In a neonatal intensive care unit, a diagnosis that arrives in five weeks is often a diagnosis that arrives too late. Treatment decisions have already been made. Procedures have already happened. Sometimes the child has already died.
A program in Dubai has now published what happens when that timeline collapses. Across 100 critically ill infants and children, rapid whole-genome sequencing returned results in a median of 3.4 days. In a matched group of patients treated at the same hospitals before the program existed, standard genetic testing took a median of 38 days.
The results, published in Nature Medicine, come from Little Falcon, a citywide program run through Dubai Health across the centralized neonatal and pediatric intensive care units at Latifa Hospital and Al Jalila Children's Hospital. The fastest result came back in 47.7 hours.
Half the Children Got an Answer, and Half Got a Different Plan
Of the 100 patients, 53 received at least one molecular diagnosis related to their admission to intensive care, a diagnostic yield of 53 percent. In the historical comparison group, the yield was 30 percent.
The number that matters more clinically is a separate 53 percent. In 53 patients, the sequencing result led to a documented change in management. That included 45 of the 53 children who received a diagnosis and 8 of the 47 who did not, in whom ruling out a genetic cause narrowed the differential and stopped further invasive testing.
Broken down, the results prompted targeted drug or dietary intervention in 36 percent of the cohort, decisions about surgery or procedures in 13 percent, consolidation of the diagnostic pathway in 43 percent, subspecialty referrals in 37 percent, and a shift toward palliative care in 9 percent. In the historical group, testing changed management in 18 percent. The authors also report that sequencing altered the actual disease trajectory in 16 percent of patients, split evenly between favorable outcomes and end-of-life decisions.
The median age at presentation was 17 days. Eighty-seven percent of the children were under a year old. Sixty-two came from the neonatal unit, and 38 from the pediatric unit, and the diagnostic yield was nearly identical across both. Metabolic presentations produced the highest yield of any category, at 80 percent. Five children turned out to have two distinct genetic conditions at once, and another seven carried actionable findings unrelated to why they had been admitted.
The Cases Where Days Mattered
The paper documents specific trajectories, and they are the clearest argument for speed.
A two-month-old with abdominal distention, adrenal calcifications, and deteriorating clinical status was diagnosed with Wolman disease in 50.6 hours through a deletion in the LIPA gene. Enzyme replacement therapy started roughly two weeks before enzymatic testing would have confirmed it. Untreated infantile Wolman disease is characterized by rapidly progressive liver dysfunction and high mortality in the first year.
An 18-day-old with severe hypoglycemia requiring concentrated dextrose was classified within 55 hours as having focal ABCC8-associated disease, which pointed directly to surgery rather than prolonged empirical medical management.
A three-month-old admitted with infantile spasms turned out, within 83 hours, to have two separate conditions at once: thiamine metabolism dysfunction syndrome 2 and biotinidase deficiency. Both respond to oral thiamine and biotin.
And a five-month-old on diazoxide for presumed hyperinsulinemic hypoglycemia had negative sequencing results in 89.7 hours. That negative result ruled out the suspected genes, supported a diagnosis of transient nongenetic hyperinsulinemia, and led to the drug being stopped.
Why This Population Produced Such High Numbers
The cohort's genetics are unusual, and that is central to interpreting the yield. Parents came from 18 countries, predominantly from the Middle East and Asia. Genomic analysis found evidence of parental relatedness in 52 percent of patients.
Among children born to consanguineous parents, the diagnostic yield reached 80 percent, compared with 34 percent among children of unrelated parents. About 70 percent of the diagnoses were autosomal recessive conditions, most of which were driven by homozygous variants. In cohorts drawn from largely outbred populations, that ratio typically runs the other way, with dominant and de novo variants accounting for 50 to 85 percent of diagnoses.
This is the main caveat for American readers. Programs in the United States, Europe, and Australia operate in populations where reported consanguinity ranges from 3 to 20 percent, and the 30 to 60 percent yields reported across international efforts reflect a different genetic architecture. Little Falcon's 53 percent sits at the upper end of that range, and the 80 percent figure among consanguineous families does not transfer.
Cost, Which Has Been the Sticking Point
The authors report roughly 3,500 US dollars per case in their setting, benchmarked against American commercial laboratories, and 6,603 dollars per diagnosis. Earlier published estimates put rapid sequencing at a median of $ 9,239 per proband and $ 23,602 per diagnosis.
The study was not designed as a formal cost-effectiveness analysis, and the authors state this explicitly. They argue instead that sequencing cost is no longer the primary barrier, and that the remaining obstacles are infrastructure, trained staff, and referral pathways. They describe the test as functioning not only as a diagnostic tool but also as "a diagnostic safety net" in children whose presentations are too nonspecific to narrow the diagnosis on clinical grounds.
Several limitations deserve weight. The comparison group is historical rather than randomized, which means changes in care over time could contribute to the differences. Middle Eastern populations remain underrepresented in variant databases, which the authors say has likely pushed their yield below the true disease burden. And nonviable neonates were excluded by design, which the authors flag as a missed opportunity for families who might have wanted recurrence information.
Rapid genomic testing is expanding in the United States, and pooled analyses of critically ill infants support its diagnostic utility, but access remains uneven, and insurance coverage varies. Families facing an undiagnosed critical illness should ask their care team directly whether rapid sequencing is available and covered at their institution.
Key Questions Answered
What is rapid whole genome sequencing?
Sequencing a patient's entire genome, usually alongside both parents, with a workflow built for speed, so results reach the intensive care team in days rather than weeks.
How fast was Little Falcon?
A median of 81.1 hours, or 3.4 calendar days, from sample submission to reporting. The fastest case returned in 47.7 hours.
How many children did it help?
Fifty-three of 100 received a molecular diagnosis, and 53 of 100 had a documented change in clinical management. Those groups overlap but are not identical.
Does a negative result help?
Sometimes. In eight children, excluding a genetic cause narrowed the diagnostic possibilities and stopped further invasive testing.
Do these numbers apply to American hospitals?
Not directly. Just over half the families showed parental relatedness, which drove an unusually high proportion of recessive diagnoses. Yields in the United States would be expected to differ.
What are the main limitations?
The comparison group was historical rather than randomized, no formal cost-effectiveness analysis was performed, and underrepresentation of Middle Eastern genomes in reference databases may have suppressed the yield.
