Computation of Life Expectation in Chronic Glomerulonephritis

Abstract

A multiple regression equation for the life expectation of patients with chronic glomerulonephritis was derived by the least-squares method from the laboratory investigations on 13 patients admitted to our clinic during 1962-65. Components of the regression were blood urea nitrogen, maximum urinary specific gravity, and hemoglobin (per cent, Sahli) obtained at the first examination, which had apparently the closest correlation with the life length of the patients. First-order regression reduced the dispersion already more than any of a single component, and second-order regression did the same further.
This regression equation is useful in: 1) Estimation of the severity of the patients' condition; 2) classification of glomerulonephritis into slowly and quickly progressing types, if such exist, by comparing the difference between two estimations made at different time and the actual lapse of time between the estimations; and 3) evaluating therapeutic efficacy.
The same regression equation (first-order) seems to be applicable to a wide range of the severity of glomerulonephritis, since it permits to predict changes in the maximum urinary specific gravity over the age from 20 to 80 years in Japanese subjects.