The STS Score (Society of Thoracic Surgeons Risk Score) is a comprehensive, procedure-specific risk prediction model used to estimate the likelihood of adverse outcomes following cardiac surgery. Unlike EuroSCORE II which provides a general cardiac surgery risk estimate, STS Scores are highly specific to individual surgical procedures.

Purpose and Clinical Significance

Primary Objectives

• Risk Prediction: Estimate procedure-specific morbidity and mortality
• Quality Benchmarking: Enable hospital and surgeon performance comparison
• Informed Decision-Making: Facilitate patient-clinician discussions
• Resource Allocation: Guide preoperative planning and resource utilization

Key Differentiators from EuroSCORE II

• Procedure-specific models rather than a single general score
• Multiple outcome predictions (not just mortality)
• North American population-based (primarily US data)
• Continuous updates with contemporary data
• Public reporting for quality transparency

Components and Variables

The STS models incorporate 200+ variables across multiple domains:

Patient Demographics

• Age, gender, race/ethnicity
• Body mass index (BMI)
• Insurance status

Comorbidities

• Cardiac: Previous MI, heart failure, arrhythmias
• Vascular: Peripheral arterial disease, cerebrovascular disease
• Respiratory: COPD, pulmonary hypertension
• Renal: Chronic kidney disease, dialysis
• Metabolic: Diabetes mellitus (with insulin dependence)
• Other: Immunosuppression, liver disease, cancer

Cardiac Status

• Left ventricular ejection fraction (LVEF)
• NYHA functional class
• CCS angina class
• Recent myocardial infarction
• Number of diseased vessels
• Presence of left main disease

Procedure-Specific Factors

• Urgency of surgery (elective, urgent, emergent, salvage)
• Surgical approach (sternotomy, minimally invasive)
• Concomitant procedures
• Previous cardiac interventions

Major Procedure-Specific Models

1. STS Adult Cardiac Surgery Database Models

• Isolated CABG (Coronary Artery Bypass Grafting)
• Aortic Valve Replacement (AVR)
• Mitral Valve Repair/Replacement
• Combined CABG + Valve Procedures
• Thoracic Aortic Surgery

2. Predicted Outcomes for Each Procedure

• Operative Mortality (primary endpoint)
• Stroke (permanent neurological deficit)
• Renal Failure (requiring dialysis)
• Prolonged Ventilation (>24 hours)
• Deep Sternal Wound Infection
• Reoperation for any reason
• Short/Long Length of Stay
• Composite Morbidity & Mortality

Scoring Methodology

Statistical Approach

• Logistic Regression Models for each outcome
• Bayesian Modeling for some complex predictions
• Continuous Recalibration with new data
• Procedure-Specific Coefficients for each variable

Risk Stratification

• Low Risk: <1% mortality (CABG)
• Medium Risk: 1-3% mortality
• High Risk: 3-8% mortality
• Very High Risk: >8% mortality

Clinical Interpretation

Output Presentation

• Percentage probabilities for each outcome
• Risk-adjusted observed/expected ratios for institutions
• Graphical displays for patient counseling
• Comparative benchmarks against national averages

Example STS Score Report

text

Procedure: Isolated CABG
Predicted Outcomes:
- Operative Mortality: 1.2% (National Average: 1.8%)
- Stroke: 0.8%
- Renal Failure: 2.1%
- Prolonged Ventilation: 5.3%
- Deep Sternal Infection: 0.5%
- Composite Morbidity/Mortality: 8.7%

Clinical Applications

Preoperative Planning

1. Patient Selection: Identify appropriate surgical candidates
2. Procedure Timing: Determine optimal timing (elective vs. urgent)
3. Surgical Approach: Guide choice of conventional vs. minimally invasive
4. Resource Planning: Anticipate ICU needs, specialized equipment

Quality Improvement

1. Hospital Benchmarking: Compare outcomes against national standards
2. Surgeon Performance: Monitor individual surgeon results
3. Process Improvement: Identify areas for clinical pathway optimization
4. Public Reporting: Support transparency initiatives

Patient Communication

1. Risk Education: Provide objective risk estimates
2. Informed Consent: Support shared decision-making
3. Expectation Management: Set realistic postoperative expectations
4. Alternative Discussions: Consider non-surgical options when appropriate

Advantages

Compared to EuroSCORE II

1. Greater Specificity: Procedure-specific models
2. Multiple Outcomes: Predicts morbidity beyond just mortality
3. Contemporary Data: Regular updates with current practice patterns
4. US Population: More relevant for North American patients
5. Quality Focus: Directly linked to quality improvement initiatives

Technical Strengths

1. Large Database: >1 million procedures in STS National Database
2. Statistical Robustness: Extensive validation studies
3. Clinical Relevance: Developed by practicing cardiac surgeons
4. User-Friendly: Available through online calculators and EHR integration

Limitations

Methodological Constraints

1. US-Centric: Primarily validated in North American populations
2. Database Dependency: Requires accurate data entry
3. Procedure-Specific: Cannot compare across different procedure types
4. Static Assessment: Does not account for intraoperative findings

Clinical Limitations

1. Does Not Replace Judgment: Should complement clinical expertise
2. Limited Emergent Cases: Less accurate in true emergency situations
3. Frailty Not Incorporated: Does not include comprehensive geriatric assessment
4. Institutional Factors: Does not account for hospital volume or surgeon experience

Access and Implementation

Calculation Methods

1. Online Calculators: STS Risk Calculator website
2. EHR Integration: Built into electronic health record systems
3. Mobile Applications: Smartphone apps for point-of-care calculation
4. Manual Calculation: Complex formulas for each procedure (rarely used)

Data Requirements

• Complete patient demographic and clinical information
• Accurate comorbidity documentation
• Specific procedure details
• Institutional data submission for benchmarking

Future Directions

Ongoing Developments

1. New Procedure Models: Transcatheter valves, minimally invasive approaches
2. Long-Term Outcomes: Beyond 30-day mortality
3. Patient-Reported Outcomes: Quality of life measures
4. Machine Learning: Enhanced prediction algorithms
5. Global Adaptation: International validation studies

Integration Trends

1. Multidisciplinary Decision Support: Heart Team applications
2. Real-Time Risk Adjustment: Intraoperative updates
3. Patient-Facing Tools: Direct patient access to risk estimates
4. Value-Based Care: Linking risk prediction to cost and outcomes

Key Clinical Pearls

1. Procedure-Specific: Always use the model specific to the planned operation
2. Multiple Outcomes: Consider all predicted complications, not just mortality
3. Dynamic Tool: Recalculate if patient condition changes preoperatively
4. Complementary Use: Combine with EuroSCORE II for comprehensive assessment
5. Discussion Tool: Use percentages to facilitate shared decision-making

The STS Score represents the gold standard for procedure-specific risk prediction in cardiac surgery in North America, providing invaluable data for clinical decision-making, quality improvement, and patient counseling.