Grading of qualitative and quantitative responses in the PI proficiency survey of the College of American Pathologists for mailings in 1997–2000

The PI Survey offered jointly by the American Association of Blood Banks (AABB) and the College of American Pathologists (CAP) provides biological samples for proficiency testing that mirror paternity casework samples. A grading scheme providing for three levels of performance (good, acceptable, and unacceptable) has been developed for responses provided by subscribers. Reported here is a review of the performance of participants using RFLP and PCR/STR methods for the 1997–2000 period. In general, laboratories reporting phenotypes and paternity index values using RFLP methods exhibited a fairly constant rate of unacceptable responses over the 4-year period. In contrast, laboratories using PCR/STR systems improved dramatically in the frequency of unacceptable responses (an average of 3.5% for mailings in 1997 vs. 0.7% for mailings in 2000). A comparison of unacceptable responses for RFLP and PCR/STR methods over the 4-year period shows that whereas phenotype designations for PCR/STR systems is better than for RFLP systems, the opposite is true for reported paternity index values.