Use of point-of-care blood ketone testing to diagnose and monitor patients with diabetic ketoacidosis

In a recent editorial, Dhatariya ¹ discussed the use of point-of-care blood ketone monitors in the management of diabetic ketoacidosis (DKA) in adults. Following the publication of the Joint British Diabetes Society guidelines on the management of DKA in adults, ² we introduced blood ketone testing at the bedside using Abbott Precision Xceed Pro meters.

Where possible, it is advisable that point-of-care testing (POCT) should be supported by a laboratory method for the same analyte, so that results outside the measurable range of the device, or results which are not consistent with the clinical picture, may be verified or otherwise. We, therefore, evaluated and introduced a laboratory method for betahydroxybutyrate (BHB). Like others, we found good agreement between results of POCT capillary ketone measurement using bedside meters, and laboratory BHB results obtained on venous serum samples, up to a concentration of 3 mmol/L.^3,4 Above 3 mmol/L, POCT results were consistently lower than serum BHB measured on venous samples. For example, one patient had a capillary POCT ketone result of 5.4 mmol/L but a laboratory venous serum BHB of 13.7 mmol/L. Dhatariya is correct to point out that both results give a diagnosis of DKA and that the rate of decline is rarely if ever used in isolation to guide treatment. However, we think it is important that all users of POCT ketone meters are aware of this limitation. In our comparison study, some patients appeared to show no or very little response to treatment using capillary POCT ketone tests, whereas laboratory measurements on venous samples showed an appropriate decrease in BHB (>0.5 mmol/L/h). ² For example, one patient had a capillary POCT ketone result of 3.6 mmol/L initially, which decreased to 3.3 mmol/L after 5 h of treatment. By contrast, BHB laboratory results on venous samples collected simultaneously showed a fall from 12.7 mmol/L to 3.1 mmol/L. Capillary POCT ketone measurement alone may give a false impression that ketoacidosis is not resolving.

We also measured venous blood ketones using the same POCT meters and found good agreement with venous blood BHB measurements, suggesting that the falsely low results observed using capillary samples reflect the characteristics of capillary blood rather than the meter itself. Therefore, this limitation may potentially affect all POCT meters using capillary samples to measure ketones.

We have introduced routine laboratory BHB analysis and incorporated venous BHB measurement into our local DKA guideline. We have also included a statement in our guideline to ensure that blood ketone monitoring is never used on its own to guide treatment:

When blood ketones are >3.0 mmol/L, ketone meter measurements using capillary blood may underestimate the actual value and rate of change may also be falsely low. Review venous ketone results and also assess change in capillary glucose and venous bicarbonate to help determine if IV insulin rate needs to be changed.

We advise all hospitals using POCT capillary blood ketone testing to be aware of the possible limitations when using any bedside meters to measure blood ketones. The laboratory has an important role as with any POCT to ensure that all users are aware of any limitations.