Invasive Hemodynamic Assessment and Classification of In-Hospital Mortality Risk Among Patients With Cardiogenic Shock
What is New?
-
Using data from the Cardiogenic Shock Working Group registry inclusive of contemporary short-term, percutaneous mechanical circulatory support devices and invasive hemodynamic data, we report a novel validation analysis showing that Society for Cardiovascular Angiography and Intervention stages directly associate with in-hospital mortality.
-
We provide new insight into the distribution of short-term mechanical circulatory support use across Society for Cardiovascular Angiography and Intervention stages.
-
We show that elevated right heart filling pressures (venous congestion) are common and associated with worsening shock severity and in-hospital mortality.
What are the Clinical Implications?
-
Our findings suggest that more clinical data inclusive of hemodynamic and metabolic variables are required to confirm the specific definitions of Society for Cardiovascular Angiography and Intervention stages for patients with cardiogenic shock due to myocardial infarction or heart failure and further identify venous congestion as an important marker of risk for in-hospital mortality.
-
Future studies exploring whether a strategy of early venous decongestion improves clinical outcomes in cardiogenic shock are required.
Introduction
See Editorial by Jentzer
Cardiogenic shock (CS) is a complex clinical syndrome that begins with impaired cardiac function leading to systemic hypoperfusion and results in hemodynamic, neurohormonal, and metabolic changes that progressively worsen without treatment. Despite major advances in drug and short-term mechanical circulatory support (MCS) device therapies over the past 2 decades, reported 30-day mortality due to CS remains largely unchanged, ranging between 30% and 60%.1–4 One explanation for the broad range and inconsistent mortality over time may be that the lack of clear criteria for risk stratification of patients at the time of presentation obscures survival trends over time in specific subgroups, be they at low, intermediate, or high risk. This issue is further complicated by the fact that most studies involving CS focus on patients with myocardial infarction (MI).5–7 However, the number of patients with CS in the setting of decompensated heart failure (HF) has grown, owing to exponential growth of the HF population.8 Survival trends and risk stratification for CS have not been adequately investigated in the HF population.
The use of short-term MCS devices has also increased and device options for CS now include the intraaortic balloon pump, trans-valvular axial flow pumps (Impella; Abiomed Inc), left atrial to femoral artery pumps (TandemHeart; LivaNova Inc), venoarterial extracorporeal membrane oxygenation, and extracorporeal centrifugal flow pumps.9,10 With an increasing number of device options for these critically ill patients, risk stratification of patients presenting with CS is now more important than ever, since clarification of mortality and hemodynamic deficits in risk subsets may inform the development of treatment algorithms and the design of registry studies and randomized controlled trials which are necessary to evaluate clinical benefits.
Recently, a proposed staging system for CS based on input from a multi-disciplinary panel of clinical experts was proposed by the Society for Coronary Angiography and Intervention (SCAI) and endorsed by 4 other American medical associations.11 The SCAI system includes 5 classes of CS: (1) at risk for CS, (2) beginning CS, (3) classic CS, (4) deteriorating CS, and (5) extreme CS. Each stage is defined by physical exam, biochemical, and hemodynamic findings and were intentionally left as general definitions to accommodate the variability among clinical parameters available at the time of presentation. The SCAI staging system also proposes that increasing intensity of drug and device treatment over time accompanies clinical deterioration.
Two recent studies used markers of hypoperfusion and lactate levels, respectively, to define SCAI stages and showed a direct association between mortality and increasing SCAI stage. Limitations of these studies include the single-center study design, the lack of invasive hemodynamic data, and skewed distribution of short-term MCS devices in the study population.12,13 Accordingly, additional studies are required to explore the utility of SCAI stages with contemporary real-world experience and to further determine the importance of hemodynamic parameters in risk stratifying CS patients.
To begin addressing these critical gaps in knowledge, we employed a multicenter registry of patients with CS due to decompensated HF, MI, or other causes, hospitalized at 8 medical centers in the United States. The primary objective of this study was to test whether the SCAI classification system successfully stratifies patients at risk of all-cause in-hospital mortality and to further assess associations between hemodynamic parameters at presentation with mortality.
Methods
Data Source
The authors declare that all supporting data are available within the article and in the Data Supplement. The CS Working Group (CSWG) is an academic research consortium of hospitals in the United States inclusive of a national registry of all-cause CS that began in 2016 with 4 initial sites across the United States contributing data on at least 100 adult refractory patients with CS annually. The registry grew to include 8 total contributing sites by 2019. The registry includes a standardized set of data elements which were defined by principal investigators from the CSWG. These include patient, procedural, and hospital characteristics. Data represent discrete CS in patient cases treated at each institution between 2016 and 2019. Patient demographic, laboratory, and hemodynamic data were collected at a single time point as close to admission as possible, before initiation of mechanical support, in the hospital records. Information about pharmacological and device therapies represented the maximum therapies provided during the hospitalization (detailed further below). CS diagnosis was physician-adjudicated at each site and was defined as a sustained episode of systolic blood pressure <90 mm Hg for at least 30 minutes and a cardiac index (CI) <2.2 L/(min·m2) determined to be secondary to cardiac dysfunction, and the requirement for either pharmacological support (vasopressors or inotropes) or short-term MCS (ie, intraaortic balloon pump, Impella, venoarterial extracorporeal membrane oxygenation, or extracorporeal centrifugal flow pumps) at any time throughout a patient’s hospitalization. Quality assurance was achieved through adjudication at each site by the respective clinical coordinators and principal investigator. In addition, values were centrally audited and screened by the CSWG research team (K.L. Thayer, S. Newman, L. Jorde, J.L. Haywood, N.M. Harwani, M. Ayouty, E. Zweck, Dr Kapur) for any discrepancies or major outliers and resolved with the submitting site.
Study Population
Between 2016 and 2019, data from 1565 individual patient hospital admissions with a diagnosis of CS were collected. Proper Institutional Review Board approval was obtained to access this data from medical records, and patient consent was not required. CS cause was reported by each site as due to MI, HF, or other. MI was defined as any primary diagnosis of either non–ST-segment–elevation MI or ST-segment–elevation MI. HF was defined as any primary diagnosis of acute on chronic HF, not otherwise related to MI. Other causes included postcardiotomy, myocarditis, or not otherwise specified CS. Patients under the age of 18 years (n=1, 0.06%) and those with unknown mortality status at the time of hospital discharge (n=150, 9.6%) were excluded leaving a study population of 1414 patients with CS from 8 hospitals for analysis.
SCAI Classification
Patients were stratified according to the maximum SCAI classification stage reached during hospitalization to assess CS severity compared with in-hospital mortality.11 According to the SCAI definition of stages, clinical deterioration based on persistent hypotension and hypoperfusion is the main determinant of a patient’s SCAI stage and is associated with a need for intensification of treatment. Therefore, treatment escalation during hospitalization for CS was used as a proxy for persistent hypotension and hypoperfusion to retrospectively define maximum deterioration since hemo-metabolic parameters were only assessed at admission. A CSWG-adapted definition of SCAI stages was applied in our study cohort based on total use of vasopressors, inotropes, and MCS across a patient’s hospital stay as follows (Figure 1): SCAI defines stage A patients as those at risk for CS and stage A was, therefore, not captured in our study population. Stage B patients are those exhibiting early symptoms not including hypoperfusion and, therefore, do not require pharmacological or mechanical support. Stage C patients are those with hypoperfusion requiring initial intervention with up to either one drug or one MCS device. Stage D patients are those whose condition deteriorates despite initial intervention, defined in our data set by the need for additional drugs or MCS treatment. Finally, stage E patients are those who have deteriorated further and require maximal support, defined in our data set as requiring at least 2 MCS devices and 2 drugs during their hospitalization. While timing of maximal vasopressor/inotrope treatment is not known in comparison to the timing of device treatment, each progression of treatment is considered a form of escalation and therefore, deterioration as defined by SCAI, so can be assessed independently when assigning maximal patient SCAI stage.
Figure 1. Definition, retrospective adjudication, and distribution of Society for Cardiovascular Angiography and Interventions (SCAI) stages within the Cardiogenic Shock Working Group (CSWG) registry. CS indicates cardiogenic shock.
A sensitivity analysis incorporating lactate into SCAI stage definitions was performed. Stage B was defined as having a baseline lactate <2 meq/L and having received no drugs or devices throughout hospitalization; stage C was defined by a baseline lactate <5 meq/L and having received either 1 drug or 1 device; stage D patients had a baseline lactate < 5 meq/L but received >1 drug or 1 device; and stage E patients were defined by a baseline lactate of ≥5 (Figure I in the Data Supplement).
Hemodynamic Congestion and Clinical Outcomes
Hemodynamic associations with mortality and their distribution across SCAI stages were evaluated according to 4 specific profiles of congestion. Pulmonary capillary wedge pressure (PCWP) was considered elevated at ≥18 mm Hg and right atrial pressure (RAP) was considered elevated when ≥12 mm Hg. Values of RAP and PCWP in excess of these upper limits were used to stratify patients into 1 of the following 4 congestion profiles: right-ventricular (RV) (elevated RAP) congestion, left-ventricular (LV) (elevated PCWP) congestion, bi-sided (BiV, both RAP and PCWP elevated) congestion, or euvolemic (EuV, both RAP and PCWP below cutoff values).
Statistical Analyses
The primary outcome of interest was all-cause, in-hospital mortality; all mortality outcomes analyzed in this report refer exclusively to in-hospital mortality. Secondary analyses explored descriptive statistics comparing characteristics and outcomes of SCAI stages and congestion profiles (as described above). All analyses were performed on an all-cause CS cohort, an MI CS subcohort, and an HF-CS subcohort. Univariate logistic regression models were used to estimate odds and 95% CIs of mortality in association with SCAI stages and congestion profile. Multivariate analyses were then performed, adjusting by significant comorbidities, to assess the independent associations of SCAI, congestion profile, and shock cause. Descriptive statistics for categorical variables were reported as percentages and compared by χ2 tests, and continuous variables were reported as means with standard deviations and were compared using t tests or ANOVA as appropriate to report P values with a significance level of α=0.05.
Results
Patient Characteristics
Data from a total of 1414 patient hospitalizations for CS were analyzed. Patient characteristics are summarized in Table 1. Mean CI in the total cohort and across each SCAI subcohort ranged between 1.8 and 1.9 L/(min·m2). The majority of the study population was male and White. From the total population, the primary cause of CS was identified as HF in 50.4% (n=712), MI in 34.9% (n=494), and other causes in 14.71% (n=208). Stage B, C, and D patients were also more commonly HF patients while stage E was primarily patients with MI. While short-term MCS devices were broadly represented in different treatment combinations among the overall study (Figure 2) cohort, intraaortic balloon pump was the most commonly used device in the overall cohort (n=770, 54.5%). This was also the case in stage C (n=121, 46.0%) and D (n=464, 61.2%) and ECMO devices were the most commonly used devices in stage E patients (n=154, 72.6%). Prior percutaneous coronary intervention (PCI), hypertension, elevated AST, elevated lactate, and elevated filling pressures were also more common among stage E patients. Characteristics of shock cause sub-cohorts are presented in Table 2. Compared with patients with HF-CS, patients with MI were older with higher lactate and lower serum creatinine levels. Additionally, left-sided ejection fraction was higher among patients with MI, and mean pulmonary arterial pressure was lower. No differences in cardiac filling pressures, cardiac output (CO), mean arterial pressure, or heart rate were noted between the HF and MI cohorts. Characteristics of congestion sub-cohorts are presented in Table 3.
| All (N=1414) | SCAI Stage | P Value | ||||
|---|---|---|---|---|---|---|
| B (n=46) | C (n=263) | D (n=758) | E (n=212) | |||
| n (%) | n (%) | n (%) | n (%) | n (%) | ||
| Nonsurvivors | 431 (30.4) | 0 (0) | 28 (10.7) | 250 (33.0) | 117 (55.2) | <0.001 |
| Male | 1025 (72.5) | 33 (71.7) | 199 (75.7) | 540 (71.2) | 155 (73.1) | 0.58 |
| Shock cause | ||||||
| MI | 494 (34.9) | 2 (4.4) | 81 (30.8) | 244 (32.3) | 130 (61.32) | <0.001 |
| HF | 712 (50.4) | 40 (87.0) | 149 (56.7) | 432 (57.2) | 55 (25.9) | |
| Other | 208 (14.7) | 4 (8.7) | 33 (12.6) | 79 (10.5) | 27 (12.7) | |
| No. of pressors/inotropes | <0.001 | |||||
| 0 | 236 (16.7) | 46 (100.0) | 171 (65.0) | 19 (2.5) | 0 (0) | |
| 1 | 393 (27.8) | 0 (0) | 92 (35.0) | 301 (39.7) | 0 (0) | |
| 2+ | 650 (46.0) | 0 (0) | 0 (0) | 438 (57.8) | 212 (100.0) | |
| No. of devices | <0.001 | |||||
| 0 | 224 (15.8) | 46 (100.0) | 92 (35.0) | 61 (8.1) | 0 (0) | |
| 1 | 882 (62.4) | 0 (0) | 171 (65.0) | 620 (81.8) | 0 (0) | |
| 2+ | 308 (21.8) | 0 (0) | 0 (0) | 77 (10.2) | 212 (100.0) | |
| Type of MCS | ||||||
| Impella | 410 (29.0) | 0 (0) | 38 (14.5) | 186 (24.5) | 137 (64.62) | <0.001 |
| ECMO | 333 (23.6) | 0 (0) | 12 (4.6) | 127 (16.8) | 154 (72.6) | <0.001 |
| IABP | 770 (54.5) | 0 (0) | 121 (46.0) | 464 (61.2) | 145 (68.4) | <0.001 |
| Race | 0.002 | |||||
| White | 647 (45.8) | 32 (69.6) | 152 (57.8) | 306 (40.4) | 98 (46.3) | |
| Hispanic/Latino | 31 (2.2) | 1 (2.2) | 9 (3.4) | 13 (1.7) | 3 (1.4) | |
| Black | 31 (2.2) | 0 (0) | 2 (0.8) | 15 (2.0) | 3 (1.4) | |
| Asian | 28 (2.0) | 0 (0) | 8 (3.0) | 11 (1.5) | 7 (3.3) | |
| Other | 82 (5.8) | 13 (28.3) | 19 (7.2) | 44 (5.8) | 4 (1.9) | |
| Medical history | ||||||
| HTN | 681 (48.2) | 12 (26.1) | 118 (44.9) | 380 (50.1) | 115 (54.3) | <0.001 |
| DM2 | 489 (34.6) | 11 (23.9) | 87 (33.1) | 262 (34.6) | 89 (42.0) | 0.06 |
| Afib/flutter | 296 (20.9) | 14 (30.4) | 49 (18.6) | 168 (22.2) | 46 (21.7) | 0.08 |
| CKD (any stage) | 323 (22.8) | 14 (30.4) | 64 (24.3) | 182 (24.0) | 34 (16.0) | 0.24 |
| PVD | 60 (4.2) | 1 (2.2) | 12 (4.6) | 33 (4.4) | 10 (4.7) | 0.55 |
| COPD | 101 (7.1) | 6 (13.0) | 16 (6.1) | 56 (7.4) | 16 (7.6) | 0.55 |
| CVA/TIA | 159 (11.2) | 4 (8.7) | 28 (10.7) | 101 (13.3) | 15 (7.1) | 0.01 |
| Valvular disease | 214 (15.1) | 12 (26.1) | 51 (24.5) | 126 (25.8) | 19 (15.1) | 0.09 |
| PCI | 293 (20.7) | 11 (23.9) | 43 (21.6) | 136 (29.7) | 87 (44.9) | <0.001 |
| CABG | 114 (8.1) | 3 (6.5) | 16 (7.3) | 64 (11.5) | 21 (10.5) | 0.29 |
| VT | 216 (15.3) | 11 (23.9) | 39 (18.5) | 107 (20.7) | 45 (32.6) | 0.01 |
| ICD | 329 (23.3) | 23 (50.0) | 71 (34.0) | 173 (33.5) | 42 (30.7) | 0.11 |
| CRT | 97 (6.9) | 7 (15.2) | 13 (6.2) | 65 (12.6) | 10 (7.3) | 0.03 |
| All (N=1414) | SCAI Stage | P Value | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| B (n=46) | C (n=263) | D (n=758) | E (n=212) | |||||||||||||
| n | Mean | SD | n | Mean | SD | n | Mean | SD | n | Mean | SD | n | Mean | SD | ||
| Demographic | ||||||||||||||||
| Age | 1412 | 59.9 | 14.8 | 46 | 54.6 | 16.0 | 263 | 60.5 | 15.2 | 758 | 60.5 | 14.6 | 212 | 57.6 | 13.6 | 0.004 |
| Weight, kg | 1138 | 85.3 | 22.6 | 46 | 87.0 | 20.2 | 231 | 87.4 | 22.2 | 589 | 83.7 | 23.6 | 155 | 86.1 | 19.9 | 0.15 |
| Metabolic | ||||||||||||||||
| AST | 788 | 459.4 | 1492.6 | 37 | 32.0 | 19.9 | 124 | 153.5 | 547.8 | 424 | 355.6 | 1168.9 | 174 | 1023.4 | 2446.2 | <0.001 |
| BUN | 1026 | 32.4 | 20.5 | 46 | 28.6 | 16.6 | 196 | 29.8 | 19.9 | 538 | 33.6 | 21.2 | 199 | 33.2 | 20.9 | 0.08 |
| Lactate | 676 | 4.4 | 4.2 | 1 | 1.4 | 0 | 62 | 4.2 | 4.0 | 401 | 3.7 | 3.8 | 165 | 5.9 | 4.9 | <0.001 |
| HCO3 | 836 | 22.1 | 5.4 | 44 | 25.7 | 3.0 | 159 | 23.8 | 4.8 | 444 | 22.0 | 5.3 | 170 | 20.0 | 5.9 | <0.001 |
| Serum creatinine | 1295 | 1.8 | 1.1 | 46 | 1.3 | 0.4 | 248 | 1.5 | 0.8 | 739 | 1.8 | 1.2 | 203 | 1.9 | 1.1 | <0.001 |
| pH | 577 | 7.3 | 0.2 | 2 | 7.4 | 0.1 | 51 | 7.3 | 0.1 | 312 | 7.3 | 0.1 | 168 | 7.3 | 0.1 | 0.18 |
| Hemodynamic | ||||||||||||||||
| Admission EF, % | 771 | 24.9 | 15.5 | 1 | 65.0 | 0 | 126 | 28.1 | 16.5 | 490 | 24.2 | 15.0 | 111 | 24.1 | 17.1 | 0.005 |
| RAP | 1037 | 14.2 | 6.9 | 44 | 8.8 | 6.2 | 177 | 12.9 | 6.7 | 619 | 14.3 | 6.9 | 165 | 16.2 | 6.3 | <0.001 |
| PCWP | 847 | 24.5 | 8.9 | 45 | 16.5 | 7.3 | 177 | 24.3 | 8.3 | 473 | 25.2 | 8.8 | 131 | 24.6 | 9.2 | <0.001 |
| Mean PAP | 904 | 32.8 | 9.8 | 44 | 27.0 | 11.3 | 178 | 33.3 | 9.5 | 646 | 33.5 | 9.6 | 169 | 30.8 | 9.4 | <0.001 |
| CO | 1062 | 3.8 | 2.4 | 45 | 3.8 | 0.7 | 188 | 3.5 | 1.2 | 651 | 3.8 | 2.4 | 153 | 4.4 | 3.6 | 0.003 |
| CPO | 999 | 0.6 | 0.4 | 45 | 0.6 | 0.1 | 178 | 0.6 | 0.3 | 607 | 0.6 | 0.4 | 146 | 0.7 | 0.6 | 0.44 |
| Heart rate | 1248 | 92.0 | 22.7 | 46 | 75.2 | 12.8 | 234 | 85.9 | 19.7 | 685 | 93.8 | 22.2 | 193 | 97.3 | 24.8 | <0.001 |
| Cardiac index | 1071 | 1.9 | 0.6 | 45 | 1.9 | 0.3 | 191 | 1.8 | 0.5 | 659 | 1.9 | 0.6 | 151 | 1.9 | 0.6 | 0.09 |
| MAP | 1230 | 74.5 | 14.7 | 46 | 71.8 | 7.5 | 250 | 80.3 | 15.7 | 724 | 74.4 | 14.3 | 205 | 67.8 | 12.9 | <0.001 |
| Overall (N=1414) | Shock Cause | P Value | |||||
|---|---|---|---|---|---|---|---|
| MI (N=494) | HF (N=712) | ||||||
| n | % | n | % | n | % | ||
| SCAI stage | <0.001 | ||||||
| B | 1 | 0.1 | 0 | 0 | 1 | 0.34 | |
| C | 232 | 16.4 | 75 | 26.0 | 139 | 47.9 | |
| D | 220 | 15.6 | 114 | 39.6 | 85 | 29.3 | |
| E | 192 | 13.6 | 99 | 34.4 | 65 | 22.4 | |
| No. of pressors/inotropes | <0.001 | ||||||
| 0 | 236 | 16.7 | 86 | 18.8 | 119 | 17.6 | |
| 1 | 393 | 27.8 | 115 | 25.2 | 241 | 35.7 | |
| 2+ | 650 | 46.0 | 256 | 56.0 | 316 | 46.8 | |
| No. of MCS devices | <0.001 | ||||||
| 0 | 224 | 15.8 | 21 | 4.3 | 161 | 22.6 | |
| 1 | 882 | 62.4 | 294 | 59.5 | 465 | 65.3 | |
| 2+ | 308 | 21.8 | 179 | 36.2 | 86 | 12.1 | |
| Type of MCS | |||||||
| Impella | 410 | 29.0 | 210 | 42.5 | 148 | 20.8 | <0.001 |
| ECMO | 333 | 23.6 | 169 | 43.2 | 106 | 14.9 | <0.001 |
| IABP | 770 | 54.5 | 292 | 59.1 | 382 | 53.7 | 0.06 |
| Gender | 0.005 | ||||||
| Female | 387 | 27.4 | 153 | 31.0 | 169 | 23.7 | |
| Male | 1025 | 72.5 | 340 | 68.8 | 543 | 76.3 | |
| Race | <0.001 | ||||||
| White | 647 | 45.8 | 175 | 35.4 | 321 | 45.1 | |
| Hispanic/Latino | 31 | 2.2 | 16 | 3.2 | 7 | 1.0 | |
| Asian | 31 | 2.2 | 18 | 3.6 | 4 | 0.6 | |
| Black | 28 | 2.0 | 8 | 1.6 | 13 | 1.8 | |
| Other | 82 | 5.8 | 15 | 3.0 | 55 | 7.7 | |
| Medical history | |||||||
| HTN | 681 | 48.2 | 321 | 65.0 | 276 | 38.8 | <0.001 |
| DM | 489 | 34.6 | 220 | 44.5 | 222 | 31.2 | <0.001 |
| Afib/flutter | 296 | 20.9 | 37 | 7.5 | 227 | 31.9 | <0.001 |
| CKD (any stage) | 323 | 22.8 | 84 | 17.0 | 207 | 29.1 | <0.001 |
| PVD | 60 | 4.2 | 27 | 5.5 | 22 | 3.1 | 0.0177 |
| COPD | 101 | 7.1 | 27 | 5.5 | 62 | 8.7 | 0.0028 |
| CVA/TIA | 159 | 11.2 | 60 | 12.1 | 88 | 12.4 | 0.2023 |
| Valvular disease | 214 | 15.1 | 24 | 4.9 | 154 | 21.6 | <0.001 |
| PCI | 293 | 20.7 | 160 | 32.4 | 101 | 14.2 | <0.001 |
| CABG | 114 | 8.1 | 40 | 8.1 | 60 | 8.4 | 0.0135 |
| VT | 216 | 15.3 | 37 | 7.5 | 154 | 21.6 | <0.001 |
| ICD | 329 | 23.3 | 15 | 3.0 | 287 | 40.3 | <0.001 |
| CRT | 97 | 6.9 | 10 | 2.0 | 86 | 12.1 | <0.001 |
Figure 2. Device usage among Cardiogenic Shock Working Group (CSWG) patients with available hemodynamic data among the entire study cohort and each Society for Cardiovascular Angiography and Interventions (SCAI) stage. aMCS indicates acute mechanical circulatory support; ECMO, extracorporeal membrane oxygenation; and IABP, intraaortic balloon pump.
In-Hospital Outcomes
In-hospital mortality in the study cohort was 30.5%. In-hospital mortality was higher among patients with MI (39.5%) than HF patients (25.3%; P<0.0001; Table 4). Overall, survivors were younger and exhibited lower prevalence of arterial hypertension, type 2 diabetes mellitus, and prior coronary artery bypass grafting compared with nonsurvivors (Table 4). Clinical variables stratified by survivorship among patients with MI and HF are shown in Tables I and II in the Data Supplement. MI survivors were less likely to receive ventricular assist devices or heart transplant compared with HF survivors (Figure II in the Data Supplement).
| Overall (N=1414) | Shock Cause | P Value | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| MI (N=494) | HF (N=712) | |||||||||
| n | Mean | SD | n | Mean | SD | n | Mean | SD | ||
| Demographic | ||||||||||
| Age | 1412 | 59.9 | 14.8 | 493 | 64.9 | 12.8 | 712 | 57.9 | 14.1 | <0.001 |
| Weight (kg) | 1138 | 85.3 | 22.6 | 403 | 83.2 | 19.5 | 534 | 86.5 | 24.6 | 0.027 |
| Metabolic | ||||||||||
| AST | 788 | 459.4 | 1492.6 | 345 | 448.5 | 1066.3 | 328 | 441.4 | 1805.2 | 0.950 |
| BUN | 1026 | 32.4 | 20.5 | 416 | 28.3 | 17.8 | 456 | 37.7 | 22.6 | <0.001 |
| Lactate | 676 | 4.4 | 4.2 | 292 | 4.7 | 4.1 | 307 | 3.8 | 4.1 | 0.011 |
| HCO3 | 836 | 22.1 | 5.4 | 367 | 20.2 | 4.9 | 330 | 24.3 | 5.3 | <0.001 |
| Serum creatinine | 1295 | 1.8 | 1.1 | 448 | 1.7 | 1.2 | 687 | 1.9 | 1.1 | 0.003 |
| pH | 577 | 7.3 | 0.2 | 306 | 7.3 | 0.2 | 179 | 7.3 | 0.1 | <0.001 |
| Hemodynamic | ||||||||||
| Admission EF | 771 | 24.9 | 15.5 | 260 | 30.9 | 15.9 | 429 | 20.2 | 12.4 | <0.001 |
| RAP | 1037 | 14.2 | 6.9 | 303 | 14.6 | 6.5 | 626 | 14.0 | 7.2 | 0.176 |
| PCWP | 847 | 24.5 | 8.9 | 271 | 24.2 | 9.2 | 486 | 24.8 | 8.8 | 0.354 |
| Mean PAP | 904 | 32.8 | 9.8 | 257 | 30.2 | 9.0 | 549 | 34.4 | 9.8 | <0.001 |
| Cardiac output | 1062 | 3.8 | 2.4 | 329 | 3.8 | 2.1 | 630 | 3.8 | 2.4 | 0.826 |
| CPO | 999 | 0.6 | 0.4 | 314 | 0.6 | 0.4 | 584 | 0.6 | 0.4 | 0.638 |
| Heart rate | 1248 | 92.0 | 22.7 | 407 | 91.2 | 23.0 | 660 | 92.2 | 22.1 | 0.474 |
| Cardiac index | 1071 | 1.9 | 0.6 | 335 | 1.9 | 0.6 | 635 | 1.8 | 0.6 | 0.096 |
| MAP | 1230 | 74.5 | 14.7 | 433 | 74.9 | 16.9 | 628 | 74.0 | 12.7 | 0.483 |
Association of SCAI Stages With Outcomes
Patients with known drug and device data (n=1279) were classified into SCAI stages based on the number of drug and device treatments (Figure 1). Increasing drug or device treatment was directly associated with in-hospital mortality (Figure III in the Data Supplement, Table 4). All stage B patients survived to hospital discharge. Thereafter, each increased stage was associated with an increased risk of in-hospital mortality (Figure 3). Compared with SCAI stage C, stage D had 4.1 (95% CI, 2.7–6.3) times the odds of in-hospital mortality while stage E had 10.3 (95% CI, 6.4–16.6) times the odds of in-hospital mortality. Additionally, stage D had less than half the odds of in-hospital mortality of stage E (odds ratio [OR], 0.4 [95% CI, 0.29–0.55]). This was also true among the MI cohort with mortality ORs of 3.9 (95% CI, 2.0–7.6) and 8.1 (95% CI, 4.0–16.3) among stage D and E patients, respectively, compared with stage C patients and an OR of 0.49 (95% CI, 0.32–0.75) among stage D patients compared with those in stage E. The same trend was also observed in patients with HF. HF stage D and E patients had ORs of 3.5 (95% CI, 2.0–6.1) and 10.0 (95% CI, 4.7–21.0) compared with stage C patients and stage D patients had 0.35× the odds of mortality (95% CI, 0.20–0.61) compared with stage E patients.
Figure 3. In-hospital mortality by Society for Cardiovascular Angiography and Interventions (SCAI) stage among different causes of shock. HF indicates heart failure; and MI, myocardial infarction. ***P≤0.001.
| Congestion Profile | P Value | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Euvolemic | Left Ventricular | Right Ventricular | Biventricular | ||||||
| n | % | n | % | n | % | n | % | ||
| Mortality | 24 | 16.9 | 35 | 18.8 | 23 | 34.9 | 143 | 36.9 | <0.001 |
| Shock cause | 0.01 | ||||||||
| MI | 39 | 27.5 | 45 | 24.2 | 29 | 43.9 | 111 | 28.6 | |
| HF | 88 | 62.0 | 120 | 64.5 | 25 | 37.9 | 238 | 61.5 | |
| Other | 15 | 10.6 | 21 | 11.3 | 12 | 18.2 | 38 | 9.8 | |
| SCAI stage | 0.06 | ||||||||
| B | 1 | 2.4 | 0 | 0 | 0 | 0 | 0 | 0 | |
| C | 14 | 33.3 | 29 | 43.9 | 9 | 22.0 | 75 | 35.9 | |
| D | 17 | 40.5 | 23 | 34.9 | 14 | 34.2 | 71 | 34.0 | |
| E | 10 | 23.8 | 14 | 21.2 | 18 | 43.9 | 63 | 30.1 | |
| No. of pressors/inotropes | <0.001 | ||||||||
| 0 | 40 | 29.0 | 25 | 13.6 | 7 | 10.8 | 60 | 15.9 | |
| 1 | 49 | 35.5 | 75 | 40.8 | 19 | 29.2 | 116 | 30.7 | |
| 2+ | 49 | 35.5 | 84 | 45.7 | 39 | 60.0 | 202 | 53.4 | |
| No. of devices | <0.001 | ||||||||
| 0 | 64 | 45.1 | 50 | 26.9 | 10 | 15.2 | 60 | 15.5 | |
| 1 | 56 | 39.4 | 104 | 55.9 | 30 | 45.5 | 233 | 60.1 | |
| 2+ | 22 | 15.5 | 32 | 17.2 | 26 | 39.4 | 95 | 24.5 | |
| Device type | |||||||||
| Impella | 29 | 20.4 | 48 | 25.8 | 24 | 36.4 | 116 | 30.0 | 0.06 |
| ECMO | 19 | 13.4 | 25 | 13.4 | 22 | 33.3 | 89 | 22.9 | <0.001 |
| IABP | 52 | 36.6 | 96 | 51.6 | 36 | 54.6 | 223 | 57.5 | <0.001 |
| Male | 99 | 69.7 | 141 | 75.8 | 45 | 68.2 | 267 | 68.8 | 0.35 |
| Race | 0.10 | ||||||||
| White | 79 | 71.8 | 105 | 80.8 | 38 | 84.4 | 183 | 77.2 | |
| Hispanic/Latino | 1 | 0.9 | 1 | 0.8 | 2 | 4.4 | 10 | 4.2 | |
| Black | 4 | 3.6 | 3 | 2.3 | 1 | 2.2 | 8 | 3.4 | |
| Asian | 3 | 2.7 | 1 | 0.8 | 2 | 4.4 | 9 | 3.8 | |
| Other | 23 | 20.9 | 20 | 15.4 | 2 | 4.4 | 27 | 11.4 | |
| Medical history | |||||||||
| HTN | 59 | 43.1 | 76 | 43.7 | 32 | 50.8 | 180 | 51.9 | 0.35 |
| DM2 | 41 | 29.1 | 57 | 30.7 | 22 | 33.9 | 141 | 36.4 | 0.33 |
| Afib/flutter | 33 | 27.1 | 51 | 33.3 | 18 | 33.3 | 123 | 40.2 | 0.07 |
| CKD (any stage) | 36 | 26.9 | 45 | 26.6 | 11 | 18.3 | 104 | 30.6 | 0.24 |
| PVD | 6 | 4.6 | 6 | 3.7 | 5 | 8.6 | 20 | 6.4 | 0.44 |
| COPD | 12 | 8.8 | 12 | 6.9 | 3 | 4.8 | 40 | 11.1 | 0.23 |
| CVA/TIA | 22 | 16.1 | 28 | 16.2 | 6 | 9.5 | 52 | 14.4 | 0.60 |
| Valvular disease | 25 | 21.2 | 39 | 26.5 | 16 | 29.6 | 75 | 26.3 | 0.61 |
| PCI | 25 | 20.8 | 43 | 29.5 | 23 | 42.6 | 71 | 25.7 | 0.02 |
| CABG | 9 | 6.8 | 13 | 8.2 | 8 | 13.8 | 32 | 10.4 | 0.39 |
| VT | 36 | 29.5 | 41 | 26.6 | 15 | 27.8 | 71 | 23.1 | 0.54 |
| ICD | 54 | 44.3 | 77 | 50.0 | 11 | 20.0 | 116 | 37.9 | <0.001 |
| CRT | 16 | 13.1 | 18 | 11.7 | 1 | 1.8 | 36 | 11.8 | 0.14 |
| Congestion Profile | P Value | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Euvolemic | Left Ventricular | Right Ventricular | Biventricular | ||||||
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||
| Demographic | |||||||||
| Age | 56.5 | 15.4 | 57.6 | 15.4 | 60.6 | 14.9 | 59.4 | 14.7 | 0.11 |
| Weight, kg | 81.6 | 20.1 | 82.5 | 20.3 | 83.2 | 22.8 | 87.1 | 22.1 | 0.03 |
| Metabolic | |||||||||
| AST | 452.3 | 1552.6 | 340.2 | 1067.9 | 860.9 | 3070.4 | 447.0 | 1447.0 | 0.29 |
| BUN | 28.3 | 15.7 | 31.2 | 19.2 | 32.2 | 22.1 | 37.6 | 22.8 | <0.001 |
| Lactate | 3.6 | 3.3 | 3.7 | 4.5 | 5.1 | 3.4 | 4.7 | 3.4 | 0.13 |
| HCO3 | 23.7 | 4.9 | 23.9 | 5.2 | 20.0 | 5.5 | 22.2 | 5.7 | <0.001 |
| Serum creatinine | 1.6 | 0.8 | 1.5 | 1.0 | 1.8 | 1.5 | 1.9 | 1.2 | 0.002 |
| pH | 7.3 | 0.1 | 7.4 | 0.1 | 7.3 | 0.1 | 7.3 | 0.1 | <0.001 |
| Hemodynamic | |||||||||
| EF, % | 22.6 | 12.2 | 24.0 | 16.8 | 25.2 | 16.8 | 23.5 | 15.2 | 0.85 |
| RAP | 6.7 | 3.1 | 8.7 | 2.9 | 16.7 | 3.5 | 19.2 | 5.1 | <0.001 |
| PCWP | 13.2 | 3.9 | 25.6 | 5.3 | 15.8 | 2.4 | 29.3 | 7.4 | <0.001 |
| Mean PAP | 23.3 | 6.8 | 33.6 | 6.9 | 25.8 | 8.3 | 37.3 | 9.0 | <0.001 |
| Cardiac output | 4.0 | 1.6 | 3.6 | 0.9 | 4.2 | 2.9 | 3.9 | 3.3 | 0.40 |
| CPO | 0.6 | 0.3 | 0.6 | 0.2 | 0.7 | 0.5 | 0.6 | 0.5 | 0.57 |
| Heart rate | 83.8 | 19.0 | 90.2 | 19.2 | 91.6 | 23.9 | 93.8 | 21.4 | <0.001 |
| Cardiac index | 2.0 | 0.5 | 1.9 | 0.5 | 1.9 | 0.6 | 1.8 | 0.6 | <0.001 |
| Mean arterial pressure | 73.2 | 13.6 | 74.7 | 13.6 | 71.9 | 19.1 | 73.7 | 13.1 | 0.53 |
Lactate and drug and device data were available in 645 patients and were used to perform a sensitivity analysis of SCAI staging incorporating lactate levels. Of these 645 patients, 1 patient (0.1%) was classified as stage B, 232 (35.9%) were stage C, 220 (34.1%) were stage D, and 192 (29.8%) were stage E. This distribution differed significantly from the entire study cohort using only drug and device escalation. However, a similar trend in mortality was observed in this sensitivity analysis with 0% mortality in stage B, 32.3% in stage C, 48.6% in stage D, and 57.3% in stage E (Figure I in the Data Supplement).
Association of Congestion Profiles With Outcomes
We next explored the impact of hemodynamic congestion on mortality. Pulmonary artery catheters were used to collect any hemodynamic data in 79% of the total study population (n=1116) with both RAP and PCWP assessed in 55% of the total population (n=781). Mean CI was 1.9±0.6 across the study population. A positive correlation between RAP and PCWP (R2=0.26, P<0.001) was observed in these patients. Using these data, we grouped patients with CS into one of 4 congestion profiles as defined above (Figure 4). BiV congestion (ie, elevated right and left heart filling pressures) was most commonly observed (50%, n=390). Both BiV and right-sided congestion profiles were associated with the highest in-hospital mortality among the total cohort and among either MI or HF subgroups (Figure 5). Stage B patients were comprised mainly of euvolemic patients. The frequency of BiV congestion increased with increased SCAI stage among the entire cohort and in the MI and HF subgroups (Figure 5).
Figure 4. Adjudication and distribution of congestion profiles among the Cardiogenic Shock Working Group (CSWG) study population with available hemodynamic data. BiV indicates bi-ventricular; EuV, euvolemic; LV, left-ventricular; PCWP, pulmonary capillary wedge pressure; RAP, right arterial pressure; and RV, right-ventricular.
Figure 5. Congestion profiles as indicators of mortality and cardiogenic shock severity.A, Comparisons of mortality across congestion profiles among the overall, myocardial infarction (MI), and heart failure (HF) study cohorts. Comparisons adjusted by Bonferroni. B. Distribution of congestion profiles across Society for Cardiovascular Angiography and Interventions (SCAI) stages. BiV indicates bi-ventricular; EuV, euvolemic; LV, left-ventricular; and RV, right-ventricular. *P<0.05, ***P<0.001.
| Mortality | P Value | ||||
|---|---|---|---|---|---|
| Survivors (n=938) | Nonsurvivors (n=431) | ||||
| n | % | n | % | ||
| SCAI stage | <0.001 | ||||
| B | 1 | 0.3 | 0 | 0 | |
| C | 157 | 44.5 | 75 | 25.7 | |
| D | 113 | 32.0 | 107 | 36.6 | |
| E | 82 | 23.2 | 110 | 37.7 | |
| No. of pressors/inotropes | <0.001 | ||||
| 0 | 206 | 23.3 | 30 | 7.6 | |
| 1 | 305 | 34.5 | 88 | 22.3 | |
| 2+ | 373 | 42.2 | 277 | 70.1 | |
| No. of MCS devices | <0.001 | ||||
| 0 | 197 | 20.0 | 27 | 6.3 | |
| 1 | 631 | 64.2 | 251 | 58.2 | |
| 2+ | 155 | 15.8 | 153 | 35.5 | |
| Types of MCS | |||||
| Impella | 218 | 22.2 | 192 | 44.6 | <0.001 |
| ECMO | 168 | 17.1 | 165 | 38.3 | <0.001 |
| IABP | 560 | 57.0 | 210 | 48.7 | 0.004 |
| Cause | <0.001 | ||||
| MI | 299 | 30.4 | 195 | 45.2 | |
| HF | 532 | 54.1 | 180 | 41.8 | |
| Gender | 0.236 | ||||
| Female | 260 | 26.5 | 127 | 29.5 | |
| Male | 722 | 73.5 | 303 | 70.3 | |
| Race | 0.259 | ||||
| White | 460 | 46.8 | 187 | 43.4 | |
| Hispanic/Latino | 20 | 2.0 | 11 | 2.6 | |
| Asian | 24 | 2.4 | 7 | 1.6 | |
| Black | 18 | 1.8 | 10 | 2.3 | |
| Other | 66 | 6.7 | 16 | 3.7 | |
| Medical history | |||||
| HTN | 426 | 43.3 | 255 | 59.2 | <0.001 |
| Diabetes mellitus | 310 | 31.5 | 179 | 41.5 | <0.001 |
| Afib/flutter | 207 | 21.1 | 89 | 20.6 | 0.462 |
| CKD (any stage) | 218 | 22.2 | 105 | 24.4 | 0.339 |
| PVD | 37 | 3.8 | 23 | 5.3 | 0.040 |
| COPD | 68 | 6.9 | 33 | 7.7 | 0.883 |
| CVA/TIA | 109 | 11.1 | 50 | 11.6 | 0.682 |
| Valvular disease | 161 | 16.4 | 53 | 12.3 | 0.296 |
| PCI | 187 | 19.0 | 106 | 24.6 | 0.040 |
| CABG | 59 | 6.0 | 55 | 12.8 | <0.001 |
| VT | 143 | 14.5 | 73 | 16.9 | 0.049 |
| ICD | 250 | 25.4 | 79 | 18.3 | 0.020 |
| CRT | 69 | 7.0 | 28 | 6.5 | 0.985 |
| Mortality | P Value | ||||||
|---|---|---|---|---|---|---|---|
| Survivors (n=938) | Nonsurvivors (n=431) | ||||||
| n | Mean | SD | n | Mean | SD | ||
| Demographic | |||||||
| Age | 982 | 58.3 | 15.0 | 430 | 63.6 | 13.5 | <0.001 |
| Weight, kg | 803 | 85.3 | 22.8 | 335 | 85.2 | 22.1 | 0.970 |
| Metabolic | |||||||
| AST | 526 | 364.1 | 1324.0 | 262 | 650.8 | 1771.0 | 0.011 |
| BUN | 703 | 30.3 | 18.7 | 323 | 37.0 | 23.3 | <0.001 |
| Lactate | 377 | 3.6 | 3.4 | 299 | 5.4 | 4.9 | <0.001 |
| HCO3 | 576 | 23.2 | 5.2 | 260 | 19.8 | 5.3 | <0.001 |
| Serum creatinine | 907 | 1.7 | 1.0 | 388 | 2.0 | 1.3 | <0.001 |
| pH | 336 | 7.3 | 0.1 | 241 | 7.3 | 0.2 | <0.001 |
| Hemodynamic | |||||||
| Admission EF | 522 | 24.6 | 15.2 | 249 | 25.7 | 16.1 | 0.335 |
| RAP | 747 | 13.2 | 6.5 | 290 | 16.6 | 7.4 | <0.001 |
| PCWP | 595 | 24.0 | 8.9 | 252 | 25.6 | 8.8 | 0.018 |
| Mean PAP | 665 | 32.8 | 9.8 | 239 | 32.9 | 10.0 | 0.899 |
| Cardiac output | 747 | 3.7 | 2.0 | 315 | 4.1 | 3.2 | <0.001 |
| CPO | 704 | 0.6 | 0.4 | 295 | 0.6 | 0.5 | 0.381 |
| Heart rate | 894 | 91.0 | 22.3 | 354 | 94.6 | 23.5 | 0.012 |
| Cardiac index | 760 | 1.8 | 0.6 | 311 | 1.9 | 0.7 | 0.120 |
| MAP | 849 | 76.3 | 14.2 | 381 | 70.6 | 15.2 | <0.001 |
Multivariate Analyses
To better understand the relationship between SCAI stages, shock cause, and hemodynamics with in-hospital mortality, we ran several multivariate analyses. In the entire study cohort, after adjusting for shock cause, congestion profile, and other comorbidities (hypertension, age, type 2 diabetes mellitus, prior PCI, and ventricular tachycardia [VT]), we found that SCAI stages were still a significant predictor of mortality with stage D and E patients having aORs of 11.8 (95% CI, 4.6–30.5) and 21.3 (95% CI, 7.7–59.0), respectively, compared with stage C patients and stage D patients having an adjusted odds ratio (aOR) of 0.6 (95% CI, 0.3–0.9) compared with stage E patients. After adjustment, shock cause was not a significant independent predictor of mortality while biventricular congestion remained a signifcant independent predictor of mortality compared with left ventricular congestion or no congestion (BiV versus LV aOR, 2.4 [95% CI, 1.4–3.7]; BiV versus euvolemic aOR, 2.1 [95% CI, 1.1–4.0]). Additionally, after adjusting for SCAI stage in a separate multivariable model, RAP remained a significant predictor of mortality (OR, 1.06 [95% CI, 1.03–1.08]).
Discussion
Using a large, multicenter registry inclusive of invasive hemodynamics and contemporary short-term MCS strategies, we identified that the proposed SCAI staging system is associated with in-hospital mortality among patients with CS due to HF and MI. Compared with HF, patients with MI have higher mortality, but MI survivors have a greater likelihood of recovery to discharge, with few patients bridging to durable ventricular assist devices or orthotopic heart transplantation. Given the availability of hemodynamic data, we confirmed a low mean CI in the study population and observed a high prevalence of both right- and left-sided (biventricular) congestion in the study population. Worsening congestion was associated with both increasing SCAI stages and in-hospital mortality. These findings address critical gaps in our understanding of CS by confirming not only that SCAI stages identify patients at risk for in-hospital mortality in a population that reflects contemporary clinical practice, but also that basic hemodynamic data may be used to further stratify risk among patients with CS.
We assigned SCAI stages based on the consensus statement parameters focused on treatment intensity, defined by the number of drug and device therapies used during admission for CS (Figure 1). Drug or device escalation were each directly associated with in-hospital mortality. This observation is particularly important given the broad range of short-term MCS devices included in the analysis and supports the need for future prospective studies exploring the utility of device-based CS algorithms. Progression from one SCAI classification to the next represents a deteriorating clinical course of CS as indicated by increasing intensity of medical and device-based therapies to stabilize a critically ill patient, ultimately ending with use of all resources at hand in stage E. Therefore, we employed a matrix of drug and device escalation and identified that SCAI stages are directly associated with in-hospital mortality.
To validate our method of assigning SCAI stages, a sensitivity analysis incorporating lactate levels was performed. While a different distribution of patients across SCAI stages was observed, the relationship between SCAI stage and mortality remained unchanged (Figure I in the Data Supplement). Patient characteristics for this alternate definition of SCAI stages are displayed in Table III in the Data Supplement. Since lactate levels are not uniformly collected, these observations suggest that stratifying patients based on maximal drug or device utilization may be a reasonable approach to defining SCAI stages for the purpose of data analysis. Clinically, a more uniform definition for SCAI stages is needed, and prospective registries should incorporate lactate levels in addition to measures of hypotension, hypoperfusion, and drug/device utilization.
Recent reports exploring the utility of SCAI stages have employed different definitions for each stage. Jentzer et al12 defined SCAI stages based on clinical indices of hypotension and hypoperfusion with inclusion of a change in lactate from admission to maximal value recorded as a marker of deterioration. This group showed a correlation with in-hospital mortality in a single-center database that largely focused on intraaortic balloon pump use with minimal exposure to other short-term MCS devices. Schrage et al13 defined SCAI stages based primarily on lactate levels. This single-center study also identified a correlation between SCAI stages and in-hospital mortality. Neither study included invasive hemodynamic data. Our findings now employ a multicenter registry inclusive of contemporary short-term MCS devices and provide new information derived from invasive hemodynamic data that support our distinct approach to assigning SCAI stages.
We observed 0% mortality in SCAI stage B patients, who represent patients with early-stage shock. Since our report evaluated maximal SCAI stage during a patient’s hospitalization, these patients did not progress into CS and a low mortality rate may be expected and is consistent with prior reports.12,13 Furthermore, in our sensitivity analysis incorporating lactate levels, SCAI stage B patients continued to have the lowest mortality rate of 0%. More study of this unique population of preshock patients is required.
Accordingly, our findings strengthen the proposed SCAI classification structure by (1) providing contemporary evidence that treatment escalation may be an objective means of defining deterioration irrespective of the cause of CS, (2) enabling future analyses to evaluate both escalation and de-escalation of therapies, and (3) informing the development of future registry and randomized clinical trials where different CS strategies can be tested in patient populations with similar expected outcomes.
A unique aspect of the CSWG Registry is the availability of invasive hemodynamic data for analysis. Across survivors and nonsurvivors, cardiac filling pressures were elevated and CO, CI and cardiac power output (CPO) were low. Nonsurvivors had higher filling pressures and no significant difference in CPO or CI compared with survivors. CPO and CI were also not significantly changed across SCAI stages, but CO was paradoxically higher among stage E patients and among nonsurvivors. This may reflect variability in how CO is calculated (ie, Fick or thermodilution method) and the impact of maximal drug and device treatment to increase CO in sicker patients (ie, stage E). Furthermore, CPO has been validated primarily in MI populations14 but is less well understood in HF populations, where low CO does not always correlate with low mean arterial pressure. The SHOCK trial also did not include multiple short-term MCS approaches. Our data suggest that CPO requires further validation in MI and HF shock populations treated with contemporary short-term MCS devices.
In contrast to CO measurements, cardiac filling pressures were consistently elevated across all shock cohorts. Both RAP and PCWP were significantly higher among nonsurvivors and increased across SCAI stages. We further characterized the impact of cardiac filling pressures on clinical outcomes by defining congestive profiles based on RAP and PCWP. In-hospital mortality was highest among patients with biventricular or right ventricular congestive profiles. Furthermore, the distribution of congestive profiles across SCAI stages suggests that sicker patients are more likely to have biventricular congestion. These findings suggest that venous congestion is potentially an important determinant of clinical outcomes and may be explained by the fact that venous congestion is associated with worsening renal function and congestive hepatopathy,15,16 which may exacerbate metabolic derangement. Prior reports have also illustrated the association between venous congestion and poor outcomes in HF and MI.17 These observations suggest that approaches to decongest patients with CS may improve clinical outcomes.
We further observed that the presence of biventricular congestion was associated with worsening kidney and liver function and elevated lactate levels compared with other congestive profiles. These data are also consistent with studies suggesting that the presence of right HF in the setting of CS is associated with increased mortality.18 Recent data from prospective shock registries using congestive profiles as part of a treatment strategy algorithm showed improvement in mortality due to acute MI and CS.19,20 These findings suggest that CS algorithms that include an assessment of congestive profile may lead to improved outcomes by identifying and managing patients with venous congestion before metabolic failure worsens.
Limitations
The retrospective nature of the registry limits the ability to account for missing data elements, is subject to clinical and selection bias, and further limits our ability to adjust for metabolic and hemodynamic indicators of prognosis. Since the exact timing of data collection cannot be ascertained, it would be inappropriate to assess these measures as confounders relative to each other. A limitation of the current analysis is the lack of detail regarding drug dosage, sequence of device application, and timing of therapy as well as specific vasopressors or inotropes used. Future studies specifically looking at drug and device escalation across a patient’s hospitalization for CS are required. Furthermore, information about cardiac arrest was not available for analysis. However, even without serial data available, maximal escalation of treatment serves a reasonable marker of overall clinical deterioration. Though, this approach prevents drawing inferences about treatment strategies at each SCAI stage and may be influenced by other factors including institutional availability of devices, physician preference, variations in shock treatment algorithms, and other clinical or anatomic limitations to drug or device implementation. For these reasons, an additional sensitivity analysis incorporating baseline lactate levels into the SCAI staging scheme was performed. As hemodynamic data in this study were assessed after index hospital admission, data was most likely acquired after initiation of drug or device therapy in the case of transfer patients. Future studies involving a more granular retrospective data set or prospective studies are required to put these findings into context.
Conclusions
In this large, multicenter analysis of a national registry, we provide new insight into the characteristics, contemporary treatment strategies, and predictors of in-hospital mortality among patients with all-cause CS or CS due to MI or HF. We provide real-world validation of the SCAI staging scheme as an approach to identify patients with CS at risk of in-hospital mortality. We also identified venous congestion as a critical marker of risk, thus potentially identifying an important target of therapy for patients with CS. Future prospective studies are required to confirm the long-term prognostic significance of these findings.
| BiV |
bi-ventricular |
| CI |
cardiac index |
| CO |
cardiac output |
| CS |
cardiogenic shock |
| CSWG |
Cardiogenic Shock Working Group |
| HF |
heart failure |
| MCS |
mechanical circulatory support |
| MI |
myocardial infarction |
| OR |
odds ratio |
| PCI |
percutaneous coronary intervention |
| PCWP |
pulmonary capillary wedge pressure |
| RAP |
right atrial pressure |
| SCAI |
Society for Cardiovascular Angiography and Interventions |
| VT |
ventricular tachycardia |
Sources of Funding
This work was supported by a National Institutes of Health RO1 grant to Dr Kapur (RO1HL139785-01) and institutional grants from Abiomed Inc (Danvers, MA), Boston Scientific Inc (Minneapolis, MN), and Abbott Laboratories (Abbott Park, IL) to Tufts Medical Center.
Disclosures
Dr Kapur receives consulting/speaker honoraria and institutional grant support from Abbott Laboratories, Abiomed Inc, Boston Scientific, Medtronic, LivaNova, MDStart, and Precardia. Dr Hernandez-Montfort is a consultant for Abiomed Inc. Dr Abraham is a consultant for Abbott Laboratories, Abiomed Inc. Dr Burkhoff reports an unrestricted, educational grant from Abiomed Inc. to Cardiovascular Research Foundation. Dr Sinha is a consultant for Abiomed Inc (Critical Care Advisory Board). Dr O’Neill receives consulting/speaker honoraria from Abiomed Inc. The other authors report no conflicts.
Footnotes
References
- 1.
Kolte D, Khera S, Aronow WS, Mujib M, Palaniswamy C, Sule S, Jain D, Gotsis W, Ahmed A, Frishman WH, . Trends in incidence, management, and outcomes of cardiogenic shock complicating ST-elevation myocardial infarction in the United States.J Am Heart Assoc. 2014; 3:e000590. doi: 10.1161/JAHA.113.000590LinkGoogle Scholar - 2.
van Diepen S, Katz JN, Albert NM, Henry TD, Jacobs AK, Kapur NK, Kilic A, Menon V, Ohman EM, Sweitzer NK, ; American Heart Association Council on Clinical Cardiology; Council on Cardiovascular and Stroke Nursing; Council on Quality of Care and Outcomes Research; and Mission: Lifeline. Contemporary management of cardiogenic shock: a scientific statement from the American Heart Association.Circulation. 2017; 136:e232–e268. doi: 10.1161/CIR.0000000000000525LinkGoogle Scholar - 3.
Goldberg RJ, Makam RC, Yarzebski J, McManus DD, Lessard D, Gore JM . Decade-long trends (2001-2011) in the incidence and hospital death rates associated with the in-hospital development of cardiogenic shock after acute myocardial infarction.Circ Cardiovasc Qual Outcomes. 2016; 9:117–125. doi: 10.1161/CIRCOUTCOMES.115.002359LinkGoogle Scholar - 4.
Becher PM, Schrage B, Sinning CR, Schmack B, Fluschnik N, Schwarzl M, Waldeyer C, Lindner D, Seiffert M, Neumann JT, . Venoarterial extracorporeal membrane oxygenation for cardiopulmonary support.Circulation. 2018; 138:2298–2300. doi: 10.1161/CIRCULATIONAHA.118.036691LinkGoogle Scholar - 5.
Thiele H, Zeymer U, Neumann FJ, Ferenc M, Olbrich HG, Hausleiter J, Richardt G, Hennersdorf M, Empen K, Fuernau G, ; IABP-SHOCK II Trial Investigators. Intraaortic balloon support for myocardial infarction with cardiogenic shock.N Engl J Med. 2012; 367:1287–1296. doi: 10.1056/NEJMoa1208410CrossrefMedlineGoogle Scholar - 6.
Hochman JS, Sleeper LA, Webb JG, Sanborn TA, White HD, Talley JD, Buller CE, Jacobs AK, Slater JN, Col J, . Early revascularization in acute myocardial infarction complicated by cardiogenic shock. SHOCK Investigators. Should we emergently revascularize occluded coronaries for cardiogenic shock.N Engl J Med. 1999; 341:625–634. doi: 10.1056/NEJM199908263410901CrossrefMedlineGoogle Scholar - 7.
Ouweneel DM, Eriksen E, Sjauw KD, van Dongen IM, Hirsch A, Packer EJ, Vis MM, Wykrzykowska JJ, Koch KT, Baan J, . Percutaneous mechanical circulatory support versus intra-aortic balloon pump in cardiogenic shock after acute myocardial infarction.J Am Coll Cardiol. 2017; 69:278–287. doi: 10.1016/j.jacc.2016.10.022CrossrefMedlineGoogle Scholar - 8.
Heidenreich PA, Albert NM, Allen LA, Bluemke DA, Butler J, Fonarow GC, Ikonomidis JS, Khavjou O, Konstam MA, Maddox TM, ; American Heart Association Advocacy Coordinating Committee; Council on Arteriosclerosis, Thrombosis and Vascular Biology; Council on Cardiovascular Radiology and Intervention; Council on Clinical Cardiology; Council on Epidemiology and Prevention; Stroke Council. Forecasting the impact of heart failure in the United States: a policy statement from the American Heart Association.Circ Heart Fail. 2013; 6:606–619. doi: 10.1161/HHF.0b013e318291329aLinkGoogle Scholar - 9.
Strom JB, Zhao Y, Shen C, Chung M, Pinto DS, Popma JJ, Yeh RW . National trends, predictors of use, and in-hospital outcomes in mechanical circulatory support for cardiogenic shock.EuroIntervention. 2018; 13:e2152–e2159. doi: 10.4244/EIJ-D-17-00947CrossrefMedlineGoogle Scholar - 10.
Stretch R, Sauer CM, Yuh DD, Bonde P . National trends in the utilization of short-term mechanical circulatory support: incidence, outcomes, and cost analysis.J Am Coll Cardiol. 2014; 64:1407–1415. doi: 10.1016/j.jacc.2014.07.958CrossrefMedlineGoogle Scholar - 11.
Baran DA, Grines CL, Bailey S, Burkhoff D, Hall SA, Henry TD, Hollenberg SM, Kapur NK, O’Neill W, Ornato JP, . SCAI clinical expert consensus statement on the classification of cardiogenic shock: This document was endorsed by the American College of Cardiology (ACC), the American Heart Association (AHA), the Society of Critical Care Medicine (SCCM), and the Society of Thoracic Surgeons (STS) in April 2019.Catheter Cardiovasc Interv. 2019; 94:29–37. doi: 10.1002/ccd.28329MedlineGoogle Scholar - 12.
Jentzer JC, van Diepen S, Barsness GW, Henry TD, Menon V, Rihal CS, Naidu SS, Baran DA . Cardiogenic shock classification to predict mortality in the cardiac intensive care unit.J Am Coll Cardiol. 2019; 74:2117–2128. doi: 10.1016/j.jacc.2019.07.077CrossrefMedlineGoogle Scholar - 13.
Schrage B, Dabboura S, Yan I, Hilal R, Neumann JT, Sorensen NA, Gossling A, Becher PM, Grahn H, Wagner T, . Application of the SCAI classification in a cohort of patients with cardiogenic shock [published online ahead of print January 11, 2020]. Catheter Cardiovasc Interv.2020.Google Scholar - 14.
Fincke R, Hochman JS, Lowe AM, Menon V, Slater JN, Webb JG, LeJemtel TH, Cotter G ; SHOCK Investigators. Cardiac power is the strongest hemodynamic correlate of mortality in cardiogenic shock: a report from the SHOCK trial registry.J Am Coll Cardiol. 2004; 44:340–348. doi: 10.1016/j.jacc.2004.03.060CrossrefMedlineGoogle Scholar - 15.
Mullens W, Abrahams Z, Francis GS, Sokos G, Taylor DO, Starling RC, Young JB, Tang WHW . Importance of venous congestion for worsening of renal function in advanced decompensated heart failure.J Am Coll Cardiol. 2009; 53:589–596. doi: 10.1016/j.jacc.2008.05.068CrossrefMedlineGoogle Scholar - 16.
Samsky MD, Patel CB, DeWald TA, Smith AD, Felker GM, Rogers JG, Hernandez AF . Cardiohepatic interactions in heart failure: an overview and clinical implications.J Am Coll Cardiol. 2013; 61:2397–2405. doi: 10.1016/j.jacc.2013.03.042CrossrefMedlineGoogle Scholar - 17.
Brinkley DM, Ho KKL, Drazner MH, Kociol RD . The prognostic value of the relationship between right atrial and pulmonary capillary wedge pressure in diverse cardiovascular conditions.Am Heart J. 2018; 199:31–36. doi: 10.1016/j.ahj.2018.01.006CrossrefMedlineGoogle Scholar - 18.
Jacobs AK, Leopold JA, Bates E, Mendes LA, Sleeper LA, White H, Davidoff R, Boland J, Modur S, Forman R, . Cardiogenic shock caused by right ventricular infarction: a report from the SHOCK registry.J Am Coll Cardiol. 2003; 41:1273–1279. doi: 10.1016/s0735-1097(03)00120-7CrossrefMedlineGoogle Scholar - 19.
Tehrani BN, Truesdell AG, Sherwood MW, Desai S, Tran HA, Epps KC, Singh R, Psotka M, Shah P, Cooper LB, . Standardized team-based care for cardiogenic shock.J Am Coll Cardiol. 2019; 73:1659–1669. doi: 10.1016/j.jacc.2018.12.084CrossrefMedlineGoogle Scholar - 20.
Basir MB, Kapur NK, Patel K, Salam MA, Schreiber T, Kaki A, Hanson I, Almany S, Timmis S, Dixon S, ; National Cardiogenic Shock Initiative Investigators. Improved outcomes associated with the use of shock protocols: updates from the national cardiogenic shock initiative.Catheter Cardiovasc Interv. 2019; 93:1173–1183. doi: 10.1002/ccd.28307MedlineGoogle Scholar