Skip to content
Back to search results

Score for Neonatal Acute Physiology-II (SNAP-II)

Introduced by
David K. Richardson
Year
2001
Type
Clinician-rated
Family
severity-stratification
Font size

What does it do?

SNAP-II is a six-variable physiological scoring system designed to quantify acute illness severity in very low birth weight (VLBW) neonates and predict mortality risk.

Developed by Richardson and colleagues in 2001 as a refinement of the original SNAP, it incorporates readily available bedside physiological variables (mean blood pressure, lowest body temperature, hypoxemia, seizures, urine output, and sepsis indicators) measured within the first 12 hours of life. SNAP-II is widely used in neonatal quality improvement, clinical research, and benchmarking of NICU outcomes.

The intuition behind the method

SNAP-II is grounded in the principle that acute physiological derangement in the first hours of life reflects the severity of perinatal insult and predicts short-term mortality risk. Unlike scales that assess neurological examination findings (which may lag behind physiological deterioration), SNAP-II captures real-time physiological dysregulation: hypotension reflects inadequate perfusion; hypothermia reflects metabolic dysfunction; hypoxemia reflects respiratory/circulatory insufficiency; and oliguria reflects renal hypoperfusion. The scale is parsimonious (only six variables, all routinely monitored), making it feasible for rapid bedside calculation and risk stratification. By quantifying acute physiology, SNAP-II allows clinicians to identify which infants are at highest mortality risk, enabling triage decisions, prognostic counseling, and quality benchmarking.

History

The original SNAP was published by Richardson et al. in 1993, incorporating 34 physiological variables in extremely low birth weight infants. Recognition that SNAP was complex and difficult to calculate at the bedside prompted development of a simplified version. SNAP-II was published in 2001, retaining only the six most predictive variables while maintaining strong discriminatory power. Validation studies across North American NICUs (Vermont Oxford Network institutions) confirmed SNAP-II's accuracy and generalizability. SNAP-II remains the primary illness severity score recommended by the National Institute of Child Health and Human Development (NICHD) for NICU quality improvement and has been adopted by the National Quality Forum for NICU outcome reporting.

When to use it

SNAP-II is specifically validated for very low birth weight (VLBW, <1500 grams) and extremely low birth weight (ELBW, <1000 grams) neonates during the first 12 hours of NICU admission. It is indicated for: (1) prognostic risk stratification and mortality prediction in acutely ill VLBW/ELBW infants; (2) quality assessment: comparing observed mortality to SNAP-II-predicted mortality as measure of NICU quality and outcome benchmarking across institutions; (3) clinical research: adjustment for illness severity when comparing treatment groups or outcomes across different populations; (4) parental counseling: objective physiological data to communicate prognosis; (5) triage decisions in resource-limited settings where high-risk infants may require transfer to higher-level centers.

Limitations and considerations

  • Limited to VLBW/ELBW populations (<1500 g); not validated for normal or large for gestational age infants, limiting applicability in term infant ICUs.
  • First 12-hour window captures acute illness at birth; does not assess subsequent physiological changes or adaptation during first week of life; infants may improve or deteriorate significantly after initial SNAP-II calculation.
  • Predicted mortality is population-average risk; individual infants may have substantially different prognosis based on unmeasured factors (congenital anomalies, genetic conditions, quality of neonatal resuscitation, parental preferences).
  • Some variables are crude surrogates: seizures present/absent is binary (does not account for frequency/duration); sepsis is clinical presumption (culture results not yet available at 12 hours); PaO2/FiO2 ratio may be unavailable if infant on non-invasive support.
  • Does not account for subsequent deterioration or secondary insults (post-natal infection, NEC, IVH) developing after first 12 hours; requires serial reassessment.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Parsimonious six-variable design uses only routinely available bedside data; no special testing required; rapid bedside calculation feasible within first 12 hours of admission.
  • Strong predictive validity: Original validation on 6,093 VLBW infants demonstrated SNAP-II mortality prediction accuracy (area under the ROC curve 0.82–0.85) superior to clinical judgment alone.
  • Physiologically coherent: Each variable (blood pressure, temperature, oxygenation, seizure, urine output, sepsis) reflects distinct aspect of acute organ system dysfunction, reducing redundancy.
  • Widely adopted and externally validated: Used across >500 NICUs in North America and internationally; multiple validation studies in diverse populations confirm generalizability.
  • Supports quality improvement: Standardized mortality ratio (observed/expected mortality) enables benchmarking, quality monitoring, and identification of NICUs with superior outcomes.

Applications

  • Neonatal mortality prediction and prognostic counseling: Clinicians calculate SNAP-II for newly admitted VLBW infants and communicate mortality risk estimates to parents, supporting informed decision-making about intensity of care and advance planning.
  • NICU quality improvement and benchmarking: Institutional quality programs calculate standardized mortality ratios (observed mortality / SNAP-II predicted mortality) to assess quality and identify performance gaps; ratios <1.0 indicate better-than-expected outcomes.
  • Clinical research adjustment for severity: Comparative effectiveness trials use SNAP-II scores to adjust for baseline illness severity, ensuring groups are balanced; this is particularly important in observational studies where treatment assignment may correlate with severity.
  • Resource allocation and triage: In resource-limited settings or during surge capacity events, SNAP-II identifies infants at highest mortality risk who warrant referral to centers with advanced capabilities (ECMO, surgical intervention).
  • Trends and outcome monitoring: Serial SNAP-II scores (at 12 hours, 24 hours, 48 hours) or similar physiological scoring may track improvement or deterioration over time, informing care escalation or de-escalation decisions.

How it works

How it works

1

SNAP-II is calculated from physiological data routinely obtained during the first 12 hours of neonatal life in an intensive care setting. No special assessment procedure is required; clinicians extract values from existing vital sign monitoring, laboratory results, and clinical observation. The six variables are: (1) mean arterial blood pressure (mmHg); (2) lowest body temperature (°C); (3) P/F ratio (partial pressure of oxygen divided by fraction of inspired oxygen, or surrogate SpO2/FiO2); (4) presence of seizures (yes/no); (5) urine output (mL/kg/hour); (6) sepsis or antibiotic use (yes/no, indicating suspicion of infection). Data collection takes <5 minutes once vital signs and labs are available.

2

SNAP-II scoring is as follows: Mean Arterial Blood Pressure: <30 mmHg=3 points; 30–45 mmHg=0 points; 46–55 mmHg=2 points; >55 mmHg=0 points. Lowest Temperature: <35.0°C=3 points; 35.0–35.9°C=0 points; 36.0–36.9°C=0 points; ≥37.0°C=0 points. PaO2/FiO2 (or SpO2/FiO2 if PaO2 unavailable): <100=3 points; 100–299=0 points; ≥300=0 points. Seizures: Present=3 points; Absent=0 points. Urine Output: <0.5 mL/kg/hour=3 points; 0.5–1.0 mL/kg/hour=1 point; >1.0 mL/kg/hour=0 points. Sepsis Treated or Presumed: Yes (antibiotics started)=3 points; No=0 points. Total SNAP-II score ranges from 0 to 18 points. Each variable is scored independently; higher scores indicate greater physiological derangement. The total score is used to predict mortality risk.

3

SNAP-II scores stratify mortality risk as follows: 0–4 points: Low risk (mortality approximately 1–2%). 5–8 points: Intermediate-low risk (mortality approximately 3–5%). 9–12 points: Intermediate-high risk (mortality approximately 10–15%). 13–18 points: High risk (mortality approximately 30–50% or higher). These are population-level risk estimates; individual mortality risk varies based on gestational age, birth weight, and other clinical factors. SNAP-II is used clinically to: (1) identify which infants are at highest risk and warrant aggressive intervention; (2) provide parents with prognostic information; (3) guide resource allocation in resource-limited settings; (4) enable NICU quality improvement by comparing observed vs. predicted mortality (measured-to-expected mortality ratio). A measured mortality lower than predicted suggests better-than-average quality; higher measured mortality may indicate need for quality improvement initiatives.

Common pitfalls

  • Using SNAP-II alone for prognostic counseling without incorporating gestational age and birth weight; a SNAP-II of 12 has very different mortality meaning in a 24-week vs. 30-week infant.
  • Applying SNAP-II to infants outside the VLBW range where the prediction model is not validated; a term infant with SNAP-II 10 cannot be interpreted using published mortality thresholds.
  • Misinterpreting low SNAP-II scores as indicating safe, stable condition; scores of 0–4 reflect low mortality risk but do not mean the infant is healthy or without serious concerns (e.g., a term infant with meconium aspiration and good physiology may have low SNAP-II but require intensive support).
  • Fixed cutoffs without re-assessment; SNAP-II is calculated at 12 hours; an infant improving by day 2 or deteriorating requires updated calculation, not reliance on initial 12-hour score.
  • Confusing SNAP-II (acute physiology, first 12 hours) with chronic severity scales like CRIB (Clinical Risk Index for Babies), which incorporates birth weight and Apgar score; these scales are complementary but not interchangeable.

Frequently asked questions

A baby has a SNAP-II of 9. What does that mean for mortality?

A SNAP-II of 9 places the infant in the intermediate-high risk category, corresponding to approximately 10–15% mortality risk. This does not mean the infant will necessarily die; rather, it indicates substantial acute illness and increased risk compared to infants with lower scores. Clinically, this suggests the need for aggressive supportive care, close monitoring, and prognostic discussion with parents about realistic outcomes including the possibility of death or severe morbidity.

Can SNAP-II be used in term infants?

SNAP-II was developed specifically for VLBW (<1500 g) infants and is not validated in term infants. While the physiological variables are measured in all neonates, the prediction model and published mortality thresholds are not applicable to term populations. Use disease-specific severity scales or alternate risk stratification for term infants (e.g., PIM-3 for pediatric ICU patients, or condition-specific scores for meconium aspiration, sepsis, etc.).

What if an infant's SNAP-II variables change after 12 hours? Should I recalculate?

The original SNAP-II is calculated once at 12 hours post-admission. Clinically, physiological status changes substantially in the first days of life, and an infant improving may have markedly lower scores on day 3. Some research groups have extended SNAP to later time points (SNAP-PE at 24–48 hours), but the standard SNAP-II is not intended for serial recalculation. Instead, clinicians use SNAP-II as a baseline risk assessment and integrate subsequent clinical changes into ongoing prognostic updates.

How does SNAP-II differ from CRIB (Clinical Risk Index for Babies)?

SNAP-II and CRIB are complementary but distinct. SNAP-II uses six physiological variables measured in the first 12 hours and predicts mortality in VLBW infants. CRIB incorporates birth weight, gestational age, gender, and Apgar score at admission and was originally designed for UK populations; different prediction models and thresholds apply. In practice, some NICUs calculate both scores; SNAP-II focuses on acute physiology whereas CRIB emphasizes demographic and delivery factors.

Related methods

Sources

  1. Richardson, D. K., Gray, J. E., Gortmaker, S. L., Goldmann, D. A., Purohit, D. M., & Paige, D. (2001). Declining Severity Adjusted Mortality: Evidence of Improving Neonatal Intensive Care. Pediatrics, 108(2), 331-337.

    Open publication reference
    Copied.
  2. Richardson, D. K., Corcoran, J. D., Escobar, G. J., & Lee, S. K. (1993). SNAP-II: Simplified Newborn Physiology Score and Estimation of Mortality Risk in Very Low Birth Weight Infants. Pediatrics, 91(1), 33-41.

    Open publication reference
    Copied.