Nairobi’s second governor, the colourful Mike Sonko, took office on 21 August.

In his inaugural speech outlining his priorities for the Kenyan capital, Sonko took exception to health standards in the city of an estimated 3.5-million people.

“It is completely unacceptable that, today, a mother or a newborn child has a higher chance of dying at birth in Nairobi than in Mandera,” he said, referring to a county in the north-eastern region that shares a border with Somalia and Ethiopia.

Could one of Kenya’s more remote regions have better survival rates than its main city, a regional hub and magnet for multinationals? We checked.

(Note: Africa Check contacted Sonko and his office for the source of the data informing this claim, but they are yet to respond. We will update this report if they do.)

What counts as a maternal or newborn death

Maternal death: Maternal mortality is the death of a woman during pregnancy, childbirth or 42 days after birth or the termination of a pregnancy. This does not include accidental causes.

Maternal mortality ratio: The risk of death associated with each pregnancy expressed in the number of maternal deaths per 100,000 of the female population of reproductive age (15-49).

Lifetime risk: The probability of a 15-year-old woman of eventually dying from a pregnancy or childbirth related cause in her lifetime.

Newborn mortality: Newborn, or neonatal, mortality is the death of a child in the first 28 days of life.  

Claim: “A mother has a higher chance of dying [during] birth in Nairobi than in Mandera”
Verdict: unproven

One way to look at maternal mortality is to estimate the risk of death associated with each pregnancy in women of reproductive age. Another is to calculate a woman’s risk of dying from such a cause in her lifetime.

As for the first kind of estimate, the United Nations Population Fund placed the national figure at 510 deaths per 100,000 live births in 2015, its most recent data. (Note: As Africa Check reported before, live births are used instead of the number of pregnancies in a country, which are difficult to determine.)

The State of the World Population 2016 said this figure was sourced from the United Nations Maternal Mortality Estimation Inter-agency Group. However, the data was not broken down by county.

The 2014 Kenya Demographic and Health Survey (DHS), released in January 2016, estimated a lower figure of 362 deaths per 100,000 live births between 2007 and 2014.

UN didn’t include Kenyan survey

he difference between the rates can be explained in part in that the Kenya survey was not available to the UN group when its estimate was modelled, Dr Doris Chou, a medical officer at the World Health Organization’s department of reproductive health and research, told Africa Check.

“When we update our estimates, the 2014 DHS information will be included, and its inclusion will likely make the differences observed between the internationally comparable [inter-agency group] estimates and the DHS smaller,” she added.

The Demographic and Health Survey also did not break down maternal mortality data by region, Dr Elizabeth Kimani-Murage, the head of the maternal and child wellbeing unit at the African Population and Health Research Center, told Africa Check.

99% of maternal deaths in 15 counties

The country’s official data agency, the Kenya National Bureau of Statistics, has data by county in its 2009 census report.

The agency found that in 2008, Mandera had the country’s highest maternal mortality rate by far, with 3,795 deaths per 100,000 live births. It was lowest in Nairobi, at 212 deaths for every 100,000 live births.

Just 15 of the country’s 47 counties accounted for nearly 99% of maternal deaths, the United Nations Population Fund highlighted in a report based on the census figures.

Mandera’s mortality rate sparked international headlines that it is the world’s most dangerous place to give birth in. However, a Bureau of Statistics spokesman told Africa Check that 2009 census data from north-eastern Kenya needs to be treated with caution. More deaths other than the targeted 12-49 age range for females may have been captured due to poor enumerator training.

The data agency did not have more recent data on maternal mortality at county level, the spokesman added.

As for a woman’s lifetime risk, the 2014 Demographic and Health Survey estimated that 2% of women of reproductive age, or about 1 in 67, will have a maternal death. 

In the absence of newer or reliable data on counties, we rate this claim unproven.     

Claim: “A newborn has a higher chance of dying at birth in Nairobi than in Mandera”
Verdict: mostly-correct

The 2014 Kenya Demographic and Health Survey does not break down newborn mortality by county, giving a national estimate of 22 deaths per 1,000 live births between 2010 and 2014.

However, governor Sonko’s claim could be close to the truth, Kimani-Murage from the African Population and Health Research Center told Africa Check.

In 2014, newborn mortality in Nairobi was 39 deaths in every 1,000 live births. In the north-eastern region, where Mandera is one of three counties, there were 24 deaths per 1,000 live births, the survey showed.

“The situation of poorer outcomes in Nairobi compared to other regions is attributable to the rapid growth of slums in Nairobi, with close to two-thirds of the urban residents being urban poor,” said Kimani-Murage, who has studied childhood mortality in Kenya’s urban and rural areas.

“Our previous work indicates high mortality rates among urban poor residents compared to other non-slum counterparts in Kenya,” she added.

Researched by Lee Mwiti, Africa Check, a non-partisan organisation which promotes accuracy in public debate and the media. Twitter @AfricaCheck and www.africacheck.org