Abstract
Objective. Albumin/creatinine and protein/creatinine ratios were measured with the ARKRAY AUTION Eleven reflectometer using AUTION Screen and AUTION Sticks 10PA strips, respectively, against quantitative Siemens Advia reference procedures from 368 patient urines, as an evaluation of their applicability for use in points‐of‐care and small laboratories. Material and methods. Direct reflectance measurements were utilized to estimate imprecision, as well as to suggest reclassification of ordinal scale categories into normoalbuminuria, microalbuminuria and macroalbuminuria groups (3.4 g/mol and 34 g/mol cut‐off limits, corresponding to 30 mg/g and 300 mg/g creatinine in conventional units). Results. Analytically, ordinal scale albumin/creatinine ratios agreed in 86 % of cases with those obtained from Advia measurements, resulting in a kappa coefficient of 0.79. Protein/creatinine ratios of the AUTION Sticks 10PA strip were classified into three groups at limits of 11.3 g/mol and 56.6 g/mol (100 mg/g and 500 mg/g in conventional units), with an agreement of 77 % and a kappa coefficient of 0.65 against Advia procedures. To optimize clinical outcomes, cut‐off reflectances of ordinal scale categories of AUTION Eleven were adjusted. The clinical specificity of detecting an increased albumin/creatinine ratio was then increased from 81 % to 95 %, with clinical sensitivity kept at 88 % at the 3.4 g/mol limit of the reference procedure. Clinical specificity of the albuminuria field alone (at a clinical sensitivity of 88 %) was only 73 %. Adjustments to cut‐off reflectances of the reported categories for protein/creatinine ratios increased clinical specificity from 54 % to 94 %, while losing clinical sensitivity from 97 % to 89 % only, with an improved concordance of 83 % and a kappa coefficient of 0.75 against Advia measurements. The combination to creatinine measurements improved clinical specificity compared to 50 % by the protein field alone. In economic terms, it is estimated that population screening for microalbuminuria using the AUTION Eleven reflectometer is cheaper than by quantitative albumin/creatinine measurements alone, based on the incidence of end‐stage renal disease of 90 patients/million/year at the Northern Ostrobothnia Hospital District.
List of definitions and abbreviations | ||
Cut‐off limit, cut‐off reflectance (R%) | = | an adjustable limiting R% that divides reflectometric results into different ordinal scale categories, most importantly into the negative and lowest positive (1+) category |
Grey zone | = | concentration range between the specified detection and confirmation limits of a reference procedure, where the reflectometric reading may be ambiguous (+/− zone) |
FN(C) | = | fraction of false negative cases at or above the confirmation limit L(C) |
FN(G) | = | fraction of false negative cases within the grey zone |
FP(D) | = | fraction of false positive cases below the detection limit L(D) |
L(C), Confirmation limit | = | specified analyte concentration of a reference procedure, above which all reflectometric results should be positive |
L(D), Detection limit | = | specified analyte concentration of a reference procedure, below which all reflectometric results should remain negative |
LOQ, limit of quantitation | = | highest reflectance (R%) that is still statistically below the background reflectance (mean R% of negative samples minus three standard deviations of day‐to‐day analytical variation) |
R | = | reflectance, expressed as percentage (%) of maximum value |
SD | = | Standard deviation |
List of definitions and abbreviations | ||
Cut‐off limit, cut‐off reflectance (R%) | = | an adjustable limiting R% that divides reflectometric results into different ordinal scale categories, most importantly into the negative and lowest positive (1+) category |
Grey zone | = | concentration range between the specified detection and confirmation limits of a reference procedure, where the reflectometric reading may be ambiguous (+/− zone) |
FN(C) | = | fraction of false negative cases at or above the confirmation limit L(C) |
FN(G) | = | fraction of false negative cases within the grey zone |
FP(D) | = | fraction of false positive cases below the detection limit L(D) |
L(C), Confirmation limit | = | specified analyte concentration of a reference procedure, above which all reflectometric results should be positive |
L(D), Detection limit | = | specified analyte concentration of a reference procedure, below which all reflectometric results should remain negative |
LOQ, limit of quantitation | = | highest reflectance (R%) that is still statistically below the background reflectance (mean R% of negative samples minus three standard deviations of day‐to‐day analytical variation) |
R | = | reflectance, expressed as percentage (%) of maximum value |
SD | = | Standard deviation |
Acknowledgements
This study was financially supported by ARKRAY Inc, Japan. The authors associated with the company did not, however, plan the detail of the study format, nor influence the scientific statements obtained.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.