The utility of ketones at triage: a prospective cohort study

Objective: To establish the relationship between serum point-of-care (POC) ketones at triage and moderate-to-severe dehydration based on the validated Gorelick Scales.

Design, setting and patients: Prospective unblinded study from April 2016 to February 2017 in a paediatric emergency department. Patients aged 1 month to 5 years, with vomiting and/or diarrhoea and/or decreased intake with signs of moderate or severe dehydration or clinical concern for hypoglycaemia were eligible.

Main outcome measures: The primary outcome was to describe the relationship between triage POC ketones to the two Gorelick Scales. Secondary outcomes were to examine the response of ketone levels to fluid/glucose administration and patient disposition.

Results: One-hundred and ninety-eight patients were included; median age 1.8 years. The median triage ketones were 4.6 (IQR 2.8-5.6) mmol/L. A weak correlation was identified between triage ketones and the 10-point Gorelick Scale (Spearman's ρ=0.217, p=0.002), however no correlation between triage ketones and the 4-point Gorelick Scale was identified. Those admitted had median triage ketones of 5.2 (IQR 4-6) mmol/L and repeat ketones of 4.6 (IQR 3.3-5.7) mmol/L compared with 4.2 (IQR 2.4-5.3) mmol/L and 2.9 (IQR 1.6-4.2) mmol/L in those discharged home.

Conclusion: No correlation between triage POC ketones and the 4-point Gorelick Scale was established. POC ketones at triage have poor accuracy for predicting hospital admission. The elevated profile of POC ketones in non-diabetic children with acute illness suggests a potential target of tailored treatments for further research.