Two large programs are testing whether a baby's genome can catch treatable conditions that heel-prick screening misses. Here are the published numbers and the open questions.
Newborn genome sequencing means reading a baby’s DNA at birth to look for genetic conditions that are treatable if caught early. Two large programs are testing it: the GUARDIAN study in New York and the Generation Study in England. In New York’s first 4,000 babies, sequencing flagged 120 true cases of treatable conditions, and standard screening had caught only 10 of them. England reports more than 85,000 babies recruited, with about 1 percent receiving a “condition suspected” result.
Those are early numbers from programs still running, and they raise as many questions as they settle. Below is what each program has published, how sequencing differs from the heel-prick blood spot test, and what remains open. This is general information, not medical advice.
How genome screening differs from the heel-prick test
Standard newborn screening takes a few drops of blood from the baby’s heel and measures chemicals or enzyme activity that signal specific disorders. Each state or country picks its panel. In New York, according to Columbia’s account of the GUARDIAN results, the standard program detects around 60 conditions.
Genome sequencing reads the DNA itself. It can look for disease-causing changes in hundreds of genes at once, including conditions for which no good blood marker exists. It can also fail in the other direction: it may miss a condition that the biochemical test would catch, because not every disease is caused by a change a genome test can read. Genomics England is blunt about this, stating that its study “is not a replacement” for the heel-prick test and that sequencing does not detect all forms of the conditions the blood spot test covers.
| Heel-prick blood spot | Genome sequencing | |
|---|---|---|
| What is measured | Biochemical markers in a blood spot | DNA variants, often from cord blood or a blood sample |
| Number of conditions | Around 60 in New York’s standard panel | 156 treatable in GUARDIAN’s first analysis; 450+ now; 200+ in England |
| Status | Routine public health program | Research studies with parental consent |
| Main weakness | Limited to what a biochemical test can see | Variants of uncertain meaning; cost; data and privacy questions |
GUARDIAN in New York: the JAMA results
GUARDIAN is run by Columbia University Irving Medical Center with NewYork-Presbyterian and partners that include GeneDx, Illumina, the New York State Department of Health and Sanofi, as listed on the study’s site. Parents enroll their babies voluntarily and for free.
The first report appeared in JAMA on October 24, 2024. It covered the first 4,000 newborns, born between September 2022 and July 2023. Researchers analyzed 156 treatable conditions, and parents could add 99 untreatable ones, which 90 percent chose to do. Of 147 babies who screened positive, 120 were confirmed as true positives, about 3 percent of the babies tested.
The detail that drew attention: only 10 of those 120 would have been detected by standard screening. Of the 110 others, 92 were G6PD deficiency, a condition standard New York screening does not test for, and one was severe combined immunodeficiency, treated with a bone marrow transplant. Setting G6PD aside, the positive rate was 0.6 percent, versus 0.3 percent for standard screening, a figure the authors call a floor that will rise as conditions are added.
Columbia’s summary states the cautions plainly: the results are preliminary, costs would rise at scale (about 7,700 positive screens a year if applied to New York’s roughly 210,000 births), and who would pay is unsettled. Interpretation across diverse ancestries is an open problem. The study is funded in part by companies whose employees co-authored the paper, a conflict-of-interest point worth knowing when weighing the enthusiasm.
England’s Generation Study: scale and early signs
The Generation Study, led by Genomics England with NHS England, aims to sequence the genomes of 100,000 newborns at participating hospitals. It looks at changes linked to more than 200 rare conditions that usually appear in the first few years of life and can be treated through the NHS. Cord blood is the main sample.
According to a Genomics England news item of August 21, 2026, more than 85,000 participants have been recruited at over 70 hospitals, more than 64,000 results have been returned, and more than 200 “condition suspected” results have gone to families. The study launched in 2024 and is scheduled to run until 2027. On the results page, the program says about 1 percent of babies get a suspected result and a suspected result is not a diagnosis; it must be confirmed.
The same news item tells one family story. A baby enrolled in the study was found to carry a gene change linked to adrenoleukodystrophy, a rare and often life-limiting disease, and testing then showed his four-year-old brother carries the same change. Both boys now have regular blood tests and, from age two, MRI scans so that early changes can be treated. It is a single anecdote, not evidence of benefit at population scale, but it shows the kind of result the program has in mind: a finding that changes follow-up before symptoms exist. Whether such stories add up to better health across a whole population is what the 2027 end date and any NHS review are meant to start answering.
The hard questions that remain
Does earlier detection improve outcomes?
For some conditions, plainly yes: severe immune deficiency treated early has a very different course than one found after infections. For others the answer depends on whether an effective treatment exists and when it must start. Neither program has yet reported long-term health outcomes for the babies it screened, and I found none.
What about uncertain results?
A genome contains many variants whose meaning is unclear. Programs choose gene lists and filtering rules to limit this, but some babies will get a flag that later proves harmless, along with the worry and follow-up testing that comes with it. Reference databases are still weighted toward people of European ancestry, which affects how confidently variants can be classified for everyone else.
Who owns and stores the data?
Genomics England states that data and samples are held in its National Genomic Research Library, where approved researchers can use them without seeing participants’ identities. The page also says the study is exploring the risks and benefits of keeping a person’s genome over their lifetime, and links to a separate ethics discussion covering consent, withdrawal and recruitment. The details of those policies are worth reading before enrolling.
Could it become routine?
Both programs describe themselves as research. England’s roadmap includes an NHS review of the study to inform future decisions on services, with no date given. In the US, GUARDIAN’s authors point to cost and payer questions that have to be settled first.
How results reach families
Timing matters, because some conditions need action within days. According to Genomics England, families of babies with a suspected condition are contacted by an NHS specialist team within weeks of birth, while parents of babies with no condition suspected hear a few months after birth. The suspected result then goes through confirmatory testing before anyone calls it a diagnosis.
The reassuring message needs care too. England expects over 99 percent of babies to receive “no condition suspected,” but that phrase is narrower than “healthy.” A genome screen looks at a defined list of genes. It does not rule out other genetic conditions, non-genetic illness, or conditions caused by variants the test cannot read. The same logic applies to heel-prick screening, which is one reason both programs stress that sequencing sits beside the standard test and does not replace it.
What to ask if you are offered enrollment
- Which conditions are on the list, and are any without a current treatment?
- Who will see the data, for how long, and can I withdraw it later?
- How and when will results come back, and who explains a positive one?
- What follow-up testing is covered if the result is only suspected?
A short timeline
- 2021: England holds a national public dialogue that shapes the Generation Study vision.
- 2022: England’s Baby and Mum Samples Study collects samples from 600 mothers and babies, leading to cord blood as the main sample. GUARDIAN begins enrolling in New York.
- October 24, 2024: GUARDIAN’s first 4,000-baby results appear in JAMA.
- 2024 to 2027: The Generation Study recruits and returns results.
- August 2026: Genomics England reports more than 85,000 recruited.
What parents and readers should take from this
The early numbers support a narrow claim: sequencing finds treatable conditions that the current blood spot test does not look for, in roughly three percent of babies in one New York cohort, at a cost and with an interpretation burden that are not yet resolved. They do not show that routine genome screening improves health overall, and they do not say which conditions belong on a national panel. Families who are offered enrollment can ask the study team what conditions are covered, what happens to the data, and how results are returned.
Sources
- Columbia University Irving Medical Center, GUARDIAN study results (JAMA, October 24, 2024)
- The GUARDIAN Study
- Genomics England, Newborn Genomes Programme and Generation Study
- Genomics England, Generation Study results
- Genomics England, Baby’s ALD gene discovery leads to diagnosis for older brother (August 21, 2026)
