Point-of-Care Ketone Testing for Dehydration Assessment in Children with Gastroenteritis
Point-of-care capillary ketone testing via fingerprick is a moderately useful adjunctive tool for assessing dehydration severity in children with acute gastroenteritis, with a threshold of 2.5 mmol/L demonstrating reasonable predictive value for the need for intravenous or nasogastric rehydration.
Evidence Supporting Clinical Utility
The most recent and highest-quality evidence comes from a 2026 observational study of 540 children aged 3 months to 10 years with acute gastroenteritis 1. This study established that:
- Children requiring IV/nasogastric rehydration had significantly higher median capillary ketone levels (3.45 mmol/L) compared to those managed with oral rehydration or no intervention (1.2 mmol/L), p < 0.001
- A threshold of 2.5 mmol/L demonstrated:
- 62% sensitivity
- 79% specificity
- 68% positive predictive value
- 74% negative predictive value
- ROC curve AUC of 0.75
This threshold provides moderate accuracy for predicting rehydration needs and can serve as an objective, rapid, cost-effective decision-making tool 1.
Correlation with Dehydration Severity and Acidosis
Earlier research supports the relationship between ketone levels and clinical markers:
- A 2013 study of 188 children demonstrated a significant positive correlation between serum ketone concentration and clinical dehydration scores (Spearman's rho = 0.22, p = 0.003) 2
- Children with moderate dehydration had higher median ketone levels (3.6 mmol/L) versus mild dehydration (1.4 mmol/L), p = 0.007 2
- Ketone concentration inversely correlated with serum bicarbonate (ρ = -0.26, p < 0.001), indicating relationship with metabolic acidosis severity 2
A 2006 study found that ketone levels >1.2 mmol/L had a 66.7% positive predictive value for hospital admission, with performance comparable to Pediatric Risk of Admission scores 3.
Practical Application Algorithm
When to measure:
- Children presenting with vomiting and/or diarrhea
- Clinical uncertainty about rehydration route needed
- Triage decision-making in emergency or primary care settings
How to interpret:
- <1.2 mmol/L: Mild ketosis; oral rehydration likely sufficient
- 1.2-2.5 mmol/L: Moderate ketosis; close monitoring, trial of oral rehydration with low threshold for escalation
- >2.5 mmol/L: Significant ketosis; strong consideration for IV or nasogastric rehydration
- >3.5 mmol/L: Severe ketosis; IV/nasogastric rehydration highly likely needed
Technical considerations:
- Use β-hydroxybutyrate measurement (not urine acetoacetate)
- Point-of-care blood testing preferred over urine in young children who cannot void on demand 4
- Results available within minutes at bedside
Important Caveats
Limitations of the evidence:
- The predictive accuracy is moderate (AUC 0.75), not excellent
- Ketone levels should supplement, not replace, clinical assessment of dehydration
- Individual variability exists based on duration of illness, oral intake, and fever 3
Context matters:
- Ketone levels correlate strongly with decreased oral intake (R² = 0.25, p < 0.001) 3
- Higher levels expected in children with anorexia, vomiting, and fever 3
- Normal ranges in healthy children: typically 0.06-0.20 mmol/L after overnight fast 5
This tool is NOT for diabetic ketoacidosis assessment - the guidelines and thresholds cited for diabetes management (≥3.0 mmol/L for DKA diagnosis) 5 are completely different clinical contexts and should not be confused with gastroenteritis-related dehydration assessment.
Clinical Bottom Line
Use capillary blood ketone measurement as an objective adjunct to clinical dehydration assessment in children with gastroenteritis. A level >2.5 mmol/L should prompt strong consideration for IV or nasogastric rehydration, while levels <1.2 mmol/L support a trial of oral rehydration in clinically stable children. This rapid, inexpensive test can improve triage decisions and patient referrals in both emergency and primary care settings 1.