PRISM (Pediatric Risk of Mortality)

The PRISM (Pediatric Risk of Mortality) score is a validated severity-of-illness scoring system designed for pediatric patients (newborn to 18 years) in pediatric intensive care units (PICUs). Developed in 1988 and refined in subsequent iterations (PRISM III and PRISM IV), it quantifies disease severity and predicts hospital mortality risk based on physiological and laboratory data collected within the first 24 hours of PICU admission. PRISM is widely used to assess critically ill children, excluding premature neonates in neonatal ICUs (NICUs), where scores like CRIB II or SNAP-II are preferred.

Purpose

• Clinical: Evaluates illness severity to guide resource allocation and inform prognosis discussions.
• Quality Benchmarking: Adjusts for patient risk to compare PICU performance across institutions.
• Research: Standardizes risk adjustment in pediatric critical care studies.
• Prognostic: Estimates hospital mortality risk for cohorts, not individual patient decisions.

Components of PRISM Score

The original PRISM score (1988) uses 14 physiological and laboratory variables, each scored based on deviation from age-specific normal ranges. Data is collected within the first 24 hours of PICU admission, using the worst values observed. Key variables include:

1. Cardiovascular:
   • Heart rate (beats/min)
   • Systolic blood pressure (mmHg)
   • Temperature (°C)
2. Respiratory:
   • Respiratory rate (breaths/min)
   • PaO₂/FiO₂ ratio
   • PaCO₂ (mmHg)
3. Neurological:
   • Glasgow Coma Scale (GCS)
   • Pupillary reactions
4. Metabolic/Chemical:
   • Glucose (mg/dL)
   • Potassium (mmol/L)
   • Blood urea nitrogen (BUN, mg/dL)
   • Creatinine (mg/dL)
5. Hematological:
   • White blood cell count (×10³/mm³)
   • Platelet count (×10³/mm³)

Each variable is assigned points (e.g., 0–10) based on predefined thresholds tailored to pediatric age groups (neonates, infants, children, adolescents). The total score correlates with mortality risk, with higher scores indicating greater severity.

PRISM III and PRISM IV

• PRISM III (1996): An updated version with 17 variables, refined thresholds, and age-specific adjustments. It includes additional parameters like pH, prothrombin time, and bilirubin. PRISM III is more precise and widely used in modern PICUs.
• PRISM IV (2015): Further refined with improved calibration, incorporating diagnosis-specific adjustments and better discrimination for low- and high-risk patients. It requires proprietary software for full implementation.

The original PRISM is simpler but less accurate than PRISM III/IV. This document focuses on the general PRISM framework, with notes on updates.

Scoring Example (PRISM)

• A 5-year-old with sepsis, worst values in 24 hours:
  • Heart rate 160 beats/min (3 points)
  • Systolic BP 70 mmHg (4 points)
  • GCS 10 (2 points)
  • Glucose 200 mg/dL (1 point)
  • Other variables contributing 5 points.
• Total PRISM score = 3 + 4 + 2 + 1 + 5 = 15.
• This score, with a diagnosis of sepsis, may predict a mortality risk of ~10–20% (depending on calibration and PRISM version).

Clinical Application

1. Risk Stratification: Identifies high-risk patients for intensified monitoring or interventions.
2. Prognostic Guidance: Provides objective data for family discussions, combined with clinical judgment.
3. Quality Improvement: Adjusts mortality rates for illness severity, enabling fair PICU performance comparisons.
4. Research: Standardizes risk in studies of pediatric critical care interventions or outcomes.

Advantages

• Pediatric-Specific: Tailored to children’s physiology, unlike adult scores (e.g., APACHE II).
• Validated: Extensively studied across diverse pediatric populations and conditions.
• Comprehensive: Captures multiple organ systems (cardiovascular, respiratory, neurological, etc.).
• Updated Versions: PRISM III/IV offer improved accuracy and calibration for modern PICUs.

Limitations

• Not for Neonates: Excludes premature infants in NICUs, where CRIB II or SNAP-II are used.
• Time-Sensitive: Relies on 24-hour worst values, missing later clinical changes.
• Calibration: Mortality predictions may vary by region or PICU practices, requiring periodic recalibration.
• Complexity: PRISM III/IV require more variables and proprietary tools, limiting accessibility in resource-constrained settings.
• Cohort Focus: Designed for group risk prediction, not individual prognosis.

Implementation

• Data Collection: Use standardized protocols for accurate measurement (e.g., arterial blood gas for PaO₂, consistent GCS scoring). Age-specific normal ranges are critical.
• Calculation: Use validated PRISM scoring tables or software (PRISM III/IV often require licensed tools). Online calculators exist for original PRISM.
• Interpretation: Combine score with clinical context and diagnosis. Avoid sole reliance on PRISM for treatment decisions.
• Training: PICU staff need training to ensure consistent data collection and score application.

Comparison with Other Scores

• APACHE II: For adult ICUs, not suitable for pediatric or neonatal patients.
• CRIB II/SNAP-II: Neonatal-specific, used in NICUs for preterm or low-birth-weight infants.
• PIM3 (Pediatric Index of Mortality 3): Simpler than PRISM, uses admission data only, but less comprehensive.
• SOFA (Sequential Organ Failure Assessment): Tracks organ dysfunction over time, less specific to pediatrics.

PRISM (especially III/IV) is preferred in PICUs for its pediatric focus and robust validation.

Relevance to Other Settings

• Neonatal ICUs: PRISM is not suitable for preterm neonates; use CRIB II or SNAP-II.
• Adult ICUs: Inappropriate for adults; use APACHE II or SAPS II.
• Mixed ICUs: Apply PRISM for pediatric patients only, with separate tools for adults or neonates.

Conclusion

The PRISM score, particularly PRISM III and IV, is a cornerstone tool for assessing illness severity and predicting mortality in pediatric ICU patients. Its pediatric-specific design, incorporating physiological and laboratory data, makes it invaluable for risk stratification, quality benchmarking, and research. Medical professionals should use PRISM alongside clinical judgment, ensuring accurate data collection and context-specific interpretation. While not applicable to neonatal or adult populations, PRISM remains a gold standard in pediatric critical care when implemented correctly.