Comparison of the Predictive Performance of Cardiogenic Shock Scores in a Real-World Latin America Country

Abstract

Background: Mortality in cardiogenic shock (CS) is up to 40%, and although risk scores have been proposed to stratify and assess mortality in CS, they have been shown to have inconsistent performance. The purpose was to compare CS prognostic scores and describe their performance in a real-world Latin American country. Methods: We included 872 patients with CS. The Society for Cardiovascular Angiography and Interventions (SCAI), CARDSHOCK, IABP-Shock II, Cardiogenic Shock Score, age-lactate-creatinine score, Get-With-The-Guidelines Heart Failure score, and Acute Decompensated Heart Failure National Registry scores were calculated. Decision curve analyses were performed to evaluate the net benefit of the different scoring systems. Logistic and Cox regression analyses were applied to construct area under the curve (AUC) statistics, this last one against time using the Inverse Probability of Censoring Weighting method, for in-hospital mortality prediction. Results: When logistic regression was applied, the scores had a moderate-good performance in the overall cohort that was higher AUC in the CARDSHOCK (c = 0.666). In acute myocardial infarction-related CS (AMI-CS), CARDSHOCK still is the highest AUC (0.68). In non-AMI-CS only SCAI (0.668), CARDSHOCK (0.533), and IABP-SHOCK II (0.636) had statistically significant values. When analyzed over time, significant differences arose in the AUC, suggesting that a time-sensitive component influenced the prediction of mortality. The highest AUC was for the CARDSHOCK score (0.658), followed by SCAI (0.622). In AMI-CSrelated, the highest AUC was for the CARDSHOCK score (0.671). In non-AMI-CS, SCAI was the best (0.642). Conclusions: Clinical scores show a time-sensitive AUC, suggesting that performance could be influenced by time and the type of CS. Understanding the temporal influence on the scores could provide a better prediction and be a valuable tool in CS.

Clinical Characteristics, Management, and Outcomes in Cardiogenic Shock: Insights From a High-Volume Italian Cardiac Intensive Care Unit

“…The median LVEF was 25 [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35] % with a median left ventricular outflow tract velocity-time integral of 11 [9][10][11][12] cm and a median E/e' average ratio of 15 [13][14][15][16][17][18][19][20][21][22].…”

“…The hospital length of stay was 11.5 [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] days. During the hospital stay, 26 patients (27%) progressed to SCAI 1.…”

“…In brief, NSs were older [median age 75 (63-82) years vs. 69 (56-75) years; P = 0.019], had a higher prevalence of chronic obstructive pulmonary disease (COPD) [13 (29%) vs. 6 (12%), P = 0.036], presented higher levels of white blood cell at admission [13.4 (8.1-18.9) 10 9 /L vs. 9.5 (7.3-14.2) 10 9 /L, P = 0.019], CRP [60 (27.0-128.5) mg/L vs. 7.7 (2.8-31.0) mg/L, P , 0.001], procalcitonin [1.8 (0.4-5.5) ng/mL vs. 0.2 (0.05-0.31) ng/ mL, P , 0.001], N-terminal pro-BNP type natriuretic peptide (NT-pro BNP) [13,037 (8129-31,935) pg/mL vs. 9220 (4311-20,199) pg/mL, P = 0.043], and lactate [6 (3.0-9.8) mmol/L vs. 3 (1.8-4.8) mmol/L, P = 0.001] at admission, and showed a higher increase in lactate levels at 24-48 hours from admission (delta with baseline lactate levels) [+2.25 (21.5; 5) vs. 21.05 (22.8; 0.0), P = 0.001]. NSs had a worse SCAI stage at hospital admission (P , 0.001), a more depressed right ventricular (RV) function, assessed by tricuspid annular plane systolic excursion (TAPSE) [12 (10-15) mm vs. 18 (12)(13)(14)(15)(16)(17)(18)(19) mm, P , 0.001], and a higher estimated left ventricular filling pressure assessed by the average E/e 0 ratio [21 (18)(19)(20)(21)(22)(23)(24)(25) vs. 14 (11)(12)(13)(14)(15)(16)(17)(18), P , 0.001]. NSs had a higher CardShock risk score [5 (3.5-6) vs. 3 (2.8-4), P , 0.001].…”

Considerations About Scores in Cardiogenic Shock

“…To the Editor: We appreciate the work of Jorge Ortega-Hernández (1) in which he proposes to assess the predictive capacity of clinical-prognostic models in patients with cardiogenic shock. The results of his work reveal two pieces of information that we believe deserve further reflection:…”

Beyond One-Size-Fits-All: Exploring Predictive Models for Cardiogenic Shock

“…Dear Dr. Valiente Fernández and colleagues : We would like to express our gratitude for your interest in our work titled “Comparison of the Predictive Performance of Cardiogenic Shock Scores in a Real-World Latin America Country” (1) and for taking the time to provide insightful comments and feedback.…”