Triple-marker test as screening for Down syndrome: a meta-analysis.

The effectiveness of triple-marker testing as screening for Down syndrome needs to be evaluated by means of formal meta-analytic techniques. We did a MEDLINE search to identify studies evaluating the detection of Down syndrome by use of the triple-marker test. Reference lists of articles were also checked. Papers published in either English, French, or German from 1966 to November 1996 were eligible for this review. Twenty cohort studies were identified. Results of sensitivities and false-positive rates from different subgroups of the study sample were compared by using summary receiver-operating characteristic curve analysis. Medians of sensitivities and false-positive rates were also estimated. A total of 194,326 patients were included. In women of all ages, the medians for sensitivities were 67, 71, and 73 percent when the cutoffs used were 1:190-200, 1:250-295, and 1:350-380, respectively. The median false-positive rates fluctuated between 4 and 8 percent. For women at or above 35 years old, the medians of sensitivity and false-positive rate were 89 and 25 percent, respectively, when the chosen cutoff was 1:190-200. In patients below 35 years old, the median sensitivity was 57 percent if the cutoff used was 1:250-295. Summary receiver-operating characteristic curves showed that 1:190 was the best cutoff for predicting Down syndrome. The triple-marker testing is an effective screening method of detecting Down syndrome pregnancies. It is less effective in younger than in older age groups and may be offered as an alternative to amniocentesis to pregnant women over 35.