Reliability of Reagent Strips in Veterinary Medicine: A Critical Evaluation of <scp>pH</scp> and Urinary Density in Dogs using Statistical Modeling

BACKGROUND: Urinalysis is a simple and essential test in veterinary practice, involving physical, chemical, and microscopic analysis of urine. Reagent strips are widely used due to their practicality; however, their reliability compared with reference laboratory methods remains a subject of debate. OBJECTIVES: This study compares different brands of reagent strips for the measurement of urinary pH and urine specific gravity (USG) in dogs and to evaluate their performance in relation to reference methods. METHODS: This study evaluated the reliability of six brands of reagent strips for measuring pH and urine specific gravity in dogs, comparing them with the gold standard methods (bench pH meter and refractometer). Association between methods was evaluated using Pearson and Spearman correlation analyses, and agreement was assessed using Bland-Altman analysis. Linear and polynomial regression models were applied to derive correction equations where appropriate. Ninety-seven canine urine samples were analyzed, collected by natural urination, and processed under refrigeration for preservation. RESULTS: ) greater than 69%, validating their use to calibrate results and improve accuracy. In contrast, urine specific gravity measurements obtained by reagent strips showed poor agreement with refractometry, characterized by wide limits of agreement and clinically relevant bias. CONCLUSIONS: The reagent strips are inadequate for reliable determination of urine specific gravity in dogs and are not recommended as a substitute for a refractometer, while showing acceptable agreement for urinary pH estimation. According to the proposal, its use is feasible in routine clinical practice, provided that the values are adjusted by the proposed correction equations, which minimize scaling errors and significantly improve the accuracy of the results, if the manufacturers' guidelines are followed.