Diagnostic accuracy of venous blood gas electrolytes for identifying diabetic ketoacidosis in the emergency department

Abstract

Objectives: Diagnosing diabetic ketoacidosis (DKA) has traditionally required a venous blood gas (VBG) to obtain serum pH and a serum chemistry panel to obtain electrolyte values. Because newer blood gas analyzers have the ability to report electrolyte values and glucose in addition to pH, this diagnostic process could theoretically be condensed. However, neither the diagnostic accuracy of the VBG for DKA nor the agreement between the VBG electrolytes and the serum chemistry electrolytes, including sodium, chloride, and bicarbonate, has been evaluated in the context of acute hyperglycemia. The purpose of this study was to assess the accuracy of VBG electrolytes for diagnosing DKA using serum chemistry electrolytes measures as the criterion standard and to describe the correlation between VBG and serum chemistry electrolytes in a sample of hyperglycemic patients seen in the emergency department (ED).

Methods: The authors prospectively identified a convenience sample of ED patients with serum blood glucose ≥ 250 mg/dL and examined their paired VBG and serum chemistry electrolytes. The diagnosis of DKA was made by using American Diabetes Association (ADA) criteria including serum glucose ≥ 250 mg/dL, serum anion gap > 10 mEq/L, bicarbonate ≤ 18 mEq/L, serum pH ≤ 7.30, and presence of ketosis. Serum chemistry electrolyte values were considered to be the criterion standard. Diagnostic test characteristics of VBG electrolytes including sensitivity and specificity were compared against this standard. In addition, correlation coefficients for individual electrolytes and anion gap between VBG and chemistry electrolytes were calculated.

Results: Paired VBG and serum chemistry panels were available for 342 patients, of whom 46 (13.5%) had DKA. The sensitivity and specificity of the VBG electrolytes for diagnosing DKA was 97.8% (95% confidence interval [CI] = 88.5% to 99.9%) and 100% (95% CI = 98.8% to 100%), respectively. One case of DKA was missed by the VBG. Correlation coefficients between VBG and serum chemistry were 0.90, 0.73, 0.94, and 0.81 for sodium, chloride, bicarbonate, and anion gap, respectively.

Conclusions: The VBG electrolytes were 97.8% sensitive and 100% specific for the diagnosis of DKA in hyperglycemic patients. These preliminary findings support the use of VBG electrolytes in lieu of VBG along with serum chemistry analysis to rule in or rule out DKA.