Analytical evaluation of five methods for quantitation of urine protein

Abstract EN

This study represents an analytical comparison between five methods for quantitating urinary protein.

These included : biuret method following protein precipitation using trichloroacetic acid, Coomassie brilliant blue dye binding method, ponceau S dye binding method, sulphosalicylic acid turbidimetric method and triethanolamine-ferric chloride method following protein precipitation by tannic acid.

The parameters which were assessed for each method included its linearity, sensitivity ( detection limit ), precision ( within-batch and between-batch), accuracy ( allowable limit of error and percentage error ) and comparability.

The linearity response was found to be up 1.6 g/l for tannic acid-ferric chloride, 2.0 g / l for Coomassie brilliant blue and sulphosalicylic acid, 2.1 g/l for ponceau S and 2.3 g/l for biuret method.

The sensitivity, as reflected by the detection limit, was found to be 0.013 g/l for ponceau S and tannic acid – ferric chloride methods, o.025 g/l for Coomassie brilliant blue and biuret methods, and 0.05 g/l for sulphosalicylic acid method.

The within – batch imprecision for low protein concentration ranged from 6.9 % for Coomassie brilliant blue to 9.99 % for ponceau S and for high concentration it ranged from 2.47 % also for Coomassie brilliant blue to 4.06 % for biuret ad tannic acid methods.

The between – batch imprecision for low concentration ranged from 7.43 % for Coomassie brilliant blue to 13.58 % for ponceau S, and for high concentration it ranged from 3.89 % for Coomassie brilliant blue to 4.68 % for tannic acid method.

The allowable limit of error ranged from 17.39 for sulphosalicylic acid to 19.8 % for tannic acid and the percentage error for low protein ranged form 9.40 for sulphosalicylic acid to 19.07 for ponceau S and for high concentration it ranged from 7.67 for sulphosalicylic acid to 12.26 for ponceau S.

Comparability study revealed highly significant difference in mean protein values at low protein concentration but no significant difference at high concentration.

In conclusion, all methods showed limited linearity with better sensitivity, precision and accuracy at high than at low protein concentration.

The highest linearity was that of biuret method, while tannic acid – ferric chloride method was the least linear.

The best sensitivity was obtained by ponceau S and tannic acid – ferric chloride methods, whereas sulphosalicylic acid was the least sensitive.

The Coomassie brilliant blue method combined acceptable linearity, sensitivity and appeared to be the most precise and accurate at both protein concentrations, in addition to its practicability.