Fast is fine, but accuracy is everything: making intraoperative transfusion decisions using point-of-care testing

Abstract

Accurate and timely diagnostic information is a vital adjunct to clinical assessment to inform therapeutic decision-making, including decisions to transfuse, or not transfuse, blood components. A prospective cohort study of diagnostic point-of-care (POC) haemoglobin measurements on arterial or central venous samples from adults undergoing major noncardiac surgery compared three widely used devices, HemoCue®, i-STAT™, and the Rad-67™ pulse CO-Oxymeter® finger sensor device, against standard laboratory haemoglobin measurements, but importantly not against a blood gas analyser. The study focused on haemoglobin results below 100 g L⁻¹ to establish the utility of these devices to guide red cell transfusion decisions. None of the limits of agreement between POC device and laboratory results were within the allowable difference of plus or minus 4 g L⁻¹, and no device consistently over- or underestimated laboratory haemoglobin results in the same direction. However, results from the HemoCue® had the lowest likelihood to lead to inappropriate red cell transfusion. Clinicians should be aware of the patient, sample, and device factors that can influence the accuracy of POC haemoglobin testing results.