Diabetic ketoacidosis consists of elevated blood glucose, measurable ketone bodies and metabolic acidosis. Arterial blood gas determination is considered essential in the initial evaluation of patients with suspected diabetic ketoacidosis. Arterial blood sampling is painful, may be technically difficult and must be done in addition to venous sampling when testing for electrolytes and other values. The general correlation between arterial and venous pH measurements is well established, although this correlation has not been studied in patients with diabetic ketoacidosis. Brandenburg and Dire prospectively studied the relationship between arterial and venous blood gas values in the initial evaluation of patients with suspected diabetic ketoacidosis.

Thirty-eight patients with 44 episodes of diabetic ketoacidosis who presented to an emergency department with blood glucose levels greater than 250 mg per dL (13.9 mmol per L), urine dipstick results positive for ketones and clinical suspicion of diabetic ketoacidosis were included in the study. Arterial and venous samples were obtained as temporally close to each other as possible for blood gas analysis.

The mean difference between arterial and venous pH values was 0.03 (range: 0 to 0.11). Arterial and venous pH results, arterial and venous bicarbonate measurements and arterial bicarbonate and serum carbon dioxide results were also closely correlated.

The authors conclude that the peripheral venous pH measurement is a valid and reliable substitute for arterial pH in patients with diabetic ketoacidosis. A potential disadvantage of using venous determinations to determine the presence of diabetic ketoacidosis is that it may be more difficult to detect when mixed acid-base disturbances are present, since venous blood values may not reflect hypoxia in patients with respiratory depression. In patients with suspected mixed acid–base disorders or evidence of respiratory depression or hypoxemia on pulse oximetry, arterial blood gas analysis remains an appropriate means of testing for diabetic ketoacidosis. To ensure that venous sampling does not underestimate the degree of acidosis, a larger patient sample would be needed to examine more patients with severe acidosis. In addition, a study of final patient outcomes is necessary to determine whether the use of venous rather than arterial gas values affects outcome.