Back

Application of Blood Glucose and Ketone Meters for Measuring Glucose and Ketone Levels in Urine Samples

Goluch, E. D.; Panahi, Z.; Denley, E.

2025-09-08 bioengineering
10.1101/2025.09.03.674019 bioRxiv
Show abstract

BACKGROUNDUrine glucose dipstick tests have low sensitivity, particularly in detecting glucose in early or mild cases of diabetes in cats and dogs. Ketone dipstick tests only detect acetoacetate and do not measure {beta}-hydroxybutyrate (BHB) which is the predominant ketone in diabetic patients and the most reliable indicator of diabetic ketoacidosis (DKA). OBJECTIVETo evaluate the accuracy, sensitivity, and specificity of urine dipstick tests for glucose and ketone measurements in cats and dogs. To assess whether blood-based Fora meters provide more reliable urine measurements for diabetes detection. METHODSStandard solutions of glucose, BHB, and acetoacetate in artificial urine were analyzed using Fora meters and dipstick tests. Urine samples from 37 random dogs and 27 random cats were tested and results were analyzed using regression analysis, confusion matrices, and Bland-Altman plots. RESULTSFora meters accurately detected glucose and BHP ketone levels in artificial urine with strong linearity to standards. Urine dipstick tests showed moderate accuracy (74%) but low sensitivity (57%) for glucose; performance varied by species and specific gravity. Urine dipstick test underestimate specific gravity in samples with high pH or high glucose levels. Urine dipstick tests cannot be used for BHP ketone measurements (sensitivity < 1%). CLINICAL RELEVANCEFora meters even though originally designed for blood, may offer a more precise, point-of-care alternative to dipsticks for diagnosing diabetes and DKA in cats and dogs. Improved urine glucose and ketone quantification can help reduce false diagnosis and support earlier and more accurate treatment decisions.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.