The use of record linkage for auditing the uptake and outcome of prenatal serum screening and prenatal diagnostic tests for Down syndrome.

Abstract

To pilot the use of linked routine records for auditing Down syndrome prenatal serum screening and diagnostic tests. The cohort studied were 110 272 patients of 4 London maternity units that offered the Bart's maternal serum tests any time between 1990 and 1999. Audit was based on linked data derived from obstetric records, referral data on maternal serum screening and/or prenatal diagnoses. Cytogenetic reports without matching obstetric data were retained in the cohort as they included fetal deaths or terminations. (1) Significant independent influences on uptake of serum screening (58% overall) were maternal age, ethnicity, year and referring hospital, and those on uptake of prenatal diagnosis (4% overall) were screening result (54% uptake after positive screen), maternal age, year and referring hospital; (2) detection, false-positive rates and odds of being affected after positive results were respectively 49%, 4% and 1 : 59 between 1990 and 1994, and 78%, 7% and 1 : 58 after 1994. Using maternal age alone (cut-off > or =37 at delivery), these would have been respectively 40%, 7% and 1 : 96 between 1990 and 1994, and 40%, 9% and 1 : 107 between 1995 and 1999. Ongoing audit of DS prenatal programmes could be derived from computerised maternity data sets if they included fetal deaths, and relevant laboratory and ultrasound findings.