Introduction: The purpose of this study is to compare blood glucose levels measured by a point of care (POC) device to laboratory measurement using the same sample venous blood from patients with suspected diabetic ketoacidosis (DKA).
Methods: A descriptive correlational design was used for this IRB-approved quality assurance project. The study site was the 50-bed BMC emergency department (ED) which has an annual census of over 100,000 patient visits. The convenience sample consisted of 54 blood samples from suspected DKA patients with orders for hourly blood draws for glucose measurement. Spearman correlations of the glucose POC values, reference lab values, and differences between the two, were evaluated. A chi-square test was used to evaluate the association between the acidosis status and FDA acceptability of POC values.
Results: Patient age range was 10-86 years; 63% were females; 46% had a final diagnosis of DKA. POC values underestimated glucose levels 93% of the time. There was a high correlation between the lab value and the magnitude of the difference, (lab minus POC value) indicating that the higher the true glucose value, the greater the difference between the lab and the POC value. A chi-square test showed no overall association between acidosis and FDA-acceptability.
Discussion: The POC values underestimated lab reported glucose levels in 50 of 54 cases even with the use of same venous sample sent to the lab, which make it highly unreliable for use in monitoring suspected DKA patients.