A New York study added genome sequencing to newborn screening. 3.7% screened positive

Every newborn in the United States already has a few drops of blood taken from the heel and screened for a panel of serious, treatable conditions. The GUARDIAN study asked what happens if you add genome sequencing on top — and reported its first results, from 4,000 babies in New York City, in JAMA [s1].

What the study tested

GUARDIAN — the Genomic Uniform-screening Against Rare Disease In All Newborns study — is a multisite, single-group, prospective, observational investigation run within the New York State Department of Health's newborn screening programme [s1]. Its full planned enrollment is 100,000 participants [s1]. The paper reports a prespecified interim analysis of the first 4,000 newborns, recruited at six New York City hospitals between September 2022 and July 2023 [s2][s1].

The screen does not sequence and interpret the whole genome blind. Investigators selected a defined list: 156 early-onset genetic conditions with established interventions were screened in every participant, and a further 99 neurodevelopmental disorders associated with seizures were offered as an option [s1]. The primary outcome was the screen-positive rate — how often the test flagged a condition — alongside how many families agreed to take part and how often sequencing succeeded [s1].

What it found

Over 11 months, 5,555 families were approached and 4,000, or 72.0%, consented to take part [s1]. That uptake is a central finding in its own right: a large majority of parents, offered a genomic screen for their newborn, accepted it.

The enrolled group was deliberately diverse. By parent-reported race, participants were 0.5% American Indian or Alaska Native, 16.5% Asian, 25.1% Black, 0.1% Native Hawaiian or Other Pacific Islander, 44.7% White and 13.0% two or more races; by ethnicity, 44.0% were Hispanic and 56.0% not Hispanic [s1]. Among those who enrolled, 90.6% consented to screening for both groups of conditions, and 9.4% to the disorders with established interventions only [s1].

Sequencing was technically robust: testing was successfully completed for 99.6% of cases [s1]. The screen-positive rate was 3.7%, and that figure included treatable conditions not currently covered by standard newborn screening — the main thing the genomic layer is meant to add [s1].

How to read it

The careful word in the study's own conclusion is "feasibility." GUARDIAN shows that a predefined set of genes can be interpreted from genome sequencing, in a racially and ethnically diverse population, at scale and with very high technical success [s1]. It does not yet show that doing so improves health.

That distinction matters. A 3.7% screen-positive rate is far higher than the rate at which traditional biochemical screening flags disease, and some of those positives will be for conditions that would never have caused harm, or that cannot yet be acted on effectively — the familiar trade-off of any expanded screen between finding more and worrying more [s1]. The investigators are explicit that additional studies are needed to learn whether the findings generalise to other populations, and whether adding sequencing actually changes how children are managed and whether it improves outcomes [s1].

The design also builds in parental choice in a way worth noting: the split between the 90.6% who opted into both condition groups and the 9.4% who took only the established-intervention set shows that families were given, and used, a real decision about how much they wanted to know [s1]. That choice matters because the optional tier — 99 neurodevelopmental disorders associated with seizures — covers conditions that are harder to act on than the 156 with established interventions, and GUARDIAN let parents screen for one without the other [s1].

The diversity of the enrolled group is also more than a demographic footnote. Genomic medicine has a documented history of performing worst for the populations least represented in its reference data, so a screen validated largely in one ancestry group can misfire in others [s1]. By recruiting a cohort that was 25.1% Black, 16.5% Asian and 44.0% Hispanic, and by reporting success across it, GUARDIAN is testing whether a sequencing-based screen can work equitably rather than assuming it will — though its authors caution that these interim findings still need to be confirmed in other populations [s1].

What to watch

GUARDIAN is still enrolling toward its 100,000-participant target, so these interim numbers may move [s2][s1]. The questions that will decide whether genomic newborn screening becomes routine are the ones this interim report cannot yet answer: what the positive results lead to, how often a flagged condition changes a child's care, and whether the children screened are measurably better off years later [s1]. Feasibility is the necessary first step, not the finish line.

This article describes research and is not medical advice. Decisions about newborn screening are for families and their treating clinicians.