Counting the Babies We Miss: Why Congenital Syphilis Is More Common Than the Official Numbers Suggest

Vaz and colleagues in Portland and Flores and colleagues in Chicago show how infections slip past testing before birth and past the record books afterwards

Why a very old infection has become a new problem again

Syphilis is a treatable infection that can pass from a pregnant person to their baby during pregnancy. Cases in the United States are rising again, and two new American studies suggest the true number of affected babies is higher than official counts show — some infections are found too late in pregnancy, and many treated newborns are never reported at all. Both gaps have practical fixes, and the treatment itself, penicillin, works very well.

When a baby is infected before birth, doctors call it congenital syphilis — "congenital" simply meaning present from birth. It is one of the few serious newborn infections medicine already knows how to prevent completely, because penicillin given to the pregnant person in good time almost always stops the infection reaching the baby. Preventing it depends not on a new drug but on a chain of ordinary steps: testing early in pregnancy, testing again if there is continuing risk, treating properly, checking the newborn, and telling the local health department so the case is counted.

How this problem was almost solved, and then came back

For much of the past thirty years, congenital syphilis in the United States was becoming rare. That changed after 2012, when national records showed the number of affected babies rising again, alongside a rise in syphilis among women of childbearing age [1]. Researchers then looked closely at why. Reviewing cases from 2018, they sorted each one by which step had broken down [2], and a larger 2023 study of cases from 2022 found the same pattern: most affected babies were born to people who had attended prenatal care but had either not been tested at the right time or not been treated adequately [3]. These were not families who had vanished from the health system. They were families the system had already met.

The stakes are high. Untreated syphilis in pregnancy substantially raises the risk of stillbirth, premature birth and death in the first weeks of life [4], and part of the difficulty is that syphilis is often silent — an adult can carry it for a long time with no symptoms they would notice [5]. Because of this, national guidance from the Centers for Disease Control and Prevention (CDC), the United States government's public health agency, tells doctors to decide how to treat a newborn based mainly on the mother's test results and on whether her treatment was adequate and early enough — not on the baby's own blood test [6]. Pregnancy guidance has also stressed for years that one test early on cannot catch an infection acquired later [7].

There is a useful comparison here. HIV — human immunodeficiency virus, the virus that causes AIDS — can also pass from a pregnant person to a baby, and decades ago it often did. Today that is rare in American hospitals, because everyone is offered an HIV test in pregnancy as a routine part of care and effective medicines are given to anyone who tests positive. Syphilis is, if anything, easier to stop — penicillin has worked against it for over seventy years — but it has never had the same routine, repeated testing behind it.

What the two new studies looked at, and where

The first was carried out by Vaz and colleagues at Oregon Health & Science University, a large university hospital in Portland, Oregon [8]. They went back through every syphilis test their laboratory had run over seven years (2018 to 2024) and every HIV test over four years (2021 to 2024) — 67,800 and 56,727 tests respectively — along with the records of 13,889 pregnancies. Their question was simple: are the systems built around these two infections actually working?

The second study was carried out by Flores and colleagues at Comer Children's Hospital, part of University of Chicago Medicine in Chicago, Illinois [9]. This was a quality improvement project — a structured attempt to fix a specific problem in how a hospital works, then measure whether the fix held. The team suspected their hospital was treating babies for congenital syphilis without those babies ever being reported to the city's health department, and set out to check, and to stop it.

The two teams worked independently, in different cities, using different methods, and neither study mentions the other — which is part of what makes them worth reading side by side. One examined the gap before birth; the other examined the gap after it. Together they describe a single problem with two separate leaks.

What the Portland team found about testing during pregnancy

The Oregon results point in two opposite directions. Syphilis tests came back positive far more often over time, rising from 1.8% of tests in 2018 to 4.3% in 2024, and the number of children diagnosed rose from 2 in 2018 to 16 in 2024 — every one infected by their mother during pregnancy [8]. Among newborns tested in the first four weeks of life, about 46% were positive, and the typical age at diagnosis was three days old. Over the same period, not a single newborn tested positive for HIV. Two infections moving in opposite directions inside one hospital, under two different testing policies.

The most striking finding concerns timing rather than coverage. Of the 13,889 pregnancies, 87% were tested for syphilis at some point — a good figure. But among the 202 pregnancies where syphilis was confirmed, only 64% were picked up in the first three months; 21% were first found in the middle third of pregnancy and 15% not until the final third. Roughly one in three infections was therefore discovered at a stage where finishing a full course of treatment well before birth may no longer be possible. Testing was happening. It was often happening too late.

What the Chicago team found about babies who were never counted

The Chicago findings are about record-keeping, and they are stark. Among babies born between 2011 and 2022, the team found 154 diagnosed with congenital syphilis and given penicillin [9]. Of those, 107 — about 69% — had a negative result on their own blood test, a test called a rapid plasma reagin, or RPR. When the team checked the Chicago Department of Public Health's records, not one of those 107 babies had ever been reported.

The reason was not carelessness. The hospital's reporting happened automatically whenever a baby's RPR came back positive, so babies treated for the other, more common reason — because their mother's treatment during pregnancy had been inadequate — never triggered the system. The team changed the trigger. They arranged for the children's infectious diseases team to be paged for every baby exposed to syphilis in pregnancy, added a standard checklist into the hospital's electronic record, tidied up how the diagnosis was coded, and sent monthly reminders to staff. Over the following 17 months, 44 babies were diagnosed and treated, 15 of them with a negative RPR — and all 44 were reported. When 30 staff members were surveyed six months in, every one said the change had improved care, and no one reported a downside.

Why a baby's negative blood test can be misleading

This is the piece of biology that ties both studies together. During pregnancy, antibodies — the proteins the immune system makes to fight infection — cross the placenta from the pregnant person into the baby, so a newborn's blood test partly reflects the mother's immune system rather than the baby's own. It works the other way too: a baby who really is infected may not yet have made enough of their own antibodies to show up on a test in the first days of life. A negative result in a newborn is therefore genuinely reassuring in some situations and genuinely misleading in others, which is exactly why guidance tells doctors to look at the mother's results and treatment first.

That is also why treatment is often given as a precaution — not because doctors are certain the baby is infected, but because the consequences of missing it are serious. Untreated congenital syphilis can affect the brain, the bones and other organs [10], and hearing loss is a recognised later effect still not well measured in modern studies [11]. Being treated is not a verdict about anyone's behaviour; it is a precaution taken because the medicine is safe, cheap and highly effective.

What this means for families

If you are pregnant, the practical message is about repetition rather than novelty. A syphilis test early in pregnancy is standard, and in 2025 the US Preventive Services Task Force — an independent expert panel that reviews prevention evidence — reconfirmed that everyone should be screened during pregnancy [12]. What the Oregon study adds is that one test at the start may not be enough if there is any continuing risk, so it is entirely reasonable to ask about testing again later in pregnancy or at delivery, and about whether a partner should be tested — partners are tested far less often, and untreated partners are a common route back to reinfection.

None of this is about individual failure. Researchers who have interviewed families in areas with high rates of congenital syphilis consistently find the same practical obstacles: transport, unstable housing, gaps in insurance, and clinics that are hard to reach or hard to get an appointment at [13]. Those are system problems, and a syphilis diagnosis in pregnancy carries a stigma it does not deserve. If your baby is being treated, good questions to ask are which category the team has placed your baby in, how long the penicillin course will be, what follow-up and hearing checks are planned, and whether the case has been reported — that last step protects other pregnancies rather than penalising yours.

Readers of In[Neo]Sight's coverage of newborn screening for congenital cytomegalovirus in New York and of pulse oximetry screening for critical congenital heart disease will recognise the theme: what a screening programme finds tells you as much about the programme's design as about how common a condition really is.

What researchers are working on next, and what to take from this

Both studies come from a single hospital, so neither can tell us exactly how big the national gap is. The Chicago group has already begun answering that by applying the same detective method — looking for babies who received penicillin, rather than relying on blood tests or billing codes — across a large group of children's hospitals [14]. Other useful work would be to measure how long before delivery the last negative test was taken, since that gap decides whether an infection can be caught in time, and to follow treated children for longer to understand their hearing and development.

The takeaway is not alarming so much as clarifying. Congenital syphilis is preventable, penicillin works, and the two things standing between families and that outcome are both fixable: testing that happens late, and cases that are treated but never counted. Chicago needed no new technology and no extra staff to close its half of the gap — only the recognition that a baby's negative blood test is not the end of the story.

References

  1. Bowen V, Su J, Torrone E, Kidd S, Weinstock H. Increase in incidence of congenital syphilis — United States, 2012–2014. MMWR Morbidity and Mortality Weekly Report. 2015;64(44):1241–1245. doi:10.15585/mmwr.mm6444a3 ↩
  2. Kimball A, Torrone E, Miele K, et al. Missed opportunities for prevention of congenital syphilis — United States, 2018. MMWR Morbidity and Mortality Weekly Report. 2020;69(22):661–665. doi:10.15585/mmwr.mm6922a1 ↩
  3. McDonald R, O'Callaghan K, Torrone E, et al. Vital signs: missed opportunities for preventing congenital syphilis — United States, 2022. MMWR Morbidity and Mortality Weekly Report. 2023;72(46):1269–1274. doi:10.15585/mmwr.mm7246e1 ↩
  4. Gomez GB, Kamb ML, Newman LM, Mark J, Broutet N, Hawkes SJ. Untreated maternal syphilis and adverse outcomes of pregnancy: a systematic review and meta-analysis. Bulletin of the World Health Organization. 2013;91(3):217–226. doi:10.2471/BLT.12.107623 ↩
  5. Ghanem KG, Ram S, Rice PA. The modern epidemic of syphilis. New England Journal of Medicine. 2020;382(9):845–854. doi:10.1056/NEJMra1901593 ↩
  6. Workowski KA, Bachmann LH, Chan PA, et al. Sexually transmitted infections treatment guidelines, 2021. MMWR Recommendations and Reports. 2021;70(4):1–187. doi:10.15585/mmwr.rr7004a1 ↩
  7. Adhikari EH. Syphilis in pregnancy. Obstetrics & Gynecology. 2020;135(5):1121–1135. doi:10.1097/AOG.0000000000003788 ↩
  8. Vaz L, Guzman-Cottrill JA, Downs S, et al. Rising congenital syphilis amid declining perinatal HIV: a 7-year study from a tertiary medical center in Oregon. Microbiology Spectrum. 2026;14(2):e02593-25. doi:10.1128/spectrum.02593-25 ↩
  9. Flores JM, Dewey A, Palama BK, Ahn YW, Bartlett AH. Increased reporting of infants born with congenital syphilis to a local department of public health: a quality improvement project. Open Forum Infectious Diseases. 2025;12(10):ofaf609. doi:10.1093/ofid/ofaf609 ↩
  10. Salomè S, Cambriglia MD, Montesano G, Capasso L, Raimondi F. Congenital syphilis: a re-emerging but preventable infection. Pathogens. 2024;13(6):481. doi:10.3390/pathogens13060481 ↩
  11. Hafeeez AA, Bezerra KC, Jimoh Z, et al. Scoping review of hearing loss attributed to congenital syphilis. PLOS ONE. 2024;19(4):e0302452. doi:10.1371/journal.pone.0302452 ↩
  12. US Preventive Services Task Force. Screening for syphilis infection during pregnancy: US Preventive Services Task Force reaffirmation recommendation statement. JAMA. 2025;333:2006–2012. doi:10.1001/jama.2025.5009 ↩
  13. Chan EYL, Smullin C, Clavijo S, et al. A qualitative assessment of structural barriers to prenatal care and congenital syphilis prevention in Kern County, California. PLOS ONE. 2021;16(4):e0249419. doi:10.1371/journal.pone.0249419 ↩
  14. Flores JM, Grills N, Kane JM, et al. Assessing diagnostic accuracy of congenital syphilis using penicillin administration data through the Pediatric Health Information System (PHIS) database. International Journal of STD & AIDS. 2025;36(6):475–486. doi:10.1177/09564624251324981 ↩