OPINION: Too many mothers are dying in our best hospitals and we are failing to ask why

A woman is referred to a level 6 hospital in the early hours of the morning from a county hospital. 

She has already undergone emergency surgery after a ruptured uterus and severe bleeding. 

Blood in the county hospital was scarce, stabilisation limited, and her condition had already begun to deteriorate before the referral decision was made. 

She arrives at the referral hospital many hours later, critically ill. There is no prior alert to the receiving hospital, no structured handover, and no clear sense of what has already been done. 

By the time she reaches the referral facility, she is in advanced coagulopathy – a condition in which the blood does not clot to stop bleeding. Intensive care is unavailable. She dies shortly after arrival, still on triage.

This is not an unusual case. It is a pattern repeated across the country, and it reveals something uncomfortable about why maternal deaths still occur in places we assume are best equipped to prevent them.

Hospitals such as Kenyatta National Hospital (KNH) and Moi Teaching and Referral Hospital (MTRH) often sit at the centre of these statistics, reporting some of the highest maternal mortality ratios in the country. 

On the surface, this can be read as failure. But a closer look shows a different reality. These are not routine maternity cases. They are the most severe, referred in extremely serious situations from multiple counties after long delays and repeated attempts to manage complications at lower levels of care. 

Due to chronic delays and poor management, many of these women are transferred to higher-level hospitals simply to die.

The problem . Some of the poor and unacceptable outcomes in our referral hospitals point to failures that go beyond clinical competence. They often reflect who arrives, when they arrive, and in what condition. 

Facilities receiving a broader mix of cases appear to perform better. Those receiving concentrated late referrals of haemorrhage, eclampsia, sepsis, and obstructed labour inevitably report higher mortality.

Much of this is decided long before the patient reaches the referral gate. Evidence from teaching hospitals shows that referral itself significantly increases risk, often doubling the likelihood of death. 

The issue is rarely transport. It is delayed decisions, poor communication, and lack of stabilisation before transfer. 

Women are frequently moved without prior notification or coordination. By the time they arrive, shock is already established.

To understand this, we return to the Three Delays model: delay in deciding to seek care, delay in reaching care, and delay in receiving care. Kenya has made progress on the first two. 

More women are reaching facilities, and physical access has improved. But the third delay — what happens inside the system— has become the most dangerous and least visible.

Women are arriving at hospitals, but not always reaching effective care in time. They are referred late, received without preparation, and managed in overstretched systems. This is where the third delay becomes a system failure rather than a clinical one.

This is where the global framework, Every Woman Every Newborn Everywhere (EWENE), led by WHO, UNICEF, and UNFPA  becomes important. 

EWENE reframes maternal and newborn survival around a continuum of care. It shifts the question from whether women reach facilities to whether the entire system they move through can respond in time. It links community care, referral systems, and hospital response into one chain rather than separate levels.

Where EWENE-informed tools are integrated, deterioration is detected early and escalation is faster. Where they are weak or poorly linked to referral pathways, warning signs are missed or acted on too late. 

The gap between detection and response remains one of the most dangerous points in maternal care.

This is also where MPDSR (Maternal and Perinatal Death Surveillance and Response) should make the greatest impact. 

In principle, it is meant to reconstruct every maternal death, identify delays, and trigger corrective action. 

In practice, reviews often stop at the facility where death occurs, rather than following the full journey across community, referring facility, and referral hospital. Inter-county coordination exists, but it is not consistently reconstructing the complete chain of events.

The result is a system that counts deaths well but explains them poorly. In some cases, counties may appear to perform better simply by not registering many deaths, even when referring high-risk cases too late for effective care. 

Some hospitals are trying to close this gap. At Moi Teaching and Referral Hospital, teams begun tracing referred maternal deaths back to originating facilities, closing the loop through feedback sessions and continuous medical education. 

The findings are consistent: postpartum haemorrhage, eclampsia, and sepsis, compounded by delayed referral, poor communication, and weak preparedness on both sides of the transfer.

What becomes clear is that maternal mortality is no longer mainly about access. It is about coordination across the entire system of care. 

If we are serious about reducing these deaths, national referral hospitals must be fully integrated into inter-county MPDSR processes. 

Death reviews must follow the full patient journey, not just the final facility. And EWENE-linked early warning systems must move from policy to practice so that detection leads to real-time action, not retrospective explanation.

Women are not dying because they reached the wrong hospital. They are dying because they reached the right hospital too late.

The author, Dr. Richard Mogeni Mogaka, is Chairman of the Kenya Obstetrical and Gynaecological Society (KOGS) – Northrift.