The effectiveness of blood glucose and blood ketone measurement in identifying significant acidosis in diabetic ketoacidosis patients

Background: Patients with diabetic ketoacidosis (DKA), a potentially fatal complication of type 1 diabetes, have hyperglycemia, ketonemia and metabolic acidosis. Blood glucose and blood ketone results are often used to triage patients with suspected DKA. This study aimed to establish how effective blood glucose and blood ketone (beta-hydroxybutyrate, BOHB) measurements are in identifying patients with significant acidosis and sought to validate existing diagnostic BOHB thresholds.

Methods: Initial Emergency Department results on 161 presumptive DKA episodes in 95 patients (42 F, 53 M, age range 14-89 years) containing a complete dataset of D (glucose), K (BOHB) and A (Bicarbonate [HCO3] and pH) results.

Results: Blood glucose correlated poorly with BOHB (r = 0.28 p = 0.0003), pH (r= -0.25, p = 0.002) and HCO3 (r= -0.17, p = 0.04). BOHB, though better, was still limited in predicting pH (r = -0.44, p < 0.0001) and HCO3 (r = -0.49, p < 0.0001). A HCO3 of 18mmol/L equated to a BOHB concentration of 4.3mmol/L, whilst a HCO3 of 15mmol/L equated to a BOHB of 4.7mmol/L. Of the 133 of 161 events with HCO3 < 18mmol/L, 22 were not hyperglycemic (> 13.9mmol/L, n = 8), ketonemic (≤ 3mmol/L, n = 9) or either (n = 5).

Conclusions: The commonly employed BOHB diagnostic cutoff of 3mmol/L could not be verified. Since acid-base status was poorly predicted by both glucose and BOHB, this highlights that, regardless of their results, pH and/or HCO3 should also be tested in any patient suspected of DKA.