How do I calculate the Z score for a pediatric patient's growth?

Calculating Z-Scores for Pediatric Growth Assessment

What is a Z-Score?

A Z-score indicates how many standard deviations a measurement deviates from the population mean, with the formula: Z-score = (observed value - expected value) / standard deviation. 1, 2

• A Z-score of 0 represents the population mean 3
• Each unit represents one standard deviation from the mean 3
• Z-scores are superior to percentages because they account for interindividual variability and allow tracking of changes over time 3

Growth Parameter Z-Score Calculation

Step 1: Obtain Accurate Measurements

Measure the following parameters and express all as Z-scores 3:

• Weight
• Height/length
• Head circumference (for children <36 months) 3
• Body mass index-for-age (or weight-for-length in infants <2 years) 3

Step 2: Select Appropriate Growth Charts

For children under 24 months, use WHO international growth charts; for children 24-59 months, use CDC growth charts. 4

• WHO charts are based on predominantly breastfed healthy children and represent optimal growth standards 4
• WHO charts have less variability than CDC charts for children under 24 months 4
• The CDC recommends using modified WHO curves that include the 2.3 and 97.7 percentiles 4

Step 3: Calculate Body Surface Area (BSA)

BSA is required for many cardiac and vascular measurements 1, 5:

• Use Du Bois, Haycock, or Mosteller formulas 1
• BSA is the most commonly used method to normalize measurements based on body size 1

Step 4: Apply Regression Equations

For cardiac structures, use regression equations that relate the measured dimension to BSA 5:

• Regression equations are derived from large pediatric cohorts using modern echocardiographic equipment 5
• Online calculators are available at: https://www.marfan.fr/accueil/z-score-calculus/ or https://marfan.org/dx/z-score-adults/ 1

Interpretation of Z-Scores

Normal Range Definition

Z-scores between -2.0 and +2.0 are considered normal for most pediatric measurements. 1, 3

• Z-scores below -2 (approximately 5th percentile) or above +2 (approximately 95th percentile) are clinically significant 3
• For coronary arteries specifically, normal is defined as Z-score always <2, with dilation beginning at ≥2 2
• A Z-score ≥2.5 in one coronary artery branch would occur in only 0.6% of the normal population 1

Clinical Significance by Severity

For aortic measurements 1:

• Mild dilatation: Z-score 2.0 to 3.0 1
• Moderate dilatation: Z-score 3.01 to 4.0 1
• Severe dilatation: Z-score >4.0 1

Special Considerations

Age-Specific Applications

In neonates, always use age-specific Z-scores for cardiac measurements, as absolute values vary significantly with age and body size. 3

• Never use absolute measurements alone without converting to Z-scores in neonates 3
• Ensure appropriate reference populations match the infant's gestational age and ethnicity 3

Growth Monitoring Frequency

• Reassess nutritional status at least weekly throughout hospitalization 3
• Use the same imaging modality and measurement method for serial assessments over time 1
• Deviations from expected patterns require further evaluation 4

Body Size Extremes

For patients at extremes of body weight, consider indexing by height rather than BSA alone. 1

• An aorta-height index >32.1 mm/m is associated with 12% yearly risk 1
• Z-scores may over- or underestimate in over- or underweight patients 1

Common Pitfalls to Avoid

• Do not rely solely on percentages of predicted values, as these do not account for normal distribution variability 3
• Recognize that different Z-score systems exist (Boston, Montreal, DC, Pediatric Heart Network) and may yield different results, potentially changing management in up to 22% of cases 6
• Formula-fed children tend to gain weight more quickly than breastfed children, so their growth may not always follow WHO curve patterns 4
• For coronary arteries, anatomic variations are frequent in the left main coronary artery, where Z-scores must be interpreted with caution 1