and Jacalyn M. Huband for design andoperation of infrastructure to collect, process, andarchive physiological monitoring data; and T. BruceHope and John Ainsworth for assistance with datawarehouse queries.

ADDRESS FOR CORRESPONDENCE: Dr. Eric M.Holland, Department of Medicine, University of Vir-ginia Health System, PO Box 800158, Charlottesville,Virginia 22908-0158. E-mail: eh7sb@virginia.edu.

RE F E RENCE S

1. Julian DG. Treatment of cardiac arrest in acutemyocardial ischaemia and infarction. Lancet 1961;2:840–4.

2. Julian DG. The history of coronary care units. BrHeart J 1987;57:497–502.

3. Killip T III, Kimball JT. Treatment of myocardialinfarction in a coronary care unit. A two yearexperience with 250 patients. Am J Cardiol 1967;20:457–64.

4. Katz JN, Shah BR, Volz EM, et al. Evolution ofthe coronary care unit: clinical characteristics andtemporal trends in healthcare delivery and out-comes. Crit Care Med 2010;38:375–81.

5. Le Gall JR, Lemeshow S, Saulnier F. A newSimplified Acute Physiology Score (SAPS II) basedon a European/North American multicenter study.JAMA 1993;270:2957–63.

6. Johnson AE, Kramer AA, Clifford GD. A newseverity of illness scale using a subset of Acute

Physiology and Chronic Health Evaluation dataelements shows comparable predictive accuracy.Crit Care Med 2013;41:1711–8.

7. Carrara M, Carozzi L, Moss TJ, et al. Heart ratedynamics distinguish among atrial fibrillation,normal sinus rhythm and sinus rhythm withfrequent ectopy. Physiol Meas 2015;36:1873–88.

8. Moss TJ, Lake DE, Calland JF, et al. Signaturesof subacute potentially catastrophic illness in theICU: model development and validation. Crit CareMed 2016;44:1639–48.

9. Moss TJ, Calland JF, Enfield KB, et al. New-onset atrial fibrillation in the critically ill. Crit CareMed 2017 Mar 14 [E-pub ahead of print].

10. Granger CB, Goldberg RJ, Dabbous O, et al.,Global Registry of Acute Coronary Investigators.Events predictors of hospital mortality in theGlobal Registry of Acute Coronary Events. ArchIntern Med 2003;163:2345–53.

11. Kontos MC, Scirica BM, Chen AY, et al. Car-diac arrest and clinical characteristics, treatmentsand outcomes among patients hospitalized withST-elevation myocardial infarction in contempo-rary practice: a report from the NationalCardiovascular Data Registry. Am Heart J 2015;169:515–22.e1.

12. Donzé J, Lipsitz S, Bates DW, Schnipper JL.Causes and patterns of readmissions in patientswith common comorbidities: retrospective cohortstudy. BMJ 2013;347:f7171.

13. Condon C, Lycan S, Duncan P, Bushnell C.Reducing readmissions after stroke with a struc-tured nurse practitioner/registered nurse transi-tional stroke program. Stroke 2016;47:1599–604.

14. Harrell FE Jr. Regression Modeling StrategiesWith Applications to Linear Models, Logistic andOrdinal Regression and Survival Analysis. 2ndedition. New York, NY: Springer-Verlag, 2015.

J A C C V O L . 6 9 , N O . 1 6 , 2 0 1 7 Holland and MossA P R I L 2 5 , 2 0 1 7 : 1 9 9 9 – 2 0 0 7 Outcomes in the Modern Cardiac Intensive Care Unit

2007

15. Steyerberg EW, Schemper M, Harrell FE.Logistic regression modeling and the number ofevents per variable: selection bias dominates.J Clin Epidemiol 2011;64:1464–5. author reply1463–4.

16. R Core Team. R: A language and environmentfor statistical computing. R Foundation for Sta-tistical Computing, Vienna, Austria. 2015;3.2.0.Available at: https://www.R-project.org. AccessedFebruary 24, 2017.

17. Teskey RJ, Calvin JE, McPhail I. Diseaseseverity in the coronary care unit. Chest 1991;100:1637–42.

18. Zobel C, Dörpinghaus M, Reuter H, Erdmann E.Mortality in a cardiac intensive care unit. Clin ResCardiol 2012;101:521–4.

19. Ratcliffe JA, Wilson E, Islam S, et al. Mortalityin the coronary care unit. Coron Artery Dis 2014;25:60–5.

20. Shaver CM, Chen W, Janz DR, et al. Atrialfibrillation is an independent predictor of mortalityin critically ill patients. Crit Care Med 2015;43:2104–11.

21. Rogers WJ, Frederick PD, Stoehr E, et al.Trends in presenting characteristics and hospital

mortality among patients with ST elevation andnon-ST elevation myocardial infarction in the na-tional registry of myocardial infarction from 1990to 2006. Am Heart J 2008;156:1026–34.

22. Sjoding MW, Prescott HC, Wunsch H,Iwashyna TJ, Cooke CR. Longitudinal changes inICU admissions among elderly patients in theUnited States. Crit Care Med 2016;44:1353–60.

23. van Diepen S, Lin M, Bakal JA, et al. Do stablenon–ST-segment elevation acute coronary syn-dromes require admission to coronary care units?Am Heart J 2016;175:184–92.

24. Pronovost PJ, Angus DC, Dorman T,Robinson KA, Dremsizov TT, Young TL. Physicianstaffing patterns and clinical outcomes in criticallyill patients: a systematic review. JAMA 2002;288:2151–62.

25. Na SJ, Chung CR, Jeon K, et al. Associationbetween presence of a cardiac intensivist andmortality in an adult cardiac care unit. J Am CollCardiol 2016;68:2637–48.

26. Morrow DA, Fang JC, Fintel DJ, et al. Evolu-tion of critical care cardiology: transformation ofthe cardiovascular intensive care unit and theemerging need for new medical staffing and

training models: a scientific statement from theAmerican Heart Association. Circulation 2012;126:1408–28.

27. O’Malley RG, Olenchock B, Bohula-May E,et al. Organization and staffing practices in UScardiac intensive care units: a survey on behalf ofthe American Heart Association Writing Group onthe Evolution of Critical Care Cardiology. Eur HeartJ Acute Cardiovasc Care 2013;2:3–8.

28. van Diepen S, Sligl WI, Washam JB, Gilchrist IC,Arora RC, Katz JN. Prevention of critical care com-plications in the coronary intensive care unit: pro-tocols, bundles, and insights from intensive carestudies. Can J Cardiol 2017;33:101–9.

29. Kansagara D, Englander H, Salanitro A,et al. Risk prediction models for hospitalreadmission: a systematic review. JAMA 2011;306:1688–98.

30. Katz JN, Minder M, Olenchock B, et al. Thegenesis, maturation, and future of critical carecardiology. J Am Coll Cardiol 2016;68:67–79.

KEY WORDS critical illness, length of stay,mortality, outcome, readmission, sepsis