Cardiovascular Outcomes Registry in Cardiogenic SHOCK (COR-SHOCK) (COR-SHOCK)

The Cardiovascular Outcomes Registry in Cardiogenic SHOCK (COR-SHOCK) is hybrid (retrospective and prospective), observational registry designed to evaluate the clinical performance, therapeutic management, and outcomes of patients presenting with cardiogenic shock (CS).

Contemporary CS management has evolved beyond correcting macro-hemodynamic parameters. CS is now recognized as a highly dynamic, systemic syndrome characterized by microvascular hypoperfusion, systemic inflammatory response syndrome (SIRS), endothelial dysfunction, and profound neurohormonal and metabolic derangements. To capture this complexity, the COR-SHOCK registry systematically integrates granular, real-world data reflecting these pathophysiological axes. This includes laboratory evaluation, and advanced invasive hemodynamic profiling via pulmonary artery catheterization (Swan-Ganz).

Registry Procedures and Quality Assurance Plan

To ensure high-quality data collection and compliance with scientific standards, the registry implements the following procedures:

1. Quality Assurance (QA) Plan:

   The Steering Committee, led by the Principal Investigator, oversees the registry's operations. Internal data quality audits are conducted semi-annually to review enrollment rates, evaluate protocol adherence, and identify systematic data entry discrepancies.

2. Data Checks:

   Data is transcribed from electronic health records into a dedicated electronic database. The entry platform utilizes programmed range validation and consistency rules (e.g., preventing physiological out-of-range values for hemodynamics and laboratory results, and enforcing logical chronological order for clinical events, such as ensuring discharge dates succeed admission dates).

3. Source Data Verification (SDV):

   To guarantee data accuracy and completeness, a designated researcher who is not directly involved in the primary data entry performs random source data verification on 10% of all registry entries. This process involves cross-referencing database records with original paper and electronic health records.

4. Data Dictionary:

   A comprehensive data dictionary has been established, detailing every variable collected. It defines clinical terminology, laboratory reference ranges, and standardized categorization rules.

5. Standard Operating Procedures (SOPs):

   Formal SOPs govern all key registry processes, including:

   • Daily screening and identification of eligible patients in the cardiac intensive care unit (UNICOR).
   • Standardized timing for dynamic data collection.
   • Data security protocols, including patient pseudonymization and password-protected access control.
   • Event reporting protocols for predefined clinical complications.
6. Sample Size Assessment:

The registry aims to include approximately 750 to 800 patients. This target is calculated based on historical admission rates at the center, comprising a retrospective cohort of approximately 500 patients (2017-2025) and a prospective recruitment target of 80 to 100 patients annually over a 2-to-3-year period. This sample size provides sufficient statistical power to characterize temporal trends, evaluate subset populations (e.g., mechanical circulatory support, specific etiologies), and perform robust multivariable risk modeling.

Statistical Analysis Plan (SAP) Descriptive statistics will characterize the study population. Continuous variables will be presented as mean ± standard deviation (SD) or median with interquartile range (IQR), based on normality (assessed via Shapiro-Wilk test). Categorical variables will be expressed as frequencies and percentages.

To evaluate management and outcome trends over the years:

• Temporal trends in clinical management, therapeutic choices (such as the use of short-duration mechanical circulatory support [VA-ECMO, Impella, IABP] or advanced therapies [LVAD, cardiac transplantation]), and survival will be assessed using linear/logistic regression models across calendar years.
• Survival curves will be estimated using the Kaplan-Meier method and compared across different clinical phenotypes, therapeutic interventions, and time epochs using the log-rank test.
• Multivariable Cox proportional hazards regression and logistic regression models will be constructed to identify independent clinical, laboratory and hemodynamic predictors of short-term (ICU and in-hospital) and long-term (1 year) mortality. Model fit and calibration will be evaluated using standard diagnostic metrics (e.g., Harrell's C-index).