Maternal mortality in Central
Province, Kenya, 2009-2010
Introduction: Maternal mortality for Kenya was 488/100,000
live births in 2009. Maternal mortality estimate for Central Province is
unknown. We retrospectively reviewed data between 1st July 2009 and 30th June
2010 to estimate the hospital based maternal mortality ratio, characterize
deaths by time, place and person and describe possible causes of deaths in
Central province, Kenya.
Methods: We abstracted data using a standard form
from maternal death notification and review forms and the district reproductive
health reports. Data was entered and analyzed using Microsoft Excel.
Results: There were 89,512 live births and 111
deaths. The facility-based maternal mortality ratio was 124/100,000 live
births. Seventy-three (66%) deaths had been audited. Thirty seven (33%) were
aged 25 to 34 years. The mean age was 31years (±6). Thirty seven (33%) had a
parity of less or equal to 2. Most case deaths (19%, n=21) had attended 2 or
less antenatal visits. The main gestation was below 37 weeks with 48% (n=53).
The main mode of delivery was vaginal (26%, n=29). Majority (35%, n=32) case
deaths had delivered a live birth. Thirty seven (33%) mothers had been stable
on admission. The main reason for admission was labor
with 12% (n=13). Thirty-eight (34%) died within 24 hours after admission.
Majority (27%, n=30) were admitted antepartum but 39%
(n=43) were postpartum at the time of death. Thirty-five (32%) died of hemorrhage and 8(7%) Eclampsia.
Conclusion: Maternal mortality is of public health
importance in the region. Most deaths occurred within 24 hours after admission.
Third delay was important. Bleeding and Eclampsia
were the main causes of death. A third (34%) of notified deaths were not reviewed.
Maternal mortality is the
death of a woman while pregnant or within 42 days of termination of pregnancy,
irrespective of the duration and the site of the pregnancy, from any cause
related to or aggravated by the pregnancy or its management but not from accidental
or incidental causes. It is the leading cause of death among women of
The global maternal mortality ratio was
estimated at 210 per 100,000 live births in 2010 and 480 per 100,000 in Africa
whereas Kenya's mortality
ratio was estimated at 360 per 100,000(1).Africa therefore shared the highest
burden especially sub-Saharan Africa. Maternal mortality ratio for Kenya was
estimated at 488 per 100,000 live births according to Kenya Demographic Health
Survey of 2009 and these results indicate that maternal mortality remains high
in Kenya .The target for Kenya is to reduce it to 147
per 100,000 live births by 2015.
Globally, the most common causes of maternal
morbidity and mortality are hemorrhage, infection,
high blood pressure, unsafe abortion, and obstructed labor
(WHO, 2013). The major causes of maternal mortality in Africa are hemorrhage (33.9%), sepsis/infections (9.7%), hypertensive
disorders (9.1%), HIV/AIDS (6.2%), obstructed labor
(4.1%), abortion (3.9%)and anemia
(3.7%) . Although reporting of
the causes of maternal deaths is incomplete in the Health Management
Information System (HMIS), the leading causes in Kenya appear to be antenatal
and postpartum hemorrhage. Others include Eclampsia, sepsis, ruptured uterus, and obstructed labor.
The fifth United Nations Millennium Development
Goal (MDG) on maternal health aims to reduce maternal mortality by
three-quarters between 1990 and 2015.This
goal still remains a challenge in Africa due to the inability to reliably
measure levels and trends of maternal mortality using existing health
information systems. Most deaths can be averted even in resource limited
settings provided information necessary to monitor levels and trends and to
guide interventions is strengthened.
A maternal death surveillance and response
system can provide the essential information to measure and monitor maternal
mortality at sub-national level and stimulate and guide actions to prevent
future maternal deaths. The system has the potential to provide real-time,
frequent monitoring of maternal mortality levels, trends, causes and
circumstances surrounding the deaths provided attention is made to ensure
completeness of reporting and data accuracy. The system would also strengthen
civil registration and vital statistics system in the long term. However, this
process requires technical innovations and financial resources .
In many sub-Saharan African countries however, these processes are not
effective in monitoring maternal mortality and their contributing factors due
to underreporting, incomplete data and lack of analysis and utilization of data
generated by the system.
Maternal mortality ratio for Central province is
unknown and accurate data to monitor progress in service delivery performance
are scarce. There are no data on trends of maternal mortality in health
institutions because of poor reporting rates by hospitals .
No study had been done in Central Province to determine maternal mortality and
the circumstances surrounding maternal deaths. This justified the need to have
a baseline data that could provide a reference for monitoring service delivery
and future investigations on maternal mortality and its contributing factors in
the region. The information derived from the findings could also be useful to
health providers in adopting interventions and policies for the reduction of
maternal mortality and improvement of maternal health services.
This investigation therefore aimed at estimating
the hospital based maternal mortality ratio, characterizing the deaths by time,
place and person and describing the causes of maternal mortality in Central
province, Kenya. The purpose was to formulate recommendations to be shared with
stakeholders in reproductive health so as to improve maternal death
surveillance, reviews and notifications and maternal health care services in
general for the achievement of the fifth millennium development goal by 2015.
The review was carried
out in Central Province, Kenya, in December 2012. Based on the national
population census of 2009, the province had a total population of 4,383,743 with
an annual growth rate of 1.6%.The rate of skilled deliveries was 73.8%, family
planning coverage of 66.7% and approximately 92.7% clients attending at least
one antenatal visit. There were 24 hospitals, 81 health centres and 304
dispensaries. Approximately 89,512 live babies were born in health facilities
during the study period.
We carried out a descriptive retrospective
review of maternal deaths reported from health facilities between 1st
July 2009 and 30th June 2010. The study population comprised of
women who had given birth in health facilities in Central Province during the
study period. Data was abstracted using a standard form from the Maternal Death
Notification and Review forms, and the monthly district reproductive health
Maternal death was defined as any death of a
woman while pregnant or within 42 days of termination of pregnancy, from any
cause related to or aggravated by the pregnancy or its management but not from
accidental or incidental causes (ICD-10).
Maternal Death Review (MDR) or Audit was defined
according to the National Guidelines for Maternal and Perinatal Death
Notification and Review as the qualitative in-depth investigation of the
causes of, and the circumstances surrounding maternal deaths with the purpose
of tracing the path of the woman through the health care system and within the
facility, to identify any avoidable remedial factors to improve maternal care
in the future. The process is incomplete unless an attempt is made to respond
to the findings with appropriate action. An audited maternal death was
ascertained if a notified death was accompanied by a filled Maternal Death
We structured the abstraction form on a Ms.
Excel spreadsheet with columns containing variables similar to the Maternal
Death Notification and Review forms. The variables extracted included: age,
parity, ANC attendance, gestation, mode of delivery, status of the baby at
birth, mother's status on admission, reasons for
admission, timing of death since admission, stage of pregnancy on admission,
stage of pregnancy at death and the reported cause of death. Variables on the
Maternal Death Notification and Review forms that were severely incomplete were
excluded from analysis. These included marital status, educational level,
details of antenatal care, delivery care and newborn care, interventions, any
other contributing factors, comments on potential avoidable factors and action
We entered data on the spreadsheet for each
month starting from 1st July 2009 to 30th June 2010 to make a line list of the
cases. The dataset was examined for obvious errors by comparing the computer
entries with those in the Maternal Death Notification and Review forms. We made
corrections before onset of data analysis. We finally presented the information
derived from the analysis using tables, frequencies, proportions, measures of
central tendency and dispersion.
One hundred and eleven
(111) deaths were reported from 1st July 2009 to 30th
June 2010 among 89,512 live births. The reporting was aligned to the government
fiscal year that runs from 1st July to 30th June the
following year. Among the 111 deaths, only 73 (66%) had been audited. It is a
requirement that after every occurrence of death, a maternal death review committee
audits the death within 7 days and develops action points to mitigate the gaps
that may have resulted to the death.
shows the maternal and obstetric characteristics for the 111 maternal deaths.
The case deaths had a mean age of 31 years (±6) with age ranging from 18 to 51
years. Thirty seven (33%) of the maternal deaths occurred in
women aged between 25 to 34 years. Thirty seven (33%) of the mothers had
a parity of less or equal to 2. Most of the case deaths (19%, n=21) had attended
or less antenatal visits. The gestation age was commonly less than 37 weeks,
with a proportion of 48% (n=53).The average gestation was 31 weeks with a range
of 7-44 weeks. A high proportion (26%, n=29) of the deliveries were vaginal.
For majority (35%, n=32) of the case deaths, the baby was live at birth. In
thirty seven (33%) of the deaths, the mother had been stable at the time of
admission. The main reason for admission was labor
accounting for 13 (12%) of the women followed by both post-partum hemorrhage and pre-Eclampsia and Eclampsia at 9 (8%) respectively. Thirty-eight (34%) died
within 24 hours after admission. Majority of the mothers (27%, n=30) who died
were admitted during the antepartum stage of
pregnancy (which is the period before onset of labor),
and at the time of death, majority (39%, n=43) of them were in the postpartum
stage of pregnancy (which is the period after delivery).
Pregnancy complications contributing to the
maternal deaths in Central Province are shown in Table
2. Thirty-five (32%) of the deaths were due to hemorrhage,
8(7%) Eclampsia, cardiovascular disease 4(4%),
amniotic embolism 4(4%), HIV/AIDS and infections contributed 3 (3%) each. The
rest contributing 1 (1%) each were reported as having been caused by either
cardiac arrest, disseminated intravascular coagulation (DIC) or the cause was
unknown. It is important to note however, that there was a high proportion of
missing information for all the parameters that were evaluated.
We documented a
facility-based maternal mortality ratio of 124/100,000 which was lower than the
national estimates, and also lower than the national targets of 2015. The
mortality ratio was lower compared to the 190 per 100,000 live births reported
in the study that reviewed data collected for the period between January 1981
and September 1988 in Thika, a constituent district
of Central Province . Auditing of deaths had been fairly
done, with 34% of the maternal deaths having not been reviewed.
Majority of the maternal deaths reported had
occurred within 24 hours after admission. This was consistent with findings
from a retrospective study conducted in Rift Valley Provincial General Hospital
where 109 deaths that were recorded between 1994 and 1998 were evaluated. The
study had revealed that 53% of the deaths had occurred within 24 hours of admission. This is a clear indication
that most maternal deaths are of emergency nature.
The deaths were prevalent among age group 25 to
34 years. This is consistent with the findings from a study in Moi Teaching and Referral hospital where a retrospective
audit of 150 maternal deaths that had been recorded from 2004 to 2011 revealed
that majority of the deaths had occurred in women between 25 and 34 years . Similarly, a study at Kenyatta National Hospital among 253
maternal deaths found two-thirds of the women to have been between 25 and 35
years of age . The study by Juma
et al, 2000, revealed that the deaths were mainly from the 14 to 24 years age
group. The variation could be explained by differences in demographic and
Thirty seven of the case deaths were less or
equal to para 2. This was in agreement with findings
in the study in Kenyatta National Hospital . The
highest proportion of mothers had attended two antenatal visits or less with
slightly more having attended two visits and a fifth having not attended any
visit. Majority had been less than 37 weeks gestation. This was observed in the
study in Moi Teaching and Referral hospital .
Most case deaths were admitted during ante
partum period. This does not concur with the findings in Moi
Teaching and Referral Hospital where mothers were mainly admitted during the
ante intrapartum period . The
reason for admission was labor followed by postpartum
hemorrhage and pre-Eclampsia/Eclampsia and majority were admitted in stable condition.
However, the highest proportion died after delivery. The main mode of delivery
was spontaneous vertex delivery. For majority of the women, the baby was live
at birth. The main causes of death were hemorrhage
and Eclampsia, with hemorrhage
contributing more than half the deaths. Others were cardiovascular disease,
amniotic embolism, HIV/AIDS, infections, unsuccessful general anesthesia, cardiac arrest, disseminated intravascular
coagulation and unknown causes. This did not concur with the findings from the
study in Moi Teaching and referral hospital where Eclampsia was the leading pregnancy complication leading
death . The findings were however consistent with those
in the Kilifi study .
This review encountered a number of limitations
which included using facility-based rather than community-based data which
theoretically limits representativeness and generalizability.
Other limitations included incompleteness of data on most of the variables that
were reviewed. Despite these limitations, the findings of the investigation
represent an approximate situation of maternal mortality, its causes and
circumstances surrounding maternal deaths in Central Province, Kenya.
maternal mortality ratio was 124 per 100,000 live births. This was lower than
the national average and also below the national target of year 2015. Most of
the deaths reviewed were preventable. The most affected age group was between 25
and 34 years. The highest deaths occurred within 24 hours after admission.
There were higher proportion of deaths among mothers with a parity of 2 or
less, those who had attended 2 or less antenatal visits and those with less
than 37 weeks gestation. The fact that most mothers were admitted during ante
partum period, in normal labor and in stable
condition indicates that there could be factors within health facilities (third
delay) that make mothers particularly at risk in this region. This is contrary
to the notion that the first and second delays are most important. Hemorrhage and Eclampsia were the
main causes of maternal death.
The authors declare that
they have no competing interests.
Onesmus Maina Muchemi; Data management
and analysis, writing of the manuscript. Agnes Wangechi Gichogo; Data management
We wish to acknowledge
the health managers and staff of Central Province, Kenya, for their committed
efforts towards maternal death surveillance and response and Dr Juma Gachau Mwangi
for his expert guidance regarding case definitions and other critical
information on maternal death surveillance and response. Our gratitude is
similarly granted to Field Epidemiology and Training Program, Kenya for their
expert training and support throughout the manuscript preparation.
Table 1: Maternal and obstetric
characteristics of maternal deaths in Central Province, Kenya, from July 2009
to June 2010
Table 2: Pregnancy complications for
maternal deaths in Central Province, Kenya, from July 2009 to June 2010
Health statistics and information systems: Maternal mortality ratio (per
100 000 live births).
Accessed 15 August 2013.
National Bureau of Statistics (KNBS) and ICF Macro. Kenya Demographic and
Health Survey 2008-09.
http://statistics.knbs.or.ke/nada/index.php/catalog/23. Accessed 12 March
KS, Wojdyla D, Say L, Gulmezoqlu
AM, Van Look PF. WHO analysis of Causes of maternal death: A systematic
review. The Lancet. 2006; 367(9516):1066-74. Google
MDG 5: improve maternal health.
Accessed 12 March 201.
Danel. Maternal death surveillance and response.
Bulletin of the World Health Organization. 2011; 89(11):779-779A. PubMed
Health Sector Reforms Secretariat. Kenya Health Policy Framework 1994 -
2010: analysis of performance. Analytical review of health progress, and
systems performance. 2010.
Accesseed 12 March 2014. PubMed
JK. Socio-demographic and gynaecological variables of maternal mortality
in a Kenyan subdistrict: January 1981-September
1988. 1990; 67(2):118-125 Google
EA, Odiyo FN. Maternal mortality occurring at
the Rift Valley Provincial General Hospital, Nakuru.
East African Medical Journal. 2000; 77(7):382-5. PubMed
Yego. A retrospective analysis of maternal and
neonatal mortality at a teaching and referral hospital in Kenya.
Reproductive Health. 2013; 10:13. PubMed
JB, Obore S, Kigondu
CS. Millennium development goal 5: a review of maternal mortality at the
Kenyatta National Hospital, Nairobi. East Afr
Med J. 2006; 83(1):4-9. PubMed
E, Mochamah G, Ndila
C, Makazi A, Kosgei P,
et al. Maternal mortality in a rural community of Coastal Kenya: insights
from verbal autopsy (VA) in the Kilifi health
and demographic surveillance system (KHDSS). The Africa Journa of Health Science.
http://www.ajhsjournal.or.ke/?p=760. Accessed 12 March 2014.