[Rationalization of the quantitative determination of fecal fat]

The stool wet weight/24 h correlated well with chemical fecal fat/24 h (n = 45, p less than 0.0001). The best and simpliest methods for stool homogenization were treatment of stool with starmix or stool dilution with water and shaking for 5 min. The coefficient of variance of one of this methods was as good as by the combined method. The coefficient of variance of fecal fat was significantly lower in homogenized stool specimens than in unprepared stool samples (p less than 0.01), but fat estimation in samples of not homogenized stool (n = 10) of normal persons also showed a significant difference in comparison with samples of stool (n = 15) from patients with malassimilation (p less than 0.01). The titrimetric method for estimation of fecal fat/24 h (van de Kamer) correlated very well with a simple convenient colorimetric method (Tomaszewski) (n = 52, r = 0.979, p = 0.001). There was no significant difference in fat excretion/24 h in two groups of patients with (n = 26) and without fat balance (n = 42).