The estimation of creatinine in serum.

Although the clearance of endogenous creatinine as a measure of the glomerular filtration rate has been much criticized, it remains the only possible measure where repeated or continuous investigation is needed, and has already yielded much useful information. When we started to use it, however, it became apparent that while the estimation of creatinine in urine was satisfactory, that of plasma creatinine was not. The purpose of this work was therefore to establish a standard method of estimation simple enough for routine use. When this had been evolved it was used for a series of creatinine clearances, which were compared with simultaneous inulin clearances to see whether a reasonable agreement was apparent. The basis of nearly all methods of estimation is the colour of the alkaline creatinine picrate (Jaffe, 1886) but it does not seem to have been realized how extremely sensitive this colour is to slight changes in media. The reaction was studied in detail by Bonsnes and Taussky (1945) and they outlined a procedure for urine which has not been questioned, but they only investigated a few blood filtrates and no serum or plasma. Some of the points they established are, however, important and may be summarized as follows. First the amount of coloured creatinine compound and the rate of its formation vary inversely with the amount of alkali present. Secondly, as a converse to this, the colour of the alkaline picrate, in which the creatinine compound is in solution and which is used as a blank in reading the colour, increases with the amount of alkali present. Thirdly, the colour response does not strictly follow De Beer's law, being proportional to concentration only in high dilutions. It follows from these findings that the composition of the alkaline picrate added to the creatinine solutions must be controlled, and this point has been studied, the most recent work being that of Lauson (1951). But it also follows that the reaction of the medium in which the creatinine is present may influence the colours of both the creatinine compound and the alkaline picrate by altering the proportions of alkali and picrate. This is not important when urine is the medium, as the estimation is made in dilute solutions approaching neutrality, but filtrates from blood plasma or serum may be sufficiently acid to have a significant effect. The preparation of filtrates from blood has generally been by the method of Folin (1919), using sodium tungstate and sulphuric acid. This involves a final dilution of 1 in 10 and is designed to give filtrates approaching neutrality. Brod and Sirota (1948) introduced two modifications, the dilution being reduced to 1 in 4 and the acid used being twice as strong, so that the filtrates are more acid. Another method is that of Popper, Mandel, and Mayer (1937) in which the proteins are precipitated with picric acid and NaOH alone added to the filtrate; the values found by this method tend to be lower than with the tungstate method, and it was pointed out by Smith, Finkelstein, and Smith (1940), who used it, that the amounts of chromogenic substances estimated vary markedly when different methods of precipitation are used. The matter is further complicated by the facts that the medium, even when neutral, may affect the colour, a possibility which does not appear to have been realized, and that plasma contains chromogens other than creatinine which give a colour with alkaline picrate. The following points were investigated: (1) The colour of blank solutions of alkaline picrate; (2) the colour of alkaline creatinine picrate in different media; (3) the time taken for the development of the colours of alkaline picrate, alkaline creatinine picrate, and picrate of other chromogens; (4) the relationship between creatinine and inulin clearances.