In This Issue...
More Newborn Screening Tests, More False Positives
The American College of Medical Genetics (ACMG) recommends that every baby in the U.S. be screened for a uniform panel of 29 disorders to identify rare, serious metabolic diseases. This recommendation is endorsed by the March of Dimes and American Academy of Pediatrics, but currently only the District of Columbia and five states—Iowa, Maryland, Mississippi, New Jersey, and Virginia—require that newborns be screened for the full panel, which includes using tandem mass spectrometry (MS/MS) technology to screen for more than 20 disorders at one time. While there is pressure on states to adopt this uniform screening panel, a recent study in Pediatrics warns that an increase in mandatory screening tests performed on newborns poses its own risks, including an increase in the number of false positive results. This issue of Strategies examines the implications of these findings for the clinical laboratory, and what measures need to be taken as MS/MS newborn screening expands nationwide.
Almost 67% of babies born in the U.S. this year will be screened for more than 20 disorders, based on the fourth annual March of Dimes Newborn Screening Report Card released this summer. This figure is nearly twice the 2005 estimates, when 23 states—covering 38% of U.S. babies—screened for more than 20 conditions. The jump in screening rates this year is attributed to an increasing number of states mandating that babies be screened for an expanded panel of disorders. By the beginning of June, eight more states—accounting for 64% of babies—will have initiated screening for more than 20 disorders (see Figure 1).
Figure 1. Increase in newborn screening panels by state, 1995-2005†
† Alaska (34 disorders added), Hawaii (38 disorders added) and District of Columbia (1 disorder added) are not graphically represented.
But while newborn screening allows clinicians to identify severe metabolic disorders, there are some concerns about erroneous results. “With the expansion of newborn screening we hope to improve the health of newborns,” said Beth Tarini, MD, MS, Clinical Lecturer with the University of Michigan’s Division of General Pediatrics in Ann Arbor. “However, we also need to be mindful of the fact that additional testing comes at a cost because more tests produce more false positive results.”
Tarini co-authored a study examining data on screening practices from the National Newborn Screening and Genetics Resource Center (Austin, Texas) published in the August 2006 issue of Pediatrics (118:448-456; www.pediatrics.org/cgi/doi/10.1542/peds.2005-2026). Looking at the number of mandated disorders added to state newborn screening panels between 1995 and 2005, Tarini and her colleagues estimated that more than 51,000 infants would have received false positive results through MS/MS screening in 2005, if the test specificity was 99.9%. When the researchers assumed a specificity of 99.995%, that estimation went down to approximately 2,575 infants who would have received false positive results. But test specificity is only one aspect of this false positive issue, said Tarini. In spite of high specificity values—such as when using MS/MS testing technology—testing for multiple rare conditions will increase the number of false positive results. “As technological advances increase our ability to test for many more disorders, we must also understand the potential public health implications and health outcomes that arise as we translate these technologies into widespread public use,” she added.
Based on these findings, the authors recommended that a cautious approach is needed as state newborn screening programs continue to expand. “The goal of a screening program is to balance the benefits of early detection and the harms of false results,” explained Tarini. “States should be aware that more testing always generates more false positives. As a result, states with expanded newborn screening should be attentive to the number of false positive results and the experiences of the families who receive them.”
Study Lacking Validation Data
This study focused on estimating how many newborns received false positive results in 2005, but there is no validation data on actual false positive results by state, explained Piero Rinaldo, MD, PhD, co-director of the Biochemical Genetics Laboratory at the Mayo Clinic (Rochester, Minn.). In early 2005, Rinaldo was a co-author of a report made to the U.S. Department of Health and Human Services recommending that the U.S. comply with ACMG’s uniform panel proposal.
This lack of data is distressing, especially given how far off base the estimations were for the state of Minnesota, noted Rinaldo. “If they had asked states about actual false positive rates in 2005, then these results would have been based on a predictive model compared to the reality,” he explained. “In our case, this is 36% off. Our number of false positives in 2005 by MS/MS was 52, but in this study’s best case scenario, they predict 81. So we are 36% better than their best case scenario.”
Rinaldo, who has co-authored a paper on objective performance metrics in MS/MS newborn screening currently undergoing review by a medical journal, also takes issue with the study’s reliance on sensitivity and specificity as performance metrics, rather than other measures such as positive predictive value and detection rate that can be objective and immediately appreciated. “Specificity and sensitivity don’t quite do that,” he said. “When I hear people complain about these false positive rates, I ask them about their detection rate and positive predictive value.”
For more information
- Additional information on the National Newborn Screening and Genetics Resource Center can be found online at http://genes-r-us.uthscsa.edu/.
- A full copy of the 2006 Newborn Screening Report Card can be found on the March of Dimes home page, www.marchofdimes.org