Filtration pressure equilibrium: a statistical analysis.

Abstract

Statistical analysis of data in published micropuncture studies was applied to determine: a) the number of rats and samples of glomerular capillary pressure (Pg), end-capillary oncotic pressure (COPE), and proximal tubule pressure (Pt), needed in a single study to state with 95% confidence that a measured mean net efferent effective filtration pressure (delta P*E) is within +/- mmHg of any true value, delta PE; b) the effect of intra-animal variation in Pg, COPE, and Pt values on the number of experiments needed; and c) the probability of obtaining a value for delta P*E that meets this criterion in multiple trials. Assuming no intra-animal variation in Pg, Pt, and COPE values and employing published interanimal variances, at least 42 rats should be studied in any series to show that -1 mmHg less than or equal to (delta P*E -- delta PE) less than or equal to 1 mmHg within 95% confidence limits. If intra-animal variances are 60% of the interanimal variances, 67 rats are needed to obtain that degree of confidence when a single measurement of Pt, Pg, and COPE is made in each rat, and 47 animals are required with five measurements per rat. A value for delta P*E within 2-3 mmHg of its true value is, however, attainable with confidence using a far smaller number of animals per series.