When a pregnancy tips into periviability: a Kenyan obstetrician on the hardest weeks of his ward

On a Tuesday in mid-July 2026, the Daily Nation's Healthy Nation column published a long-form clinical reflection by a Nairobi obstetrician on what he calls the hardest ward in the hospital: the one where babies arrive before viability. The essay, drawn from years of case notes at a tertiary referral centre, lays out the protocols and the heartbreak of managing periviable birth, the narrow band from 20 to 25 completed weeks of gestation where survival is possible but never promised, and where every decision carries the weight of probability, parental consent and equipment that may or may not be available when it matters.

The column is, on its surface, a doctor's essay. Read alongside Kenya's quietly improving neonatal reporting, it is also a window onto the structural constraint that defines maternal and newborn care across much of the country: the gap between what is now clinically possible, and what a given district, on a given night, can deliver.

The 20-week line

Periviability, the obstetrician writes, is conventionally defined as the gestational window between 20 weeks and 25 weeks and 6 days. Below 20 weeks, survival is effectively zero and active resuscitation is not offered. Above 26 weeks, outcomes improve sharply. Inside the band, the literature in high-income settings now quotes survival rates that climb steeply with each extra week: roughly the difference, at 24 completed weeks, between one baby in three and one baby in two going home from the neonatal unit.

Kenyan referral hospitals operate inside the same physiology. The obstetrician's case mix, drawn from a unit that takes transfers from across the Nairobi metropolitan area and beyond, includes the full gradient: women arriving in preterm labour at 23 weeks, at 24, at 25; women with pre-eclampsia severe enough that the only safe option is delivery regardless of gestational age; women whose waters break at 21 weeks and who sit in hospital for days waiting, hoping, negotiating. The clinical decision tree at each of those inflection points is, he argues, the thing that residents underestimate when they rotate through the unit.

The decision turns on three numbers: gestational age, estimated fetal weight, and the presence or absence of antenatal corticosteroids. A mother who has received two doses of betamethasone before delivery can hand her infant a measurable advantage in lung maturity, sometimes enough to move a borderline case from a predicted non-survivor to a predicted survivor with morbidity. A mother who arrives precipitously, without time for steroids, has that advantage withheld. Geography determines part of the math: a mother who has had antenatal care in a tertiary centre has had steroids considered, prescribed, and timed. A mother transferred in from a level-three facility at 24 weeks plus three days may be arriving in active labour with no time to give them.

What the parents are told

The piece is unusually direct about consent. The obstetrician describes a standing protocol in which the parents of a periviable infant are given, in the hour before delivery, a structured description of three options: comfort care only, with the infant held and offered warmth; full resuscitation including intubation, surfactant and admission to the neonatal intensive care unit; and a third option, active support but with limits set in advance on chest compressions, prolonged ventilation and escalation beyond a defined window.

The framing matters. Comfort care is presented not as abandonment but as a treatment path with its own evidence base, including evidence that, for infants born at 22 weeks or below, the burden of resuscitation may produce a few additional hours of life and no more. Full resuscitation is presented as a real option but with the survival probability and the morbidity risk attached. The middle option is reserved for parents who want the team to try, but who have begun to think about what they will not allow the team to do.

The obstetrician is candid that parents almost never arrive pre-decided. They arrive frightened, often having never heard the phrase "periviable" before the morning of the delivery. The protocol's value, he writes, is that it forces the conversation into a structure that allows a frightened person to participate in a decision that the medical system would otherwise take from them.

The limits the system imposes

What is most striking, in a Kenyan context, is what the essay does not have to dramatise: the equipment and staffing bottlenecks. Continuous positive airway pressure machines, surfactant, parenteral nutrition, blood-gas machines that function at three in the morning: the list is familiar to anyone who has worked in a Level 3 neonatal unit outside the world's wealthiest health systems. The obstetrician does not name a particular hospital and the piece is written as a clinical meditation rather than a supply-chain audit, but the implications are visible in every paragraph.

The unit has, at any given time, a finite number of cots and a finite number of nurses trained to ventilate a 700-gram infant. An additional delivery at 24 weeks may not arrive at a moment when a ventilator is free, when an incubator is clean, when an anaesthetist is in the building rather than in theatre. The calculation that a parent in Stockholm or in Singapore would receive as a probability in a high-resource setting becomes, in Nairobi, a calculation that includes the question of whether the equipment that would translate that probability into a survival is here, tonight, available, working.

What the reporting numbers show

Kenya's neonatal mortality rate, the number of newborns dying before 28 days of life per 1,000 live births, has been on a long downward trend since the early 2000s but remains, by global comparison, elevated. Public reporting from the Ministry of Health, gathered through the District Health Information Software system and aggregated in successive Kenya Demographic and Health Surveys, captures deaths but does not, at present, separately report survival or morbidity outcomes for the periviable band, the segment of births between 500 and 800 grams that drives the bulk of neonatal mortality in tertiary centres.

The obstetrician's essay contributes indirectly to a closing of that reporting gap. It does not produce a register, and a Daily Nation column is not a surveillance instrument. But it does perform a piece of administrative work that formal datasets struggle with: it makes a category, the periviable infant, visible as a category in public discussion. That visibility is a precondition for policies that fund surfactant at scale, that staff Level 2 hospitals for ventilatory support, and that make antenatal corticosteroids a default rather than a luck-of-the-draw intervention in district antenatal clinics.

The honest paragraph

The piece is not built to settle the questions it surfaces. Survival at 24 weeks plus zero days is not the same number as survival at 24 weeks plus six days, and the published survival curves in the high-income literature, from which the African protocols draw their working assumptions, were generated on cohorts with substantially higher baseline resources. Whether the Kenyan survival probability at any given week of gestation equals the Canadian or the Finnish probability is not known with the kind of precision that would let a parent be told a single number with confidence.

What the obstetrician offers instead is a discipline: a structure for the conversation, a refusal to promise, and an acknowledgement that the worst outcomes on the ward are not the ones where a baby dies, but the ones where a baby survives the resuscitation and the intensive care and the ventilator weeks only to leave hospital with a morbidity that no one prepared the family for. The column is a brief for honesty, written by a clinician who has seen what dishonesty costs at three in the morning.

This piece draws on a clinical reflection published in the Daily Nation's Healthy Nation column. Where the reporting on periviable survival in Kenyan tertiary hospitals is implicit in the essay rather than explicit, this publication has said so rather than fill the gap with figures from elsewhere.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/DailyNation
• https://t.me/DailyNation