As the first step of the joint study to standardize the presently used methods of the laboratory tests, the author investigated the data of quantitative examinations collected from various institutes. For this purpose, the same sample was first sent to the various national hospitals, and then the results obtained were collected. These results were evaluated with particular reference to their accuracy or precision, and in turn the standardization of the method of individual test was attempted.
The Investigation on the Present State.
1) The Method of Investigation;
Twelve items were selected for the study. Those include total protein, A/G ratio, blood sugar, sodium, potassium chloride, nonprotein nitrogen, urea nitrogen, total cholesterol, calcium, transaminase (GOT) and alkaline phosphatase. In the past two years, these tests were performed 3 times by the 2 sorts of common lyophilized sera.
2) The Criteria of Evaluation for the Data;
As a criteria of evaluating the data, accuracy or precision was applied. Namely, the percentages of the number of the institutes, from which assumably correct figures or ratios for the two samples were obtained (L or M), against the total numbers of institutes (N) were calculated. The former (L/N×100%) was referred as to accuracy and the latter (M/N×100%) as to precision respectively.
3) The Results and Discussion;
Using this criteria, the following results were obtained as to the accuracy and precision.
1: Comparing the data of 1963 and 1954, the accuracy was almost concurrent in the majority of the items with only ±10% difference.
2: The data of total protein, A/G ratio, total cholesterol, urea nitrogen, blood sugar (Hagedorn-Jensen's Method) and potassium revealed more than 50% of accuracy.
3: The data of blood sugar (Somogyi-Nelson's Method), alkaline phosphatase, nonprotein nitrogen, calcium and sodium revealed less than 50% of accuracy. Above all, the accuracy for sodium (especially for the sample of lower range) and GOT was low.
4: The accuracy of the hitherto applied methods for the determination of non-protein nitrogen was poor, Rappaport being 31% anb Kjeldahl-Nessler 27%. On the other hand, every method for urea nitrogen determination revealed about 70% of accuracy, and it is noteworthy that a simplified method (Unigraph) revealed 73% of accuracy, even better than the orthodox quantitative method.
5: As a rule, precision was better than accuracy. The reason for this is probably follows. Since precision is calculated from the concentration ratio between two samples, it may not be influenced so much by the difference in the concentration of the standard solution as accuracy.
6: This tendency was most remarkable in GOT. Therefore, it is well assumed that the low accuracy of the method of GOT determination is attributable not to the testing technic but to the concentration of standard solution or to the method itself.
7: Contrarily, the precision of the method for total protein determination is worse than its accuracy. This tendency is more remarkable in the data of 1964. This is probably due to the fact that the concentration of protein is not proportional to refractory index in its lower range.
The Standardization of the Method of Serum Urea Nitrogen Determination:
After the two year evaluation of the present state of laboratory examinations, in the third year (1955), the author attempted to establish the urease-indophenol method as a standard method for serum urea nitrogen determination.
1) The Selection of Item and Method;
There are two reasons that urea nitrogen was selected for the item of standardization. Firstly, the data of urea nitrogen are apt to be more accurate or precise than those of non-protein nitrogen which has been used in the similar clinical purpose so far. Secondly, it is more logical to determine urea nitrogen, the end product of metabolism, than to do non-protein nitrogen in the cases with renal
View full abstract