Journal club · Invited commentary

Probiotic-Associated Infection vs Necrotizing Enterocolitis

Weighing risks and benefits in the very preterm infant

Welcome, and thank you for joining. Since the FDA warning in 2023, most units quietly stopped giving probiotics to preterm babies — and that decision traded one risk for another. Over the next twenty minutes I want us to look at the largest safety study published so far, weigh it honestly against what NEC costs these babies every year, and then ask the harder question: what would any of this mean for our own unit?

Dr. Kashaf Qudoos PAEDS Medicine Unit-1 · Services Hospital, Lahore

14 slides · about 20 minutes · speaker notes included
Scottoline B, Jordan BK. JAMA Network Open. 2026;9(7):e2626535

Probiotics vs NEC Journal club · Dr. Kashaf Qudoos

Journal club · Invited commentary

Probiotic-Associated Infection vs Necrotizing Enterocolitis

Weighing risks and benefits in the very preterm infant

Scottoline B, Jordan BK.
JAMA Network Open. 2026;9(7):e2626535. Published July 31, 2026
Commentary on: Patel RM et al. JAMA Netw Open. 2026;9(7):e2626530

Presented by Dr. Kashaf Qudoos

The question on the table

Two risks, unequal in size

After the 2023 FDA warning, most US NICUs stopped giving probiotics to prevent NEC. That choice swaps one risk for another. These papers let us measure both.

!

The risk we acted on

Blood infection from probiotics

  • Live bacteria fed to a baby whose immune system is not yet ready
  • Bacteria could cross the leaky preterm gut into the blood
  • Reported in patients of all ages, not just newborns
  • Trials have never recorded safety data in a consistent way
!

The risk we accepted

Necrotizing enterocolitis

  • There is still no approved treatment for NEC
  • A severe inflammatory disease of the preterm gut
  • Babies who survive often have lifelong brain and gut problems
  • Some units that stopped probiotics now report more NEC

Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535.

The new evidence

Patel et al: the biggest safety study so far

20K

Nearly 20 000 very preterm babies studied — the largest such study published so far

<32

Babies born before 32 weeks made up the very preterm (VPT) group

30%

The Pediatrix NICU network covers about 30% of all NICU babies in the United States

The babies were given probiotics by mouth containing Lactobacillus and/or Bifidobacterium.

Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535.

Finding 1 — the safety signal

Infection from probiotics was rare

0.04%Rate of blood infection caused by probiotics in VPT babies
1 in ~2500Babies given probiotics developed a blood infection from them
1 in 19 900Deaths linked to probiotics in the whole group given them

These numbers match earlier, smaller studies and make the estimate more exact.

Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535.

Finding 2 — the same result elsewhere

Three separate studies agree

Study Babies studied Blood infection from probiotics Deaths linked to probiotics
Patel et al
2026
Very preterm, GA <32 wk
(~19 900 babies)
0.04%
(≈1 in 2500)
1 in 19 900
Alshaikh et al
2025
GA <34 wk
Extremely low birth weight <1000 g
0.14%
0.4% in ELBW
≈1 in 6300
Feldman et al
2025
Preterm babies
(meta-analysis)
<0.04% Potentially
1 in 20 323

Different networks, countries and methods — all landing on roughly the same number.

Patel et al 2026; Alshaikh et al, Pediatrics 2025;155(3):e2024069102; Feldman et al, Pediatr Res 2025.

The other side of the scale

What NEC costs every year

7–12%of VLBW babies affected by NEC
3500–6000babies per year in the US
~400deaths per year in the US alone
1 in 100death rate among VLBW babies

VLBW = very low birth weight (<1500 g). Most of these babies are also very preterm.

NEC is a severe inflammatory disease of the baby's gut, and there is still no approved treatment for it.

Wolf MF et al. JAMA Netw Open. 2023;6(3):e231511; Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535.

How NEC happens

Why gut bacteria are a sensible target

1

Immature baby

The preterm immune system reacts too strongly. The gut expects almost sterile amniotic fluid.

2

Wrong bacteria

The gut fills with bacteria picked up in the NICU, with too many Proteobacteria.

3

TLR4 switched on

TLR4 binds bacterial lipopolysaccharide and sets off heavy inflammation in the gut lining.

4

Damage spreads

Immune cells activate, blood vessels narrow, and the gut nerves stop working properly.

5

Severe NEC

The gut can perforate, leading to peritonitis, sepsis, shock and sometimes death.

Because of this, NEC in VLBW babies is often called “dysbiotic NEC” — one type of preterm gut injury.

Cuna A et al. Front Pediatr. 2024; Hackam DJ, Sodhi CP. Nat Rev Gastroenterol Hepatol. 2022.

Beyond death rates

Survival is not the end of the story

Babies who survive NEC live with the effects for life. Death rates alone hide how much harm it does.

Brain

White matter brain injury and delayed development

Gut

Long-term gut problems, including short gut syndrome

Lungs

Higher risk of chronic lung disease

Family

A heavy emotional toll on families that we rarely measure

Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535.

What the FDA did

How US practice changed in 2023

Before 2023

Up to 40% of US NICUs were giving probiotics to prevent NEC

September 2023

Growing use, plus scattered reports of infection and possible death, led the FDA to warn against giving live bacteria to preterm babies outside a formal research licence

After the warning

Probiotic use in US preterm babies almost completely stopped

Since then

Some units that stopped report more NEC — but we do not know whether this holds true everywhere

US Food and Drug Administration, September 29, 2023; Scottoline B, Jordan BK. JAMA Netw Open. 2026.

The central comparison

How much more common is NEC?

NEC compared with blood infection from probiotics

Lowest estimate30×
Highest estimate300×

Death from NEC compared with death from a probiotic infection

Lowest estimate60×
Highest estimate200×

Bottom line

NEC happens roughly 30 to 300 times more often than blood infection caused by probiotics.

Death from NEC is roughly 60 to 200 times more common than death from a probiotic infection.

Comparison derived from Patel et al 2026, Alshaikh et al 2025 and Feldman et al 2025.

What it means

Where the balance sits

1

Does it work?

More than 36 randomised trials disagree in the detail, but point the same way: probiotics may lower NEC

2

How risky is it?

Even in fragile very preterm babies, the risk of harm stays low

3

Which is bigger?

The disease we prevent is 10 to 100 times more common, and more deadly, than the harm we fear

Sharif S et al. Cochrane Database Syst Rev. 2023; Scottoline B, Jordan BK. JAMA Netw Open. 2026.

Caveats and next steps

What this commentary does not settle

Remaining limitations

Not a trial

This is registry data, not a randomised study designed to measure safety

Patchy reporting

Probiotic trials have never collected safety data in the same way

Products differ

Strains, mixtures and manufacturing quality vary widely between brands

What the commentary calls for

Big trials

Trials large enough to test whether probiotics really prevent NEC in VLBW babies

Approval as a drug

Licensing them as drugs would bring quality standards and proper oversight

Ongoing monitoring

Tracking patients after approval to sharpen the risk numbers

Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535.

Bringing it home

Could we run this at Services Hospital, Lahore?

The FDA warning does not apply in Pakistan — but we have the same missing evidence.

A realistic staged pathway

1

Start a registry

Track NEC and late sepsis in our VLBW babies for 6–12 months. This gives us the baseline numbers we do not have.

2

Small safety study

A small group given probiotics, watched closely for blood infection. Checks we can spot the harm before testing the benefit.

3

Multi-centre trial

Testing NEC properly needs more babies than one unit has. Only possible with other Punjab tertiary NICUs.

Prerequisites we must confirm first

How many babies?

Our yearly VLBW and very preterm admissions, and our own NEC rate — to be taken from unit records

Can the lab detect it?

Can our lab grow and identify Lactobacillus and Bifidobacterium from blood? Anaerobic culture is the main hurdle

Can we get the product?

A probiotic with known strains, batch checks and an unbroken cold chain — the hardest part to arrange

Approvals

DRAP status of the product, ethics committee approval, and a clear consent process

We cannot answer whether probiotics work on our own. But we can answer the first question nobody here has answered: what is our own NEC rate?

Local feasibility — figures still to be confirmed from unit records.

Closing

Scottoline and Jordan end by repeating a parent's point: infection from probiotics is a real danger, but a rare one. NEC is by far the more likely threat to these babies.

A point made by Jennifer Canvasser — mother of a baby who died of NEC, and founder of the NEC Society

For discussion

  • Do we know our own NEC rate well enough to even have this conversation?
  • Would the group support starting a prospective VLBW registry this year?
  • How would we counsel families about a 1-in-2500 risk against a 1-in-100 one?

Scottoline B, Jordan BK. JAMA Netw Open. 2026;9(7):e2626535. Full reference list available on request.

All slides