Abstract
The Lindholm test, a modified whole blood clotting time test designed to permit estimation of the blood compatibility of biomaterials, has been analyzed statistically with regard to sources of variance and desired sample sizes. Of five materials tested for compatibility with blood, only Cuprophane had an among-donor component of variance significantly different from zero. Analysis indicated that, under certain conditions, performance of 11 Lindholm tests upon each of two materials can be expected to furnish sufficient data for an observed average difference of 5 min in clotting time to be significant at the 5% level; performance of 36 pairs of tests gives a 98% probability of detecting a true difference of 5 min when testing at the 5% level of significance.