Arterial blood gas results rarely influence emergency physician management of patients with suspected diabetic ketoacidosis

Abstract

Diabetic ketoacidosis (DKA) is considered the most common acute, life-threatening complication of diabetes. Arterial blood gas (ABG) sampling is considered by most clinicians to be essential in the initial evaluation of patients with suspected DKA. However, ABG sampling is invasive and potentially difficult and may be accompanied by risks or complications. Venous pH sampling has been proposed as an alternative to ABG sampling; venous pH has been shown to be well correlated with arterial pH in the diagnostic evaluation of patients with DKA. It is possible that ABG data may not be needed by physicians who manage patients with suspected DKA. The hypothesis that ABG results for patients with suspected DKA have no influence on the management of emergency physicians was tested, and the correlation and precision between venous pH and arterial pH was validated. The prospective observational study was conducted among physicians in the emergency department (ED) of an urban teaching hospital with an annual volume of 55,000 visits. Patients included in the study were those with capillary blood glucose levels 200 mg/dL or greater, ketonuria, and clinical signs and symptoms of DKA. Venous pH, chemistry panel, and ABG readings were obtained before treatment. The attending emergency physicians indicated planned management and disposition on a standardized form before and after reviewing ABG and venous pH results. The study was designed to detect a 10% difference in management decisions. In 200 cases included in the study, results of ABG analysis changed the emergency physicians' diagnosis in 2 cases (1%), altered treatment in 7 cases (3.5%), and changed the disposition of 2 cases (1%). The pH value of the ABG changed the treatment or disposition in 2.5% of cases, whereas the Po2 and Pco2 results of the ABG assay changed the treatment or disposition of 1% of patients. There was good correlation between venous pH and arterial pH. These findings were supportive of the view that ABG results rarely influence the decisions of emergency physicians in regard to diagnosis, treatment, or disposition in patients with suspected DKA. Venous pH findings correlated well and with enough precision with arterial pH to substitute for ABG analysis in these patients.