Objectives: To test the hypothesis that arterial blood gas (ABG) results for patients with suspected diabetic ketoacidosis (DKA) do not influence emergency physicians' management decisions and to assess correlation and precision between venous pH and arterial pH.
Methods: Prospective, observational study of emergency physicians' decision making for consecutive ED patients with suspected DKA. Inclusion criteria were capillary blood glucose equal to or greater than 200 mg/dL, ketonuria, and clinical signs and symptoms of DKA. Venous pH, chemistry panel, and ABGs were drawn before treatment. Attending emergency physicians indicated planned management and disposition on a standardized form before and after reviewing ABG and venous pH results. This study was powered to detect a 10% difference in management decisions (n = 195). Pearson's correlation and Bland-Altman bias plot were used to compare venous pH and arterial pH.
Results: ABG analysis changed the emergency physicians' diagnosis in 2/200 cases (1.0%; 95% confidence interval [95% CI] = 0.3% to 3.6%), altered treatment in 7/200 cases (3.5%; 95% CI = 1.7% to 7.1%), and changed disposition in 2/200 cases (1.0%; 95% CI = 0.3% to 3.6%). The pH value of the ABGs changed the treatment or disposition in 5/200 patients (2.5%; 95% CI = 1.1% to 5.7%). The Po(2) and Pco(2) results of the ABGs altered treatment and disposition in 2/200 patients (1.0%; 95% CI = 0.3% to 3.6%). Venous pH correlated well with arterial pH (r = 0.951), and bias plotting yielded a bias value of -0.015 (+/- 0.006 pH units).
Conclusions: ABG results rarely influenced emergency physicians' decisions on diagnosis, treatment, or disposition in suspected DKA patients. Venous pH correlated well and was precise enough with arterial pH to serve as a substitute.