Updated March 12, 2012
One of the College of American Pathologists' Chemistry and Toxicology Checklist questions (CHM.32200 Phase II) for laboratories performing prenatal screening are whether the accuracy of the risk calculation for Down syndrome has been verified.
One approach to verifying this calculation is to use data contained in the multiple marker evaluation report that is part of the CAP Maternal Screening FP Survey. For each PT serum specimen, the laboratory’s own reported maternal age and multiples of the medians (MoM) values (AFP/uE3/hCG or AFP/uE3/hCG/DIA) are used to calculate a Down syndrome risk. This computed risk can then be compared with the actual risk reported by the laboratory.
In order to compute reliable Down syndrome risks, it is important to collect accurate estimates of maternal age and gestational age, accurately and precisely measure the needed analytes and convert those measurements using appropriate medians into multiples of the median (MoMs). The final step is combining this information into a patient specific risk for Down syndrome. This last step requires the use of published ‘parameter sets’ that mathematically describe the multi-dimensional relationships between these markers in Down syndrome and unaffected pregnancies. Without suitable parameter sets, the Down syndrome risks will not be as reliable as they could be in screening programs.
All laboratories should understand the importance of selecting and maintaining the parameter sets used for computing clinical Down syndrome risk estimates. The CAP website has been updated to focus on the best practices used when selecting parameter sets. Laboratories using the Wald et al., 1988, Wald et al., 1992 and Wald et al., 2000 parameter sets should consider updating to Knight et al. 1998, or SURUSS 2003 parameter sets. Updating parameter sets ensures that your laboratory is using the most reliable data for interpretation and allows for reported Down syndrome risks to become more harmonized both within and between laboratories. Additionally, using older parameter sets may not allow for quadruple testing, transformation of uE3 measurements to a logarithmic scale or have other weaknesses.