Serum pooling lowers cholesterol recovery.

Matrix effects hinder the transferability of accuracy for cholesterol. A general assumption is that pooled and individual samples yield similar results. We tested the hypothesis that serum-pooling affects the recovery for cholesterol. We pooled 100 serum samples, determining cholesterol of pool and of the individual samples with Hitachi 717 and 914. Over twenty days, we daily determined cholesterol of individual and pooled samples, using a Hitachi 736 and 747 analyzers. For the hundred-sample pool, the pool was 1.1 to 1.5% lower than the individual samples. With the daily pool study, the ratio of 747 to 736 was 1.7% lower for the pooled compared with the individual samples. Therefore, pooling of serum samples causes a decreased recovery, averaging from 1.1-1.7%, and representing 37-57% of the allowable bias for cholesterol (< 3%), and it is thereby significant.