health · family: it worked in the lab
listening to theheartbeat, missing the breath
Fetal distress during labor is detected too late because no non-invasive monitor can measure fetal oxygenation
Problem statement
During labor, the critical concern is whether the fetus is receiving adequate oxygen. Fetal hypoxia — insufficient oxygen supply due to umbilical cord compression, placental dysfunction, or uterine hyperstimulation — can cause brain injury, cerebral palsy, or death within minutes. The standard monitoring tool is cardiotocography (CTG), which tracks fetal heart rate patterns and uterine contractions, but CTG has a false-positive rate exceeding 60% for predicting fetal acidemia. This means the majority of emergency cesarean sections triggered by concerning fetal heart rate patterns are performed on babies who are not actually in distress. No non-invasive technology can directly measure fetal blood oxygenation through the maternal abdomen during labor.
Why this matters
Approximately 140 million births occur worldwide each year (WHO). Intrapartum-related hypoxic events caused an estimated 510,000–717,000 neonatal deaths and roughly 1.15 million new cases of neonatal encephalopathy in 2010, 96% of them in low- and middle-income countries (Lee et al. 2013). In high-resource settings, the primary consequence is unnecessary intervention: the U.S. cesarean section rate was 32.3% in 2023 (vs. 10–15% recommended by WHO), driven largely by defensive obstetric practice in response to non-reassuring but non-specific CTG patterns. Each unnecessary cesarean increases maternal morbidity (surgical complications, longer recovery, future pregnancy risks) and cost: average commercial-insurer payments for maternal and newborn care ran $27,866 for cesarean births vs. $18,329 for vaginal births (Truven Health Analytics 2013).
What’s been tried and why it hasn’t worked
CTG has been the standard of care since the 1970s, but the Cochrane review found it has not reduced perinatal mortality or cerebral palsy compared to intermittent auscultation (it does roughly halve neonatal seizures) while significantly increasing cesarean deliveries (Alfirevic et al. 2017). Fetal scalp blood sampling can directly measure fetal blood pH/lactate but is invasive (requires cervical dilation and membrane rupture), intermittent (not continuous), and rarely performed outside of European centers. Fetal pulse oximetry was developed in the late 1990s–2000s, but the only FDA-approved system (Nellcor's OxiFirst — the N-400 monitor with FS-14 fetal sensor, conditionally approved in 2000) was pulled from the market by its manufacturer after a 5,341-woman NICHD trial found that displaying fetal oxygen saturation neither reduced cesarean delivery rates nor improved newborn condition (Bloom et al. 2006) — the sensor required internal placement against the fetus after membrane rupture, was unreliable during contractions, and knowledge of its readings did not change management outcomes. STAN (ST segment analysis of the fetal ECG) showed promise in European trials but failed to demonstrate benefit in a large U.S. trial of 11,108 women (Belfort et al. 2015); the trial's authors suggest that differences between U.S. and European practice, such as more frequent fetal blood sampling in Europe, may have contributed to the different results.
What would unlock progress
A non-invasive, continuous monitoring technology that can quantify fetal cerebral oxygenation through the maternal abdomen during labor — providing a direct measure of the variable that matters (brain oxygen supply) rather than an indirect proxy (heart rate patterns). Candidate technologies include transabdominal near-infrared spectroscopy (but light scattering through maternal tissue severely limits fetal signal); transabdominal photoacoustic imaging (combines optical absorption with ultrasound resolution); or advanced signal processing of fetal ECG waveform morphology that extracts oxygenation-correlated features. Any solution must work during uterine contractions (when monitoring is most critical), on mothers of all body habitus, and be interpretable by labor and delivery nurses and obstetricians with minimal training.
Entry points for student teams
A student team could develop and test a Monte Carlo photon transport model of near-infrared light propagation through layered maternal tissue (skin, fat, uterine wall, amniotic fluid) to quantify the theoretical depth limit and signal-to-noise ratio for transabdominal fetal pulse oximetry. An engineering team could prototype a multi-wavelength photoacoustic sensor array and test its ability to detect oxygenation changes in a tissue phantom at fetal depths. Relevant disciplines: biomedical engineering, biomedical optics, signal processing, obstetrics.
Genes
Structural cousins: stuck the same way, in other fields
Sources
ARPA-H, "Making Obstetrics Care Smart (MOCS)," ARPA-H press release, "New ARPA-H obstetrics program seeks to make delivery safer for moms and babies," 2025, Lee, A.C.C., et al. (2013), "Intrapartum-related neonatal encephalopathy incidence and impairment at regional and global levels for 2010 with trends from 1990," Pediatric Research 74(S1):50–72, Bloom, S.L., et al. (2006), "Fetal Pulse Oximetry and Cesarean Delivery," New England Journal of Medicine 355(21):2195–2202, Belfort, M.A., et al. (2015), "A Randomized Trial of Intrapartum Fetal ECG ST-Segment Analysis," New England Journal of Medicine 373(7):632–641, Alfirevic, Z., et al. (2017), "Continuous cardiotocography (CTG) as a form of electronic fetal monitoring (EFM) for fetal assessment during labour," Cochrane Database of Systematic Reviews, Issue 2, CD006066, CDC/NCHS, "Births: Final Data for 2023," National Vital Statistics Reports 74(1), Truven Health Analytics (2013), "The Cost of Having a Baby in the United States," Accessed 2026-08-21 (ARPA-H pages first accessed 2026-02-23) go to source 1 ↗ go to source 2 ↗ go to source 3 ↗ go to source 4 ↗ go to source 5 ↗ go to source 6 ↗ go to source 7 ↗ go to source 8 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
Related briefs: `health-pulse-oximeter-skin-tone-bias` (pulse oximetry accuracy challenges — the same optical measurement principles face even greater barriers in transabdominal fetal monitoring); `health-postpartum-sepsis-monitoring-gap` (obstetric monitoring gap — addresses the maternal side of the same clinical setting). The `failure:wrong-problem` tag is primary: CTG has been refined for 50+ years, but the problem was never the fidelity of heart rate recording — it was that heart rate is not the right variable to measure. Clinicians need oxygenation data, not better heart rate data. The `temporal:static` tag reflects that this problem has persisted since the 1970s with minimal progress — the core technical barrier (measuring fetal oxygenation non-invasively through maternal tissue) has not fundamentally changed.
- Note reconciled 2026-08-20: the note above calls `failure:wrong-problem` primary; in the August taxonomy pass that tag was replaced on this brief's genome (heart rate as the measured variable is a proxy for the oxygenation clinicians actually need). The original note is kept verbatim as the tagging rationale of record.
Reconciliation 2026-08-21: The brief was sole-sourced to ARPA-H's MOCS program page while carrying roughly ten quantitative claims that are not on that page (the page itself contains no statistics beyond "50-year-old tool" — confirmed by fetch); every number now has its own cited source. Corrections: the hypoxia burden figure ("600,000 neonatal deaths and 1 million cases of neonatal encephalopathy annually") was a blur of the actual estimates — replaced with Lee et al. 2013 (Pediatric Research 74(S1):50–72): 510,000–717,000 intrapartum-related neonatal deaths and ~1.15 million new NE cases for 2010, 96% in low- and middle-income countries. The U.S. cesarean rate updated 32% → 32.3% (CDC/NCHS Births: Final Data for 2023, NVSR 74(1)). The unattributable "$28,000 per case" cost replaced with Truven Health Analytics 2013 commercial-payer figures ($27,866 cesarean vs. $18,329 vaginal, total maternal+newborn payments). The fetal pulse oximetry device was misnamed ("Nellcor FS-14" is the sensor only) and its exit mischaracterized as a regulatory withdrawal: the system was the Nellcor OxiFirst (N-400 monitor + FS-14 sensor, FDA conditional approval May 2000), and the manufacturer stopped selling it after the NICHD masked/open trial (Bloom et al., NEJM 2006;355:2195–2202) found no reduction in cesarean rates and no improvement in newborn condition — a commercial exit, not an FDA withdrawal. The CTG-vs-auscultation sentence sharpened per Alfirevic et al., Cochrane 2017 (CD006066): no reduction in perinatal mortality or cerebral palsy, seizures roughly halved, cesareans significantly increased (RR 1.66). Claims that verified clean and stand unchanged: CTG false-positive rate >60% for fetal acidemia (AAFP Intrapartum Fetal Monitoring, 2020 — "falsely positive for fetal acidosis approximately two-thirds of the time," sensitivity 57%/specificity 69%); ~140 million births/year (WHO maternal-health topic page); WHO 10–15% cesarean reference range; the STAN U.S. trial (verified as Belfort et al., NEJM 2015;373:632–641, n=11,108, no superior outcomes — now cited); both ARPA-H URLs and the press-release title verified live and exact. URLs checked: arpa-h.gov/explore-funding/programs/mocs, arpa-h.gov/news-and-events/new-arpa-h-obstetrics-program-seeks-make-delivery-safer-moms-and-babies, nature.com/articles/pr2013206, nejm.org/doi/full/10.1056/NEJMoa061170 (via PubMed 17124017), nejm.org/doi/full/10.1056/NEJMoa1500600 (via PubMed 26267623), cochranelibrary.com CD006066.pub3, cdc.gov/nchs NVSR 74-1, nationalpartnership.org Truven executive summary.
Reconciliation 2026-09-29: Site review round 5 read the whole write-up against its sources. All fetches were made on 2026-09-29. Read in full: both ARPA-H pages, Lee et al. 2013 and the Truven executive summary, each at the address in the Source line; Belfort et al. 2015 from the PubMed Central copy (https://pmc.ncbi.nlm.nih.gov/articles/PMC4631435/); Alfirevic et al. 2017 from the PubMed Central copy (https://pmc.ncbi.nlm.nih.gov/articles/PMC6464257/); the CDC report from the Internet Archive copy of the cited PDF, because cdc.gov refused the request; and the AAFP article named in the note above (https://www.aafp.org/pubs/afp/issues/2020/0801/p158.html). Bloom et al. 2006 was read as an abstract only, from its Europe PMC record (PubMed 17124017), because nejm.org refused the request and there is no open copy. (1) The Source line dated the ARPA-H press release 2024. The page says "Published November 17, 2025", so the year now reads 2025. (2) The text said STAN "showed promise in Scandinavian trials." Belfort et al. say the approval rested on "European studies", and the two trials they cite are the "Plymouth randomized trial" and "a Swedish randomised controlled trial". The text now says European trials. (3) The text said the U.S. trial may have failed "because the technology required expertise in interpretation that generalist obstetricians lacked." No source read here says this. Belfort et al. report that every care provider was trained and certified on the system, and that results in the first half of the trial did not differ from results in the second half. The explanation they offer is that U.S. and European practice differ, for example fetal blood sampling is more common in Europe. The sentence now gives that explanation. Confirmed and left as written: the authors, titles, journals, volumes, pages and years of the six papers and reports in the Source line; 510,000 to 717,000 neonatal deaths, 1.15 million new cases of neonatal encephalopathy and 96% in low- and middle-income countries (Lee et al.); the 32.3% U.S. cesarean rate for 2023 (CDC); $27,866 and $18,329 as average commercial payments for cesarean and vaginal births, from 2010 claims (Truven); no reduction in perinatal death or cerebral palsy, neonatal seizures halved, and more cesareans with continuous CTG (Alfirevic et al.); 5,341 women, no fewer cesareans and no better newborn condition (Bloom et al., abstract); 11,108 women (Belfort et al.); heart-rate monitoring largely unchanged since the 1970s (ARPA-H press release); false positives for fetal acidosis about two-thirds of the time (AAFP). The note above gives the risk ratio for cesareans as 1.66; the review as fetched gives 1.63. That number is not in the body text. NOT checked: the statements that the pulse oximetry sensor went in after the membranes ruptured and was unreliable during contractions, because only the abstract of Bloom et al. could be read; the device name, the 2000 approval and the withdrawal from sale, and the WHO figures of 140 million births and 10 to 15%, which rest on the earlier note and were not fetched again; and the statement that the U.S. cesarean rate is "driven largely by defensive obstetric practice." ARPA-H links the monitor to unnecessary cesareans and to lawsuits, but no source read here says how much of the rate this explains. These are left as written. Genes untouched.