What is the role of RESP (Respiratory Extracorporeal Support Priority) and Murray scores in managing severe respiratory failure?

RESP and Murray Scores in Severe Respiratory Failure Management

RESP Score: Primary Tool for ECMO Candidate Selection

The RESP (Respiratory Extracorporeal Membrane Oxygenation Survival Prediction) score is a validated tool specifically designed to predict hospital survival in adult patients being considered for ECMO support for severe acute respiratory failure, with demonstrated discrimination (c-statistic 0.74) in the original cohort. 1

RESP Score Components and Calculation

The RESP score incorporates 12 pre-ECMO variables that independently predict hospital survival 1:

• Age (younger patients score higher)
• Immunocompromised status (negative predictor)
• Duration of mechanical ventilation before ECMO (shorter duration favored)
• Diagnosis (viral/bacterial pneumonia scores higher than other causes)
• Central nervous system dysfunction (negative predictor)
• Acute associated nonpulmonary infection (negative predictor)
• Neuromuscular blockade use (positive predictor)
• Nitric oxide use (positive predictor)
• Bicarbonate infusion (negative predictor)
• Pre-ECMO cardiac arrest (strongly negative predictor)
• PaCO2 levels
• Peak inspiratory pressure

RESP Score Risk Stratification

The RESP score classifies patients into five risk categories with corresponding survival predictions 1:

• Class I: Highest survival probability
• Class II-III: Intermediate survival probability
• Class IV-V: Lower survival probability (though still potentially acceptable for ECMO consideration)

Critical Limitations of RESP Score

The RESP score performs poorly in COVID-19 patients (AUC 0.61), with COVID-19 patients in RESP class I-III having worse survival than predicted and class IV-V having better survival than predicted. 2 This represents a major limitation when managing COVID-19-related ARDS 2.

For prolonged ECMO (≥14 days), the RESP score also demonstrates poor discrimination (c-statistic 0.540), suggesting it should not be the sole decision-making tool for patients anticipated to require extended support 3.

Murray Lung Injury Score: Limited Role in Modern Ventilation

The Murray Lung Injury Score (LIS) has become less reliable in the era of lung-protective ventilation strategies, as it incorporates therapy-dependent variables (tidal volume, PEEP, compliance) that change with modern low tidal volume ventilation. 4

Murray Score Components (Traditional)

The Murray LIS traditionally incorporates 4:

• Chest radiograph findings
• PaO2/FiO2 ratio
• PEEP level
• Respiratory system compliance

Why Murray Score Is Problematic

The Murray score is inferior to simple PaO2/FiO2 ratio assessment for predicting mortality (AUC 0.67 vs 0.74), because lung-protective ventilation strategies artificially alter the therapy-dependent components of the score. 4 In 71% of patients, the simplified P/F score was higher (more accurate) than the Murray LIS 4.

Practical Clinical Algorithm

For ECMO Candidacy Assessment:

1. Calculate RESP score using all 12 pre-ECMO variables for non-COVID-19 respiratory failure patients 1
2. Use RESP score cautiously in COVID-19 patients, recognizing its poor predictive value (AUC 0.61) 2
3. Consider additional factors beyond RESP score for prolonged ECMO candidates (≥14 days anticipated), as discrimination is poor (c-statistic 0.540) 3
4. Prioritize patients with: younger age, shorter pre-ECMO mechanical ventilation duration, viral/bacterial pneumonia, absence of immunosuppression, and no pre-ECMO cardiac arrest 1, 3

For Lung Injury Severity Assessment:

1. Use PaO2/FiO2 ratio alone rather than Murray LIS for severity grading 4:
   • Grade 0: >250
   • Grade 1: 175-250
   • Grade 2: 100-174
   • Grade 3: <100
2. Avoid Murray LIS in patients receiving lung-protective ventilation, as PEEP and compliance measurements become therapy-dependent and unreliable 4

For General Severe Respiratory Failure Risk Stratification:

When ECMO is not being considered, use established pneumonia severity tools (CURB-65, PSI, IDSA/ATS criteria) rather than RESP or Murray scores, as these are designed for different clinical contexts. 5, 6, 7

Key Pitfalls to Avoid

• Do not rely solely on RESP score for ECMO decisions in COVID-19 patients, as it significantly underperforms (AUC 0.61 vs 0.74 in original cohort) 2
• Do not use Murray LIS in patients on lung-protective ventilation; substitute with simple PaO2/FiO2 ratio assessment 4
• Do not apply RESP score to predict outcomes in prolonged ECMO (≥14 days), as discrimination is poor 3
• Do not ignore ECMO complications in outcome prediction; gastrointestinal hemorrhage, neurologic complications, and CPR during ECMO significantly worsen survival 3
• Recognize that immunosuppression (relative risk 0.43) and pre-ECMO cardiac arrest (relative risk 0.48) are particularly strong negative predictors 2