ACCESS TO INFANT AND
MATERNAL HEALTH (AIM
HEALTH) PROJECT
REPORT ON MID TERM REVIEW
NICOLE LEE
Social & Behavioral Interventions Program
MSPH Essay
March 2015
First Reviewer: Larissa Jennings, Ph.D
Second Reviewer: Ingrid Friberg, Ph.D
i
Executive Summary
Background
World Vision Kenya (WVK) launched the Access to Infant and Maternal (AIM) Health project in 2012 at
Mutonguni Area Development Programme (ADP), which is located in the semi-arid land of western Kitui
County. This area has an estimated infant mortality rate of 39 deaths per 1,000 live births and maternal mortality
ratio of 137 deaths per 100,000 live births.
In support of Ministry of Health (MOH) initiatives, Kenya is one of five countries in which WV is implementing
the AIM Health Programme. Recognizing the unique vulnerability of women and young children, the AIM Health
project seeks to address their needs with a set of comprehensive interventions. The AIM Health Programme is a
contextualized roll-out of World Vision’s 7-11 strategy, which focuses on promoting seven key behavior change
interventions for pregnant women and eleven key interventions for infants and children below two years. The
programme activities are based on evidence-based interventions from the Lancet’s Maternal Survival and
Neonatal Survival series, as well as World Health Organization guidelines for improving maternal and newborn
health.
Methodology
AIM Health project carried out a mid-term review (MTR) between May and August 2014. The MTR used a
mixed methods approach, collecting primary quantitative data through Lot Quality Assurance Survey (LQAS) and
qualitative data. The review also included measurement of population-level outcomes in the results based
framework (RBF) of the project and modelling the impact of the project’s interventions in place of measuring
population-level mortality rates. Qualitative data was collected through in depth interviews (IDIs) and focus
group discussions (FGDs) with a range of program stakeholders. Fifteen IDIs and nine stratified FGDs were
coordinated and conducted by a trained team consisting of a graduate-level Global Health Fellow, local
facilitators, and WVK staff.
The Lives Saved (LiST) tool, a multi-cause software model of mortality, was used to make a retrospective
estimation of lives saved/deaths averted, and to project the same estimates for the remaining part of the project
and beyond. Coverage data were derived from default population-level survey sources and project
baseline/LQAS. Three projections were created for the main analyses, and a separate projection was created for
secondary analyses – which were carried out through extraction of results from LiST.
Key Findings
Program progress to date: The perceptions of health status changes in the AIM Health project area
were generally positive, but were not supported by improved quantitative data. Nutritional outcomes,
specifically, and care seeking behaviors have improved. Community members reported that some
negative traditional practices have fallen out of favor over the past few years, especially when it comes to
unskilled deliveries with Traditional Birth Attendants (TBAs) or harmful newborn care. Several new
facilities have opened in the past 2-3 years, and utilization of health services are said to have increased as
a result of the health messages and facility upgrades.
Lessons, Promising Practices and Key Success Factors: As a regional staff member put it,
“Community participation ensures ownership which later on makes the project be more sustainable.”
ii
The synergy created through AIM Health is evident through the noted community ownership and
local/national cohesion to achieve shared visions and objectives around MCHN. One manager admired
the project’s five-year duration as a best practice for lasting change.
Constraints and Challenges to Implementation: Environmental and socio-economic constraints, such
as poverty and a challenging landscape, hamper maximum intervention coverage. The AIM Health
catchment area is rural, the terrain is difficult, and there are very few reliable transportation options.
AIM Health relies on CHW-led health promotion, and many stakeholders believe minor illness
management to be a critical component left out of the intervention. District health services, in general,
are also constrained by occasionally delayed funding, supply and staff shortages, and poor infrastructure.
Management Issues and Lessons Learned: AIM Health has had limited staff, but MOH/WVK
integration and shared resources are key to successful implementation of the project. There is agreement
that strong points of the program are budget clarity and focused spending. There is an interest in
utilizing technology for monitoring and evaluation purposes at the district level, as long as it is not
heavily web-based. One important lesson learned is that activities need to be carefully planned but
flexible enough to respond to delays or environmental changes.
Impact: The LiST exercise found that the AIM Health project contributed to saving a total of 36
maternal and child lives (including stillbirths) since the project’s start, which is approximately an 8%
reduction in under five mortality and a 14% reduction in maternal mortality. There are an estimated 37
additional maternal and child lives that can be saved by achieving scaled up intervention coverage to RBF
target levels in 2015. The most important intervention to preventing maternal, neonatal and child
deaths is increased skilled birth attendance, particularly in a facility.
Recommendations
Big gains in mortality reduction can be made with increased care seeking (and treatment) for
pneumonia, increased oral rehydration solution (ORS) treatment for diarrhea, and continued increases
in skilled delivery. Project activities should focus on scaling up ICCM training in the next year to aid in
achieving end of program goals.
Accelerate advocacy efforts at the county government level and work to strengthen communication
among different levels of the community health system with the purpose of creating sustainable funding
solutions for the intervention components.
Develop and implement a communications strategy to better promote the community ambulance and
encourage utilization. Aid the DHMT in further work on transportation guidelines for health system
linkages in the sub-county.
Encourage the MOH to develop an alternative supervision model to ease the current gap. For example,
CHWs could be incentivized with task-shifting by creating an advancement opportunity for high-
performing CHWs to move into supervisory roles (with stipend) as they gain more skills.
iii
Acknowledgments
The author was based in Kitui and Nairobi (Kenya) as a Global Health Fellow with World Vision Kenya during
the time of this study. Thanks are owed to staff at the World Vision Kenya National Office and to members of the
World Vision Ireland AIM Health team, who sponsored and organized training in the Spectrum software in
Baltimore, who provided logistical support and fellowship accommodation, and assisted in the procurement of
background information and coverage data. Yvonne Tam, MHS and Ingrid Friberg, Ph.D. at Johns Hopkins
Bloomberg School of Public Health provided technical assistance on the Spectrum software and modelling
process. Study design and supervisory support was provided by Beulah Jayakumar, MD, MPH, public health
consultant. Additional feedback on earlier drafts and supervisory support was provided by Erin Jones, MPH,
World Vision International (Sustainable Health) and Larissa Jennings, Ph.D., Johns Hopkins Bloomberg School of
Public Health.
List of Acronyms
ADP – Area Development
Programme
AIDS – Acquired
Immunodeficiency Syndrome
AIM – Access to Infant and
Maternal (Health)
ANC – Antenatal Care
ARI – Acute Respiratory Infection
ARV – Anti-retroviral
CBO – Community Based
Organization
CHC – Community Health
Committee
CHEW – Community Health
Extension Worker
CHW – Community Health
Worker
CU – Community Unit
CVA – Citizen Voice and Action
DHMT – District Health
Management Team
DHS – Demographic and Health
Survey
FGD – Focus Group Discussion
HH - Household
ICCM – Integrated Community
Case Management
IDI – In-depth Interview
IPTP – Intermittent Preventive
Treatment in Pregnancy
ITN – Insecticide Treated Net
KEPH – Kenya Essential Package
for Health
KEPI – Kenya Expanded
Programme on Immunization
LiST – Lives Saved Tool
LQAS – Lot Quality Assurance
Survey
MCHN – Maternal & Child Health
and Nutrition
MDG – Millennium Development
Goals
MMR – Maternal Mortality
Ratio/Rate
MOH – Ministry of Health
MTR – Mid-term Review
NGO – Non-governmental
Organization
NMR – Neonatal Mortality Rate
OBA – Output Based Approach
ORS/ORT – Oral Rehydration
Solution/Treatment
PMTCT – Prevention of Mother-
To-Child Transmission
PNC – Postnatal Care
RBF – Results Based Framework
SBA – Skilled Birth Attendance
SD – Standard Deviation
SGA – Small for Gestational Age
TBA – Traditional Birth Attendant
TTC – Timed and Targeted
Counseling
U5MR – Under-5 Mortality Rate
UNICEF – United Nations
International Children's
Emergency Fund
USAID – United States Agency for
International Development
WASH – Water and Sanitation
Hygiene
WHO – World Health
Organization
WVI – World Vision Ireland
WVK – World Vision Kenya
iv
Contents
1.0 Background ......................................................................................................................... 1
2.0 Qualitative Study .................................................................................................................. 6
2.1 Scope, Design and Methods .................................................................................................. 6
2.2 Data Collection, Processing and Analysis ................................................................................. 6
2.3 Findings .......................................................................................................................... 7
2.4 Limitations ..................................................................................................................... 13
3.0 Impact Study ....................................................................................................................... 15
3.1 Scope, Design and Methods ................................................................................................. 15
3.2 Findings for Objective 1 (2014 Midterm) ............................................................................... 16
3.3 Findings for Objective 2 (2015 Endline) ................................................................................. 17
3.4 Mortality Findings ............................................................................................................ 18
3.5 Limitations ..................................................................................................................... 18
4.0 Conclusions & Recommendations............................................................................................. 20
5.0 Reflections on Field Practicum ................................................................................................ 22
References ........................................................................................................................... 23
Programmatic Context............................................................................................................ 25
Data Collection Summary ........................................................................................................ 27
LiST Data Collection .............................................................................................................. 29
LiST Results ......................................................................................................................... 33
v
Tables and Figures
Table 1: TTC components of AIM Health......................................................................................... 2
Table 2: Qualitative Data Collection .............................................................................................. 27
Table 3: Survey Results .............................................................................................................. 27
Table 4: Demographic Data ......................................................................................................... 29
Table 5: National/Region Coverage Levels Update ............................................................................ 29
Table 6: AIM Health Indicators and Values ...................................................................................... 30
Table 7: LiST Intervention Coverage Summary ................................................................................. 32
Table 8: Deaths Averted, Neonates and Children (Midterm) ................................................................ 33
Table 9: Maternal Deaths and Stillbirths Averted (Midterm) ................................................................ 34
Table 10: Changes in Nutritional Status (Midterm) ............................................................................ 35
Table 11: Cases of Illness Averted (Midterm) ................................................................................... 36
Table 12: Deaths Averted, Neonates and Children (Endline) ................................................................ 36
Table 13: Maternal Deaths and Stillbirths Averted (Endline) ................................................................ 38
Table 14: Cases of Illness Averted (Endline) ..................................................................................... 38
Table 15: Changes in Mortality ..................................................................................................... 38
Table 16: Lives Saved, Neonates ................................................................................................... 39
Table 17: Lives Saved, Children under 5 ......................................................................................... 40
Table 18: Lives Saved, Maternal .................................................................................................... 40
Table 19: Lives Saved, Stillbirths ................................................................................................... 41
Table 20: Lives Saved Summary .................................................................................................... 42
Figure 1: Map of Kitui County ...................................................................................................... 25
Figure 2: AIM Health MTR Timeline ............................................................................................. 25
Figure 3: LiST Modelling Examples ............................................................................................... 26
Figure 4: Deaths Prevented by Cause (Midterm) ............................................................................... 34
Figure 5: Deaths Prevented by Cause (Endline) ................................................................................. 37
Figure 6: Reduction in Mortality by Intervention (Midterm) ................................................................ 39
Figure 7: Changes in Mortality, Children under 5 .............................................................................. 41
Figure 8: Changes in Mortality, Neonates ........................................................................................ 42
1
1.0 Background Health overview in Kenya
Kenya has made considerable progress in maternal and child health since the Millennium Development Goal
(MDG) targets for 2015 were first conceived. Mortality has improved, fertility has declined, and health service
utilization has increased across the country over the years1. Although improvements are being seen, Kenya is not
on track to reach those targets for 2015. Maternal and child mortality rates are still higher than expected. In
2008, Kenya’s under-5 mortality rate was 74 deaths per 1,000 live births, the infant mortality rate was 52 deaths
per 1,000 live births, and the maternal mortality ratio was 488 deaths per 100,000 live births2. Although most
women received antenatal care at some point in pregnancy (92%) and almost all infants were breastfed at birth
(97%), only three-quarters of children under five were fully vaccinated (77%) and a third of children (35%) were
stunted2. In contrast, Kitui district’s under-5 mortality rate was 86 deaths per 1,000 live births, the infant
mortality rate was 63 deaths per 1,000 live births, 64% of children were fully vaccinated, and 38% of children
were stunted in 20083. Seven out of 10 facilities in Kenya offer basic child health services and antenatal care, but
normal, assisted, vaginal delivery is only possible in three out of 10 facilities4. Between a quarter to half of
facilities regularly have basic equipment and supplies for these services, a functioning transportation system,
regular water, and/or regular electricity4. Effective and efficient programs are needed to continue to address
these issues and elevate the overall health status of children in this country.
Policy Environment
Vision 2030 is the country’s 20-year development blueprint with economic, social, and political pillars to propel
the nation to middle-income status5. The vision to a globally competitive and prosperous nation is anchored on
macroeconomic stability, continuity in governance reforms, and enhanced equity and wealth creation
opportunities for the poor, among other strategies. In alignment with this vision, the overall Kenya health sector
policy goal is to “attain the highest possible standard of health in a manner responsive to the needs of the
population”6. To attain universal coverage of critical services, the government is implementing programs that
focus on
o eliminating communicable conditions;
o halting and reversing the rising burden of non-communicable conditions;
o reducing the burden of violence and injuries;
o providing essential health care;
o minimizing exposure to health risk factors; and
o strengthening collaboration with other sectors.
In 2006, Kenya introduced a community health strategy to deliver basic health care services – the Kenya Essential
Package for Health (KEPH) – to 16 million Kenyans7. The purpose of the community health strategy was to
enable communities to improve and maintain a level of health that will enable them to participate fully in national
development towards the realization of Kenya’s Vision 2030. Community Units (CU) were established, each
made up of about approximately 5,000 people and 1,000 households, with 5 persons per household on average.
The CUs are served by volunteer health workers to deliver household health promotion, and are linked to district
health facilities.
The adoption of a new constitution by Kenya in 2010 introduced a rights based approach to health in service
provision and a devolved system of government, which assigned the larger portion of delivery of health services
to the counties8. With the exception of national referral services, in which the Ministry of Health Kenya (MOH)
governs tertiary hospitals that provide complex care, the governance of community and provincial facilities is left
to the counties. The government is committed to good health for all Kenyans, acknowledging that health is not
only a right but also a responsibility for all. Promotion of good health at different levels of society is the
responsibility of all individuals, families, households, and communities.
2
World Vision’s response
There are many local, national, and international health and development actors working to realize these goals in
Kenya. To support the government, World Vision develops and implements programs in some of the neediest
areas. With funding support from World Vision Ireland (WVI), World Vision Kenya (WVK) launched the Access
to Infant and Maternal (AIM) Health Programmea in 2012 at Mutonguni Area Development Programme (ADP)
to tackle these health challenges in an area with poor health outcomes. Recognizing the unique vulnerability of
women and young children, the AIM Health project seeks to address their needs with a set of comprehensive
interventions.
Kenya is one of five countries where the AIM Health Programme is currently being implemented. The first phase
of the AIM Health project—a preparatory phase for WVK staff, local communities and MOH partners termed
AIM-Prep—engaged stakeholders in a participatory approach to program implementation from 2011 to 2012.
Officially launched in 2012, the objective of AIM Health was to be actively engaged with local stakeholders in
order to increase community access to maternal and infant health care services and in turn, improve maternal and
child health. Its overall goal is to reduce both infant mortality and maternal mortality by 25% over five years.
Mutonguni ADP is situated in western Kitui County – formerly known as Kitui West District and now referred
to as Kitui West sub-county (Figure 1) – and serving a population of approximately 50,000 people. Kitui
County is located in a semi-arid region with moderately high temperature and bi-modal rainfall. The central part
of the county is characterized by hilly ridges separated by wide low lying areas and has slightly lower elevation.
The population is rapidly growing and poverty is quite high, as many families rely on sustenance farming and
livestock herding. Agriculture, though difficult, is the major backbone to the economy, and cotton is a major
industry.
AIM Health Programme
The AIM Health Programme is a contextualized roll-out of World Vision’s 7-11 strategy9, which focuses on
promoting seven key behavior change interventions for pregnant women and eleven key interventions for infants
and children below two years (Table 1). This approach is delivered at the household for behavior change
communication and community levels with Community Health Committee (CHC), Citizen Voice and Action
(CVA), and Community Health Worker (CHW) led Timed and Targeted Counselling (TTC) programming.
Table 1: TTC components of AIM Health
Pregnant women Children (0-24 months)
Adequate diet Appropriate breastfeeding
Iron/folate supplements Essential newborn care
Tetanus toxoid immunization Hand washing with soap
Malaria prevention and treatment (including IPT) Appropriate complementary feeding
Birth preparedness and healthy timing/spacing of delivery Adequate iron
De-worming Vitamin A supplementation
Access to maternal health services (antenatal care, postnatal
care, delivery by skilled birth attendant, PMTCT,
HIV/AIDS, tuberculosis, screening for STIs)
Oral rehydration therapy (ORT/zinc)
Prevention and care seeking for malaria
Full immunization for age
a For this paper, “AIM Health programme” will always be in reference to the 5-country initiative in general, and “AIM Health project” will be in reference to the Mutonguni initiative.
3
Prevention/treatment of acute respiratory infection
De-worming (+12 months)
Timed and Targeted counselling (TTC) refers to household level counselling informed by the MOH’s community
health volunteers training curriculum, which included basic health promotion and maternal and newborn care.
Community health workers (CHWs) are community-elected volunteers and are frontline agents of the Ministry.
Each CU comprises approximately 50 CHWs supervised by a Community Health Extension Worker (CHEW).
The CHWs make a series of eleven scheduled visits to households when women become pregnant, and
throughout the child’s infancy, communicating the 7-11 message sets at the most appropriate times using
effective dialogue-based counselling techniques.
The Community Health Committee (CHC) represents the collective of stakeholders in the community focused
on Maternal and Child Health and Nutrition (MCHN) outcomes. The CHC is an integrated civil structure
empowered to coordinate community activities and manage activities leading to child well-being. In Kenya, these
groups were formed independently through the MOH under the Community Health Strategy (2006), province
by province. CHCs have an important governance role in the processes that take place to improve health at the
community level, particularly regarding the CU. They have the responsibility of mobilizing communities for
involvement in health-promotive and disease prevention activities, and they are active in supporting latrine
construction in communities as well as other small projects.
Citizen Voice and Action (CVA) is an approach that aims to improve accountability from the government (both
national and local) in order to improve the delivery, quality and efficiency of public services through increased
dialogue. AIM Health helps to mobilize and train groups of community members for advocacy work. Educated,
empowered and mobilized citizens are encouraged to assess the performance of public services that are provided
in their communities. In addition to equipping CVA with the tools they need to effect change, AIM Health also
contributes to health system strengthening through activities that are aligned with MOH priorities such as
community outreaches, capacity building, and gifts-in-kind.
Evidence supporting the programme design
When it comes to ensuring infant and child survival, key interventions have been repeatedly identified as effective
at preventing particular causes of mortality. Immunization, exclusive breastfeeding, antibiotics for pneumonia,
insecticide treated nets, and ORS, among others, have been evaluated to show efficacy in a variety of settings10–12.
While acknowledging that child mortality has a complex web of determinants, Jones and colleagues (2003)
estimated that 63% of deaths could be prevented if the most effective interventions – breastfeeding or ORS for
example – were universally available13. Unfortunately, some programs package these interventions haphazardly
out of convenience or funding requirements, rather than based on anticipated synergistic effects12. Several
researchers have noted that sustained and strong community engagement, care linkages, and health education are
crucial to meeting health needs at scale11,12.
Evidence on community-based strategies to improving maternal, neonatal, and child health outcomes is rapidly
mounting. Freeman and colleagues (2012) reviewed multiple child health interventions and identified four
programmatic approaches that the most successful interventions employ, using community empowerment as an
overarching strategy: home visitation, group educational or support meetings, health care service outreach, and
community-based treatment (case management) or referral14. Researchers have also seen that community
interventions effective at reducing mortality have elements of involving communities in designing solutions,
raising awareness through health behavior change strategies such as the interpersonal contact of caregivers and
decision makers, and encouraging dialogue through community mobilization15,16. In Kenya, recent studies have
evaluated the effectiveness of their community maternal and child health programs. In one study, a three year
project that trained volunteer CHWs in rural areas to deliver health education found that knowledge of newborn
care and the proportion of women going for skilled delivery was higher among those exposed to the messages
4
than among those who were not17. In a controlled two year trial of the implementation of Kenya’s national
Community Health Strategy, another study found that antenatal care, deliveries in facilities, latrine use, measles
vaccination, and water treatment practices significantly improved in intervention sites as compared to control
sites18.
Overview of the current MTR
The AIM Health project, which is being implemented until December 2015, carried out a mid-term review
(MTR) between May and August 2014 (Figure 2). The MTR used a mixed methods approach, collecting
primary quantitative data through a Lot Quality Assurance Survey (LQAS) and qualitative data. The purpose of
the MTR was to assess the progress made in the first two full years of implementation, program quality according
to the standards for the core implementation models and make specific recommendations for how the
implementation of AIM Health can be improved, allowing for context-specific programming feedback regarding
opportunities to increase efficiency and reach of program interventions. In addition to gauging progress and
quality, a qualitative study was conducted to help identify any limitations, opportunities, and risks connected
with the AIM Health program. The review also included measurement of population-level outcomes in the
results based framework (RBF) of the project and modelling the impact of the project’s interventions in the place
of measuring population-level mortality rates, as the latter is beyond the scope of the project. The Lives Saved
Tool (LiST) was used to make a retrospective estimation of lives saved/deaths averted, and to project the same
estimates for the remaining part of the project and beyond.
Process evaluation, or implementation research, is a common and practical tool in the field of program
evaluation19. Often there is a big gap between what is described in the protocol and what is actually implemented
on the ground. If a program fails, the reason may be unclear from standard pre- and post-measurement designs.
Process evaluations measure how well a program is operating in its specific context. They can help to modify or
strengthen the intervention, make midcourse corrections, help explain why results were achieved (or not), and
provide lessons for others. Process evaluations may contain a quantitative component, but they almost always
contain a qualitative component.
Qualitative research is a set of theoretical perspectives and methods for eliciting textual data. It is a long-standing
methodology of social science with well-known data collection strategies such as interviews with key informants,
focus group discussions, and observations20. A common strategy for qualitative data analysis include the
Grounded Theory technique, and common software for analysis include Nvivo or Atlas.ti. Qualitative research is
useful in process evaluations for understanding nuances of implementation and allowing the evaluation to be
culturally responsive. LQAS is a sampling and analysis quantitative methodology for rapid population-based
surveys, adapted from industrial quality control techniques. It is an inexpensive tool for monitoring coverage and
outcomes that requires a small sample size to determine if an area is “performing” at acceptable levels or not21. A
review found that more than 800 health-related studies using LQAS had been conducted globally since the mid-
1980s22. LiST, on the other hand, is a fairly novel tool for evaluation that has been picking up steam in the past
decade.
LiST has been described as a linear, mathematical modelling software to estimate the impact of scaling up
evidence-based intervention on various population level maternal and child health outcomes23. It is one of nine
modules nested in the Spectrum software, and it requires several pieces of input data for the models –
intervention coverages, underlying health status measures, and effect estimates of interventions on cause-specific
mortality, which general come from the Child Health Epidemiology Reference Group (CHERG)24,25. The
reduction in mortality – better known as lives saved or deaths averted – caused by a particular intervention is
calculated from the increased coverage of that intervention multiplied by the effectiveness of that intervention in
reducing mortality26 (see example, Figure 3). LiST modelling is a useful alternative to lengthy, randomized
controlled trials in that it can capture rare events (e.g. maternal mortality), overcome time lags (e.g. impact of
nutrition interventions), and can be used to construct counterfactuals to prospectively or retrospectively
interpret trends.
5
Often, LiST is used for strategically planning and validating public health programs. As an example, Friberg and
colleagues (2010) estimated using LiST that if high impact intervention were scaled up to 90% by 2015 across 42
African countries, nearly four million maternal and child deaths could be averted27. Additionally, the Catalytic
Initiative used LiST to model national scale up of five high impact intervention in five countries, and found that
under-5 mortality could be reduced by at least 20% in five years28. As LiST is adopted more as a tool in the field
of international health, more evidence of its use for evaluating programs and strengthening health systems
accumulates29. In several cases, LiST has been shown to measure mortality outcomes as good as or better than
standard epidemiological measurement. Ricca and colleagues looked at numerous USAID supported projects that
implemented community-based intervention packages for children under five years, and, using LiST in
comparison to measured DHS data, found that average coverage changes exceeded DHS trends in the majority of
cases – an overall 5.8% improvement in mortality compared to 2.5% by DHS16. A World Relief project in
Mozambique compared LiST estimations of mortality to independently collected evaluation data and found
reasonably accurate estimates of decline in under-5 mortality (39% in LiST vs. 37% independently)30. The
Accelerated Child Survival Project in Ghana found similar results, with LiST estimating a 10.7% reduction in
under-5 mortality which was within the 95% confidence interval for the measured reduction of 20.6%31.
Although LiST is not perfect – the same projects also found underestimations in some aspects – it is a practical
and sound tool for program planning and evaluation activities.
6
2.0 Qualitative Study
2.1 Scope, Design and Methods The overall purpose of the qualitative study was to provide in-depth, descriptive information on the AIM Health
project to complement available quantitative data. There were five specific objectives, which revealed key themes
from varied respondent groups. The study aimed to:
1. Describe the project’s progress to date.
2. Compile lessons, promising practices and key success factors.
3. Analyze constraints and challenges to implementation of the various project models.
4. Identify management issues that help or constrain program implementation and management lessons
learned.
5. Help clarify the probable mechanisms behind observed quantitative changes.
This qualitative study covered all stakeholders from communities and health facilities who participated in the
program, MOH partners, AIM project staff, and ADP/region/national office staff who have provided technical
or managerial support to the project. Fifteen IDIs were planned and conducted in English among community and
MOH partners, and WVK staff. Nine stratified FGDs were planned and conducted in the local language,
Kikamba, among beneficiaries, community groups, and CHWs.
IDIs and FGDs among key stakeholders were conducted by a trained team consisting of the author and local
facilitators, and these were coordinated by WVK staff. ADP staff hired local personnel to carry out FGDs in the
community. The facilitators both held Kenyan university degrees. The Fellow is a current U.S. graduate student
in public health trained in qualitative research methodology and analysis. A half-day training was conducted with
the facilitators on FGD methodology, facilitation skills, and the content of the guides.
The MTR consultant developed semi-structured interview and discussion guides. The tools were finalized
incorporating input from the evaluation team, including AIM Health staff and key technical staff from WVI and
the Global Centre. Interview and discussion guide topics focused on themes identified above, with neutral and
open lines of questioning. Probes were used to explore other relevant themes that emerged during the
discussions and interviews. Study participants were purposively sampled from among the 10 community units.
The FGD participants were identified, recruited, and introduced to the team by MOH staff with the support of
AIM Health staff. The IDI participants were identified by the MTR consultant and were asked to join the study by
AIM Health staff. All logistics of the data collection were planned, overseen, and made possible by WVK staff.
No ethical approval from an institutional review board was sought or obtained for this study. Evaluative studies of
public health programs are not typically considered human subjects research, and are thus not subject to peer
review.
2.2 Data Collection, Processing and Analysis Over the course of four weeks, nine FGDs and fifteen IDIs were conducted (Table 2). Verbal informed consent
was obtained from all study participants prior to interviews and discussions. A digital voice recorder was used to
record all IDIs and FGDs with consent from all participants. Face-to-face (mostly) in-depth interviews were
carried out at MOH/health facilities, and with staff at ADP, project, region and national levels in and around the
ADP area. The FGDs were carried out by a team of three consisting of gender appropriate local facilitators and
note-takers, and the Fellow. Each FGD called for eight to ten members in each discussion. FGDs were held with
CHWs, mothers of children under two and at least one older child, husbands or male partners, CHC and CVA
working group members across six different community units.
The main ideas and observations were reviewed between the Fellow and facilitators following each FGD. Notes
of the FGDs were written up into reports by the FGD facilitators. All FGDs and IDIs were transcribed and
7
translated to English where possible. Notes and transcripts from the IDIs and FGDs were collated and checked
for quality, in preparation for analysis. Additionally, secondary data in the form of project documents and
manuals, quantitative data from LQAS and prior reports were reviewed.
Data management for coding and analysis was aided by the use of computer software programs – Dedoose (web-
based)b and Microsoft Excel 2013. Analytic memos were used to organize and document the processes. A
combined approach was used to code the qualitative data, enabling codes to be developed both deductively from
pre-defined themes and inductively from the views of the participants. The process started with examining the
complete data set and developing a preliminary codebook based on the study objectives. After coding a few
transcripts, the codebook was revised with additional focused codes and categories that emerged. The transcripts
were checked against the revised codebook for meaning and overlaps, and then the codes were applied to the
remaining transcripts. The process of refining and applying the codebook was repeated until no new codes were
generated. The central themes were identified from clustering of codes.
All of the coded data were retrieved and sorted by codes and informant using Excel. The data were then charted
in a matrix, summarizing each theme by category of informant and noting exemplary quotes. When organizing
themes, the analyst frequently reviewed findings with other AIM Health Fellows for additional validation.
Thematic analysis was guided by the FrameWork approach32. Analysis began with reviewing the matrix and
making connections within and between informants and themes, noting contradictions and outliers.
Interpretations of possible relationships were noted in analytic memos, and possible explanations for what was
happening were developed to guide findings.
2.3 Findings Program Progress to Date
Changes in health status/outcomes
In the last few years, all stakeholders have noted changes in the health status of pregnant women and infants
under two years, mentioning each of the targeted 7-11 interventions/behaviors at least once. Markedly,
informants report that although peaking seasonally, cases of diarrhea and pneumonia are becoming harder to find.
Although it is perceived that latrine coverage, sanitation behaviors and ORS treatment have increased, survey
results show that both diarrhea and pneumonia prevalence among children under five have increased to
approximately 20% (from 15% and 12% respectively) and that ORS treatment, access to safe water, and hand
washing have all decreased since the project started (Table 3).
Members of the community also reveal that there have been reductions in diseases and that many children have
been immunized. According to the LQAS results, coverage of full immunization had dropped 30% (although no
significance testing or analysis of variance was done). Facility staff say the community is experiencing reduced
maternal, neonatal, and child deaths that could be related to increased recognition of danger signs, health care
seeking and skilled deliveries. There is a common view that women are increasingly accepting family planning to
space their children. LQAS results support this by showing improved completion of antenatal care, skilled birth
attendance, and postnatal care.
Informants relay that severe cases of malnutrition have been decreasing, and that many mothers are practicing
exclusive breastfeeding, with six months being the ideal duration. Survey data show that stunting and
underweight in children have been reduced (from 47% to 38% and from 20% to 12%, respectively), in addition
to improvements in early initiation of and exclusive breastfeeding (up to 71% and 91%, respectively). Mothers
and fathers report being more informed about complementary infant feeding and balanced diets for children and
pregnant women, which are supported by improved estimates of diverse diets.
b Dedoose Version 5.0.11, web application for managing, analyzing, and presenting qualitative and mixed method research data (2014). Los Angeles, CA: SocioCultural Research Consultants, LLC (www.dedoose.com).
8
Most stakeholders view AIM Health as a success for its contribution in delivering behavior change communication
at the household level and supporting health service improvements along with the MOH. The program works to
increase knowledge and change attitudes around specific behaviors, which fuels the uptake and adoption of these
behaviors leading to changes in health status overall.
Perceptions
AIM Health has had positive reception in the sub-county. The communities are supporting the CHWs in their
work and hold them in high regard. The perception of the health services varies from community to community,
but people are accepting the need to seek services from facilities instead of relying solely on each other or advice
from traditional healers. Community members report that some negative traditional practices have fallen out of
favor over the past few years, especially when it comes to unskilled deliveries with Traditional Birth Attendants
(TBAs) or harmful newborn care. According to informants, practices that are no longer acceptable include
putting soot on the cord stumps, rubbing ash mixtures on gums, feeding glucose and water to newborns, or
allowing women to eat soil for iron. A father in Musengo put it this way when describing changes seen in the
community, “Where there is no sickness, there is happiness.”
Demand/utilization of services
An apparent perception of change in utilization of services pertains specifically to childbirth. Many pregnant
women are gradually seeking antenatal care (ANC) earlier and more often, but facility staff note that fewer
women actually complete the fourth ANC visit. It is reported that more women are choosing to deliver in health
facilities due to counselling, birth planning, and the help of output based approach (OBA) vouchers for safe
motherhood – a government sponsored maternal health service financing intervention in place since 200533–35.
Health care seeking behavior has increased among the target population, particularly in regards to completing
scheduled immunizations and growth monitoring for children. “Defaulters” – described as people who either
refuse or delay immunizations for children – are traced and community members use social pressure and
mobilization to ensure that children are immunized. Some facilities even have queues due to increased demand.
Informants reveal that families have reduced their use of herbs, borrowing medicines from neighbors and buying
medicines from shops.
Health service accessibility and quality
Several new government and NGO health facilities have opened in the last few years in the AIM Health project
area to increase access to services for the target population. Currently, the MOH conducts door-to-door
immunization campaigns and, with facilitation from AIM Health, conducts monthly outreach services to the more
remote villages that are hard to reach. Community members report that the quality of services has been
enhanced with a noticeable improvement of attitudes by health workers. WVK also supported in provision of
KEPI (Kenya Expanded Programme for Immunization) equipment and acquisition of delivery packs to the
facilities. An example is given by a District Officer:
“Like you go in a place I have to reach, like Usiani. They’ll tell you at least there’s something good going on
because they have stayed there for a long period of time without any health facility. That facility was opened I
think one and half years ago. At least something is being done, so they say thank you because they’re seeing the
services there. They are able to access some of the services which are being provided by the Ministry. The
outreaches which have been supported by World Vision, hard to reach areas, they are really appreciating that.”
Status of women and children in household
Both husbands/male partners and mothers agree that AIM Health has contributed to playing a part in elevating
the status of women in the household. They say that, because of the counselling, men are more supportive of
wives throughout pregnancy by making efforts to accompany wives on ANC visits and for delivery. Women are
not made to work as much while pregnant or immediately after childbirth. One example was given that women
were no longer seen as slaves or beaten for secretly practicing family planning.
9
It was reported that other behavior influencers, like mother-in-laws, are accepting better practices – particularly
the avoidance of early complementary infant feeding – and husbands are learning to become champions for
nutrition and exclusive breastfeeding. One facility staff describes the change,
“If we counsel the mothers alone we encounter some problems because if you tell the mother she’s supposed to
take an extra meal, who is the provider? It’s the husband. So, if the husband is not there, he will not understand
why this woman is eating so much. But when the husband is there, he’ll be like, ‘You didn’t take your snack. You
didn’t take your porridge.’ He’ll be reminding.”
Intervention coverage and maintaining implementation standards
As far as maintaining minimum standards in project model implementation, AIM Health covers 10 CUs but at
least one sub-location does not have the minimum number of CHWs as specified by the project plan (50). CHWs
reported being responsible for 20 households of pregnant women and mothers of children under two in his/her
village, though the proposal called for a maximum of 15 per CHW. The CHWs also report visiting four or five
households each day.
The CHWs are diligent about completing all of the scheduled TTC visits, but the rainy seasons (in April and
October) can create additional adversities to overcome in order to visit each household. During the visits, the
CHWs report that they share targeted health messages to households using the counselling cards, check clinical
cards, refer women and children for services, and follow-up on referrals and births. Supervision of CHWs should
occur once per month, but constraints (discussed below) inhibit regular supervision.
Extent of support by Community Health Committees (CHC)/Citizen Voice and Action (CVA)
The CHCs report that they help to monitor progress of the CHWs, and play a role in problem solving and
conflict resolution for the CHWs. The CHCs have sensitized the community to activities related to AIM Health
and have led community dialogues and forums. CHCs are described as key for providing an appropriate and
supportive social environment for the work of CHWs and CHEWs. Some CHWs admit that the CHC has not
been that helpful for them. There seems to be tension in some areas regarding the method by which CHC
members were selected. Still, the CHC does try to support, encourage, and motivate the CHWs. They have
catered trainings for them and have also provided incentives in the form of goats, seeds, and flour.
CVA is noted as operating as a link for different groups in the community. They mostly involve themselves in
development projects and county budgeting. They have formed an umbrella CBO (community based
organization) for the CUs, and help get them registered as self-help groups. They are described by CHWs as the
force that advocates on the communities’ behalf and create awareness for their rights to health and health care.
Lessons, Promising Practices and Key Success Factors
What works: “Synergy”
The AIM Health program has been lauded at all levels for its strong design, especially for imbedding a
participatory preparation phase. The interventions were well planned, and funding was appropriately allocated.
One manager admired the project’s five-year duration as a best practice for lasting change. The programme has
allowed for novel solutions to emerge out of implementation. Prominent examples include arranging table-
banking among the CUs, acquiring a new community ambulance, and facilitating the deployment of Integrated
Community Case Management (ICCM) training to further the skills of CHWs.
Informants mention that a key practice of this program is that of community empowerment. AIM Health is
recognized as strategically working with MOH and other stakeholders to engage communities in taking
ownership of both its problems and the solutions for them. As a regional staff member put it, “Community
participation ensures ownership which later on makes the project be more sustainable.” The community-based
monitoring and data collection is also acknowledged as a best practice for programs such as AIM Health. The
communities report being aware of what is available to them in terms of health care, and are encouraging others
10
to take advantage of them. Other stakeholders suggest that the engagement on health also radiates to other facets
of well-being tangentially.
Another key practice is the indirect health system strengthening. WVK has demonstrated strong partnership with
the MOH. AIM Health is described as “harmonious”, allowing for all levels of partners to pool resources and
work together for the well-being of the area. The seamless integration of AIM Health into the existing MOH
strategy and structures provides a foundation for sustainability. WVK and the Ministry collaborated on
developing the TTC training manual, and as a result, the Ministry has taken up the TTC concept in other areas.
At the local level, capacity building has been vital to ensuring the continuum of care promoted by the program.
AIM Health, by facilitating training of CHWs and CHEWs on MOH standards and guidelines as well as the
occasional procurement of supplies, strengthens the quality of health facilities in the project area.
National learning
National level participants declare that AIM is contributing to learning throughout WVK and the MOH. For
example, the data being collected locally is used at the county level for advocacy and planning purposes.
Inclusive trainings and documentation allow lessons to be shared and applied across WVK ADPs AIM Health is
heralded for aligning with Kenya Health Policy priorities and complementing MOH strategy on health,
particularly relating to objectives 4 (Provide essential health care) and 6 (Strengthen collaboration with other
sectors). The project is also expected to contribute to long-term conversations pertaining to Kenya’s Vision 2030
Social Pillar as well as global 2015 MDG 4 and 5.
Gaps remaining
There were very few critiques towards the AIM Health program. AIM Health relies on CHW-led health
promotion, and many stakeholders believe minor illness management to be a critical component left out of the
intervention. The current CHW supervision structure is inadequate for the scope of the intervention. Whether
being a design flaw or constraint on the MOH side, the CHWs are in need of sufficient monitoring and feedback
of their work. Moreover, the community believes that AIM Health is not doing its due diligence for peripheral
target groups, such as disabled children, men, and adolescent mothers. Although a communications strategy was
implemented, engagement for these populations is perceived as a continued need.
Potential for sustainability
There are spillover effects of the AIM Health project intervention. This is evidenced by benefits of AIM Health
being seen even by those in areas the project is not covering, and those communities are demanding a similar
intervention. As one MOH representative explains, community-based care is already a “government obligation”,
in which AIM is a temporary gap-fill, and there is agreement that it should be extended. In WVK, scale-up is
desired and, already, ADPs have been identified for replicating the AIM Health strategy.
The community groups and supervision of CHWs by MOH will continue, but additional measures need to be
taken to ensure changes will last. Proper funding is a major challenge to sustaining the progress made thus far.
An MOH representative suggests that,
“You have to engage the county governments and bring them on board so that they can continue…so that they can
also safeguard some money, resources for continuing to expand coverage even in the other areas. We should talk
to the county government to continue strongly the areas where you are and then also think about moving to other
areas.”
Constraints and Challenges to Implementation
Coverage issues
Stakeholders identified several environmental constraints that have affected intervention coverage. Foremost is
the issue of distance. The AIM Health catchment area is rural, the terrain is difficult, and there are very few
reliable transportation options. The households that CHWs are required to visit can be very spread out, and they
11
sometimes cover distances up to and over five kilometers. Pregnant women struggle to deliver with skilled birth
attendants at health facilities and families struggle to follow through on referrals to the better-equipped hospitals.
Although the county government has purchased three community ambulances this year (with plans of more to
come), the issue of poverty has also constrained coverage as families have competing demands for their limited
income. Some families cannot afford the cost of transportation, or perceive facility costs to be too burdensome to
follow through on referrals – especially among the persons living with disabilities. At a more basic level, some
households cannot afford the necessities promoted through TTC for behavior change such as additional food for a
balanced diet or soap for hand washing. CHWs and project staff must often deal with people expecting material
support when visiting homes. Sometimes negative attitudes and beliefs can get in the way of implementation. For
example, the Kavanokie are a known religious sect that refuses modern medicine. CHWs also say they deal with
uncooperative people and women who hide their pregnancy from them.
Additional gaps in coverage are due to small failures in implementation. The CVA working groups are not
particularly recognized by the communities or the government – who declare that they were never introduced –
and they say they lack advocacy tools. CHWs have not been utilized fully in all areas, specifically in the Katutu
sub-location where at least three CHWs are no longer working. Fathers and CHC members in that area remark
that CHWs are not visible in the villages and they prefer to stay near the dispensary and market.
CHW motivation
For many CHWs, volunteering can be a burden as they strain to cover large areas, on top of tending to personal
responsibilities and problems. They, and other stakeholders, have expressed basic needs to do their work that
have not been fulfilled. Most prominent is a CHW kit with simple medicines, first aid, and preventive
commodities. There is also a strong demand for training to handle minor ailments, ideally, or a TTC refresher, at
minimum.
Every stakeholder discussed incentivizing CHWs as a way to motivate them, chiefly in the form of monetary
recompense. Currently, there is debate over a 2,000 Kenya shillings (KES) per month stipend recently
recommended by the government. While AIM project staff and CHC/CVA groups are making the CHWs aware
of this recommendation and advocating for its initiation by government officials, the MOH say that the
recommendation was essentially to align disparate NGO schemes. They note that even if they were to take this
on, they could not afford it because there are too many CHWs. A district official highlighted that,
“In fact, there’s a proposal that they be paid like 2,000 in a month from the government. Because now it’s kind of
disturbing their brain, ‘oh we are supposed to be given money.’ …They knew even when they were doing the
recruitment that this is a volunteer job and when you’re doing the recruitment you have to pass that message very
clear that you’re coming here. What you want? You want to serve the community and you want to assist them to
the various issues of health. But when that thing came like they expected the national government wanted to pay
them. So they are waiting. When are they going to start being paid? It’s distorting their minds, they start thinking
in terms of payment. And when they were being recruited, the issue of payment was not mentioned and they
accepted volunteer.”
A suggestion was made that AIM Health work with the MOH towards a more formal scheme of recognition or
appreciation for the CHWs before the project ends.
Monitoring and supervision issues
Some CHWs raise concerns that either 1) they have no referral forms or 2) the current referral forms are
inadequate. On top of distance issues, this makes following up on cases hard at times. The referral forms are an
important component of monitoring the project utilization, along with CHW registers. CHWs bring their
registers to CHEWs at the end of the month so they can be compiled into a report. A district official commented
that the self-selection of CHWs skews toward the less educated, which may affect the quality of information
being collected. A CHEW acknowledged possible lapses in data quality saying, “You are not sure whether this
information is true because you’ve not gone there.”
12
Everyone directly involved with project implementation agree that each CHEW has too many CHWs to account
for and supervise. CHEWs typically have a full clinical load and other MOH responsibilities, leaving almost no
time for adequate supervision of each and every CHW. One CHW mentioned going almost a year without being
visited, and others declared that CHC supervision was scant as well. In addition to similar distance and
transportation issues, a lack of appropriate tools was reported to hinder supervision visits.
Health service issues
Despite the decentralization of governance, health service delivery has been slow partly due to challenges in
securing adequate funding from the Line Ministries. Most of the health facilities in the project area are second tier
level of care dispensaries for primary care services and often do not have the capacity to support increased
demand for services. As a district officer notes,
“We talk of delivery in the facility, but the facilities they don’t have even a room where a mother can deliver.
There are some facilities that don’t even have a delivery bed. We find that it’s just a couch that they just went to
buy at the market for examining patients. So in case a mother comes possibly in second stage, she’ll just be lying
there.”
District health workers and beneficiaries raised concerns of limited services and poor infrastructure, such as a lack
of electricity, running water, and fences for security. Facilities are in need of equipment to be up to standards,
principally for maternity and lab services. Sometimes funding is delayed, causing shortages in supplies and drug
stock outs.
Health care is also hampered by restricted working hours, leaving many people without options late at night, or
when the only staff are gone for training, meetings, or are sick themselves. Health facilities are understaffed, and
those there are overworked or may not be up to date on medical skills. Monthly outreaches, or the provision of
health care services in the communities, can be expensive in terms of time by taking personnel away from the
facility.
Management Issues and Lessons Learned
Workload, monitoring and supervision
AIM Health has a single project manager and project officer. The staff are spread thin with the many moving
pieces of AIM and other projects being managed. The program requires considerable coordination with a
multitude of stakeholders, which can be taxing. Fortunately, they are significantly supported by other staff at the
ADP, region, and national levels.
While MOH officials provide monitoring and supervision support in the sub-county, there are issues in securing
enough trained personnel to conduct the visits. Until recently, supervision was conducted haphazardly due to the
lack of a standard checklist. WVK has assisted the Ministry in providing the needed monitoring tools in the sub-
county.
Data collection, reporting and use
According to a district officer, there has been a substantial improvement in data quality due to training CHWs
and CUs in data management. Reporting is streamlined, which allows the data to be shared between all
stakeholders, but regional staff note that the flow of information is problematic at times. There is an interest in
utilizing technology for monitoring and evaluation purposes, as long as it is not heavily web-based. There is need
for a dedicated and local M&E officer for AIM Health, as WVK staff have expressed dissatisfaction with the
evaluation design thus far.
Logistics, procurement and administration
Although it can take up to one month or more, the WVK procurement process is clear, and the online system has
made approvals easier. A procurement committee that includes community members ensures fairness in the
process. Respondents described the process of procuring supplies as starting with creating a purchase requisition
13
form online to be approved by management at the appropriate levels, “floating” price quotations to different
vendors, then awarding contracts by committee. One lesson learned concerning procurement is that the process
must be initiated early to account for possible delays like unreliable vendors. Although there is shared
administration and supportive integration at the ADP between projects, logistics must be carefully planned due to
limited availability of vehicles.
Line-management
AIM Health has stayed close to budget thus far, but there is a discrepancy between financial managers on
approximately how much the project has overspent. This could be due to managing spending across the different
fiscal periods that WVK and the donor have. There is agreement that strong points of the program are budget
clarity and focused spending.
When it comes to line-management, notable challenges are the inherent field realities of a community-based
program and balancing conflicting expectations. The time frame for planned activities may change rapidly with
shifting priorities of partners. Multiple stakeholders acknowledge that limited resources sometimes hinder AIM
from addressing more pressing needs. For example, the community may expect more out of the project than the
design allows for, like tangible facility improvements or health worker skills training.
Funding and reporting
Monthly financial reporting is centralized and feedback is constantly given to mitigate snags and support progress.
At times, the flow of funds is slow, but communication keeps implementation moving along. No difficulties are
found in managing the grant requirements or complying with donor expectations. The AIM team appreciates the
technical support given by WVI.
Bottlenecks and recommendations
A few bottlenecks were identified related to management and implementation. Staff turnover at WVK and staff
shortages within the Ministry impede successful completion of activities, procurement, monitoring, and
supervision. For a five-year grant, there is agreement that the program planning and budgeting has been too rigid,
overall, to meet on-the-ground needs as they arise.
“But you know when you come to the ground, things are different. Things are changing and you have to change
with the times. Yes, you’ve done TTC in the community and the household, good. But then now when they go to
the health facility they need to get equally good care, but they cannot get good care if capacity isn’t built. Five
years ago, or seven years ago, you plan for this, but in these seven years, things change. So that is where we have a
little bit of difficulty.” – AIM Project Manager
2.4 Limitations There were small challenges in executing the original protocol fully. First, the FGD training was shorter than
planned due to the late arrival of one facilitator, but the bulk of the time was spent ensuring adequate translation
of the discussion guides into the local language. Although two weeks of data collection were planned, there were
difficulties in scheduling interview or discussion time for dozens of participants. Community members were
sometimes hard to contact and mobilize due to land preparation and other home duties. One FGD had to be
rescheduled due to miscommunication concerning the start time. Both WVK and MOH staff were busy with
other work commitments and some informants were not located in the program area, therefore delaying
interviews further. Two interviews had to be completed over the phone, which resulted in muffled recordings
that were difficult to transcribe. An additional barrier encountered was fatigue with research by both FGD and
IDI participants associated with AIM Health in Mutonguni. WVK conducted a documentation exercise in
February, covering similar topics and utilizing many of the same participants. Many participants were noticeably
silent in FGDs, despite prodding from facilitators. Direct translation from the local language to English, or
verbatim transcription, of the FGDs were not possible due to limited time and only having one person to work
on them. In spite of these challenges, staff and participants alike did their best to make the data collection process
as smooth as possible. Finally, regarding analysis, internet connectivity obstructed coding progress initially, but
14
all data was downloaded for offline use afterwards. The analyst attempted to reconcile partial transcriptions with
corroboration from other sources, such as reading project documents, speaking with project managers or
searching online for additional information.
15
3.0 Impact Study
3.1 Scope, Design and Methods The overall purpose of the LiST exercise was to strengthen the evidence base of the AIM Health Programme.
Specific objectives were to:
1. Estimate the likely impact of the first half of the project (2012-’14) in terms of deaths averted and
lives saved using population-level coverage estimates obtained at baseline and at mid-term.
2. Project the probable future impact of the second half of the project (2015) with the current suite of
interventions, based on baseline and mid-term coverage estimates and establishing a temporal trend.
Protocol
The analyst attended an eight week training on the Spectrum software and conducting analyses with the LiST
module at Johns Hopkins University Bloomberg School of Public Health (JHSPH) prior to the MTR, followed by
additional refresher trainings during the MTR. The data collection and analyses were carried out over a four
week period between August and September 2014. No ethical approval was sought due to the lack of collecting
primary, human-subjects data.
The modelling process began with verifying coverage data from default sources and project baseline/LQAS. AIM
Health interventions and indicators were mapped to LiST interventions and indicators for comparison. The
decision for inclusion in final models was largely based on: degree of match between AIM Health indicator and
LiST indicators, and completeness of indicators (at least 2 time points). A baseline projection was created
adjusting national demographic, mortality, and coverage values to ADP levels. Additional change projections
were created through input of data for the relevant years. Analysis was carried out through extraction of results,
such as deaths, from the LiST software.
LiST Specifications
Projections were created with Spectrum version 5.0736. The base year of coverage and first year of intervention
were 2011 (the unifying year for sourcing data). The Mutonguni ADP AIM Health project area population was
49,055 in 2011 and population adjustments were made to align to that value. Other demographic data for the
ADP area (sourced and LiST derived) are listed in Table 4. Most default data were from the Kenya
Demographic and Health Survey 2008-2009. Other baseline coverage values, more specific to the region, were
updated from various sources (Table 5). Baseline mortality and fertility rates were sourced from Multiple
Indicator Cluster Survey 2008 (Kitui District) estimates.
When indicators had multiple potential source values, inclusion as the baseline was prioritized by relation to
project area (project/regional values chosen over national values) and year (more recent values chosen over older
values). For indicators where data was unavailable or unreliable, no coverage (0%) was assumed. These indicators
were not included in the analysis and did not have any effect on the results because coverage was not artificially
inflated to an observed percent.
HIV treatment coverages (PMTCT/ARV) were flattened from 2011 onward to eliminate any effects since the
project did not specifically focus on HIV treatment and the decision was made to not include it in the analyses. As
part of a sensitivity analysis, a projection was made without flattening these estimates and no additional lives were
saved. Vaccination coverage was not used in the models at all (flattened from 2011 onward). The analyst could
not verify the project data quality, as the indicator was listed as “percent fully immunized” without specification
of which immunizations constituted full immunization and multiple values were present in the baseline report for
individual immunization coverages. National immunization coverage (WHO/UNICEF and MOH estimates)
actually dropped in the project years.
Coverage Information
16
Three scenarios were created for the main analyses. These scenarios used values for the included AIM Health
interventions (Table 6) based on the above criteria. Table 7 lists a summary of the magnitude of changes in
intervention coverages. Midterm values moved between 1 and 31 percentage points from baseline (12.7 on
average). Endline values were scaled up between 2 and 66 percentage points (21.7 on average) from midterm
(interventions with values above 90% were held constant). Midterm and endline target values were sourced from
the AIM Health Programme RBF. Most interventions were population-level outcome indicators, but others were
listed as output indicators. The output indicator target values, which were among the population visited by
CHWs, were included because they were at a realistic level of increase from the baseline values.
AIM_baseline-point: AIM Health baseline coverage (2011) entered for the selected interventions; a
counterfactual scenario of no change in coverage for any year.
AIM_midterm-point: AIM Health LQAS coverage (2014) entered; a scenario of linear change from
2011 to 2014.
AIM_endline-targets: AIM Health endline (2015) target coverage entered; a scenario of expected
change from 2011 baseline and 2014 midterm values until the end of the program.
An additional scenario was created as secondary analyses:
AIM_midterm-targets: AIM Health MTR (2014) target coverage entered for those interventions which
did not meet targeted values (i.e. improved water source, ORS, Vitamin A supplementation, and ITN);
all others entered as the 2014 point value.
Table 7: LiST Intervention Coverage Summary
Intervention 2011 (%) 2014 (%) Change between 2011-2014 (%)
2015 (%) Change between 2014-2015 (%)
Antenatal care 43.5 58 14.5 70 12
Skilled birth attendance (facility births)
54 76 22 86 10
Postnatal care 20 30 10 32 2
Exclusive breastfeeding 83 91 8 91* 0*
Complementary feeding 50 65 15 80 15
Vitamin A supplementation
69 80 11 90 10
Improved water source 50 19 -31 85 66
Improved sanitation 91 90 -1 90* 0*
ITN ownership 70.5 71 0.5 95 24
ORS for diarrhea 55 46 -9 80 34
Case management for pneumonia
55 73 18 95 22
** = Indicator value for 2014 met or exceeded 2015 target values
3.2 Findings for Objective 1 (2014 Midterm) Deaths avertedc in children under 5
Table 8d (see Annexure) reveals that 14 total deaths were averted in all children under 5 years of age in 2014.
Ten of those deaths averted (71%) are specifically in children under one month of age. A total of 27 deaths in
c Deaths averted attribute reduction in mortality due to specific causes by each intervention. It does not assume that individuals are at risk of dying of multiple causes. d In tables on deaths/cases averted, negative numbers signify an increase in deaths due to a particular cause or intervention instead of the expected decrease in deaths. When deaths are broken down by cause or intervention, the numbers may not
17
children under 5 were averted during the project implementation period (2011-2014). In the neonates, increased
coverage of labor and delivery management (skilled birth attendance) prevented 40% of deaths, which would
have been caused by neonatal asphyxia. In children 1-59 months, increased coverage of oral antibiotics (case
management for pneumonia) prevented 36% of deaths, which would have been caused by pneumonia. Figure 4
shows what causes of death were averted in children under 5. A third of deaths averted were deaths which would
have otherwise been caused by pneumonia (n=5). The next two greatest deaths averted by cause were neonatal
asphyxia (n=5, 29%) and neonatal prematurity (n=4, 25%).
Maternal & stillborn deaths averted
Table 9 shows that one maternal death was averted, but that one death is actually made up of partial deaths due
to specific causes. Three intrapartum stillbirths were averted due to increased coverage of labor and delivery
management.
Nutritional status
Table 10 reveals stunting and wasting distributions for children of different age groups as calculated by LiST and
as calculated from the LQAS results both in 2011 and 2014. Although LQAS showed a 10% decrease in stunting
overall, LiST modelled a change in stunting status that was not more than a 2% decrease per age group. Wasting
status and birth outcomes (pre-term and term small for gestational age) were unchanged with changes in
intervention coverage.
Morbidity
Table 11 shows that although breastfeeding and vitamin A prevented over 1,132 cases of diarrhea, many more
cases were added due to decreased coverage of certain WASH (Water and Sanitation Hygiene) interventions in
2014. There is a net -366 cases of diarrhea averted – in other words, this is an additional 366 cases of diarrhea
due to intervention coverage changes. This result is consistent with LQAS findings of a 5 percentage point
increase in diarrhea prevalence over baseline values. Incidence of pneumonia increased in older children, but
LiST only modelled 4 additional cases which is not consistent with the 9 percentage point increase in suspected
acute respiratory illness (ARI) prevalence over baseline values.
3.3 Findings for Objective 2 (2015 Endline) Deaths averted in children under 5
Table 12 reveals that 31 total additional deaths will be averted in all children under 5 years of age if end of
program targets are met for 2015. Fourteen of those deaths averted (45%) will be specifically in children under
one month of age. Figure 5 shows what causes of death will be averted in children under 5. Pneumonia has the
greatest share of deaths averted by cause (n=11, 36%), followed by neonatal asphyxia (n=6, 21%) and neonatal
prematurity (n=5, 17%), and diarrhea (n=4, 14%).
In neonates, increased coverage of labor and delivery management will prevent 5 neonatal asphyxia deaths.
Increased coverage of antenatal corticosteroids will prevent 3 neonatal prematurity deaths. In children 1-59
months, increased coverage of ORS treatment and improved water source will prevent 3 and 1 deaths,
respectively, which would otherwise be caused by diarrhea.
Maternal & stillborn deaths averted
Four intrapartum stillbirths will be averted due to increased coverage of labor and delivery management, and 1
additional antepartum stillbirth will be averted due to increased coverage of syphilis detection (antenatal care) as
shown in Table 13. One additional maternal death will be averted with 2015 level coverages.
always sum up to the total because of rounding and/or fraction values. Only whole numbers were reported due to the ambiguousness of 0.3 of a life saved.
18
Morbidity
Table 14 shows that the expected increase in vitamin A coverage prevents the greatest number of cases of
diarrhea (2,023 cases), followed by improved water source coverage (1,534 cases). Approximately half of
diarrhea cases averted will occur in children between 24-59 months of age. There will be a net 2,992 cases of
diarrhea averted in children under 5 if 2015 targets are met. Cases of pneumonia will continue to occur at similar
rates in older children, with a net -10 cases averted in 2015.
3.4 Mortality Findings Mortality rates
Table 15 shows the modelled trend of different mortality rates for the project area. The intervention coverage
changes led to an 8% reductione for both the stillbirth and under-five mortality rates for 2014. They also led to a
13% reduction in the neonatal mortality rate and a 14% reduction in the maternal mortality rate for 2014.
In 2015, the intervention coverage changes will potentially lead to an 18% reduction in the under-five mortality
rates, a 12% reduction in the stillbirth rate, a 20% reduction in the neonatal mortality rate, and a 20% reduction
in the maternal mortality rate.
Figure 6 shows that (aggregated) components of skilled birth attendance contributed the most to mortality
reduction, followed by oral antibiotics for case management of pneumonia.
Deaths
Tables 16-19 make a comparison of total deaths and cause of death distribution for the 2014 scenario and the
2015 scenario. The lives savedf are approximately the same as deaths averted presented above. A secondary
analysis of MTR target values reveals that an additional 4 deaths would have been prevented in children under 5
(18 deaths averted total) if targets had been met for all interventions
Figure 7 is a graphical representation of Table 17, and Figure 8 is a graphical representation of Table 16.
Both show the trend in total deaths over the project implementation period comparing a scenario of expected no
change in coverage to a scenario that modelled AIM Health measured changes in coverage.
3.5 Limitations A decision was made early on in the analysis process to not include vaccination coverage, but a sensitivity analysis
was conducted. Had national immunization coverage been left as is (declining over the project period), 2
additional lives would have been saved for a less than 1% additional mortality reduction.
Every indicator measured in LQAS was not included in the analyses for LiST. Some indicators only had mid-term
values and no baseline values. Some indicators had both but had differences in how they were measured (e.g.
IPTp). Some indicators were not compatible enough with LiST definitions (i.e. reported hand-washing behavior).
Breastfeeding promotion was used instead of breastfeeding behavior because inadequate data was available for the
other required categories (exclusive vs. predominant vs. partial breastfeeding).
LiST makes special assumptions about antenatal care and delivery care as both are interventions that actually
represent several individual component interventions. Antenatal care components can be manually changed, but
some are calculated from the coverage of antenatal care and are scaled up or down in relation to antenatal care in
total (e.g. syphilis detection). The same goes for delivery care, or skilled birth attendance, with the additional
e Mortality reductions are interpreted as the percent change between comparison and intervention scenarios in the year of interest. f Lives saved takes into account that an individual can be at risk of dying from multiple causes, so is calculated by subtracting the total number of expected deaths in the intervention change scenario from the total deaths of the scenario depicting no intervention coverage changes.
19
assumptions that some component interventions are only at higher levels of health facilities and certain
percentages of women have access to the higher level facilities.
The impact modelled is only as valid as the quality and accuracy of its data inputs. Although LQAS is a sufficient
tool for judging program quality, it is not always a sufficient tool for precise population-level estimation due to its
small sample size requirements.
Finally, these results are also limited by the parameters of the comparison projection. The assumption made was
that with the AIM Health project in this area, the coverage values for all indicators would remain the same
throughout the years. An alternative analysis could have been made with the assumption that coverage may have
changed naturally (possibly using MOH data), and then comparing that to the evaluation results. The analysis as
presented does not take into account the potential contribution of other projects in the area. The impact cannot
be attributed to AIM Health conclusively without additional analyses.
20
4.0 Conclusions & Recommendations
Overall, the perceptions of health status changes in the AIM Health project area were positive, but are not
supported by mid-term quantitative data. Nutritional outcomes, specifically, and care seeking behaviors have
improved. Utilization of health services seems to have increased as a result of the health messages and facility
upgrades. There appeared to be widespread support for the AIM-MOH partnered interventions permeating
throughout the community.
AIM Health is well integrated into MOH strategy to strengthen the community health system. Participants note it
has been successful program at engaging the community and setting a foundation for sustainability. The synergy
created through AIM Health is evident through the noted community ownership and local/national cohesion to
achieve shared visions and objectives around MCHN.
Environmental and socio-economic constraints hamper maximum intervention coverage – primarily poverty and
a challenging landscape. Health services are also constrained by occasionally delayed funding, supply and staff
shortages, and poor infrastructure. The current CHW supervision structure is unsustainable, and CVA members
and CHWs report a need for tools to carry out their work.
AIM Health has had limited staff, but MOH/WVK integration and shared resources is key to successful
implementation of the project. Although grant compliance is satisfactory, participants share that funding has been
slow and the budget has been strict. One important lesson learned is that activities need to be carefully planned
but flexible enough to respond to delays or environmental changes.
The LiST exercise found that the AIM Health project contributed to saving a total of 36 maternal and child lives
(including stillbirths) since the project’s start, including 17 in 2014 alone (Table 20). There is an estimated 37
additional maternal and child lives that can be saved by achieving scaled up intervention coverage to RBF target
levels in 2015. The most important intervention to preventing maternal, neonatal and child deaths is increased
skilled birth attendance, particularly in a facility. Although the components (e.g. clean birth practices, neonatal
resuscitation) are what make the most impact, management of births by skilled health professionals is one of the
premiere ways to ensure child survival.
Table 20: Lives Saved Summary
Lives Saved Summary
Children under 5 years*
Children under 1 month
Maternal Stillbirths Total
2012 5 (3) 1 1 7
2013 9 (7) 1 2 12
2014 13 (10) 1 3 17
2015 31 (15) 1 5 37
Total 58 (35) 4 11 73**
*Inclusive of children under one month **Total until 2014 only: 36
() Not counted in Total column
Although at first glance these numbers seem low, they translate to an 8% reduction in under five mortality and a
14% reduction in maternal mortality. Scaled up coverage for 2015 will fall short of achieving the current RBF
targets of a 25% reduction in both under-five and maternal mortality, but the LiST analysis suggests that an
approximately 20% reduction may be achieved in both groups. The mortality impact modelled by LiST is actually
determined by several factors: intervention coverage (quality and change), intervention effectiveness, baseline
cause of death profile, and population.
The results of this exercise reflect some cautions. First, the project area population is slightly lower than
recommended for use of LiST (50,000), and even with such a high mortality rate, the absolute numbers of deaths
21
will be low. LiST estimated that pneumonia is the biggest killer of children under 5 in this area, so large
reductions in mortality would most likely be achieved with interventions focused on preventing and treating
pneumonia37. Although the choices of interventions for AIM Health are effective, some very effective
interventions had steep drops and others had high baseline values, leaving little room for improvement.
Although the biggest deficit to overcome in the next year is access to improved water source, WASH is not an
important component of the AIM Health strategy. The 2015 analysis shows that big gains can be made with
increased care seeking (and treatment) for pneumonia, increased ORS treatment for diarrhea, and continued
increases in skilled delivery. At the time of the MTR, ICCM training was beginning for MOH partners in the
project area. It is recommended that project activities focus on scaling up this intervention in the next year to aid
in achieving end of program goals. In order to increase program efficiency, reach and to realize its full potential
for sustainability, it is also recommended that the Mutonguni AIM Health project focus on the following key
aspects during the remaining year:
Accelerate advocacy efforts at the county government level and work to strengthen communication
among different levels of the community health system with the purpose of creating sustainable
funding solutions for the intervention components.
Develop and implement a communications strategy to better promote the community ambulance.
Encourage the DHMT to further work on transportation guidelines for health system linkages in the
sub-county.
Continue working with the MOH to further motivate CHW performance with training needs
assessment, refresher and ICCM trainings, procurement of or supply chain management for CHW
kits, and procurement of bicycles for transportation.
As the DHMT continues to develop CHW performance appraisal guidelines, encourage the MOH
in developing an alternative supervision model to ease the current gap. A case for task-shifting as
incentivization could be made, such as creating an advancement opportunity for high-performing
CHWs to move into supervisory roles (with stipend) as they gain more skills. Also strongly
consider the prior recommendation made to combine CHWs and CHEWs supervision with the
outreach strategy for immunization.
CVAs need to be revitalized with the support and tools they need. AIM Health should support
experiential learning visits to similar groups in nearby areas, have CVA groups introduced at
community dialogues/forums, and aid in securing an interface meeting with government
stakeholders.
22
5.0 Reflections on Field Practicum From June to September of 2014, I undertook my master’s field practicum in Kenya. I was engaged as a Global
Health Fellow, along with three other of my classmates, with World Vision – a Christian humanitarian
organization dedicated to improving the lives of children, families, and their communities worldwide. My role
was to conduct an evaluation of one of their health programs in Kenya, primarily focusing on a qualitative study
and impact modelling study. I was also to work with national office staff on implementing LiST as a monitoring
and evaluation tool for their health projects, but that portion of the fellowship had to be canceled due to
scheduling conflicts. I chose to work in Kenya because I had been there once before as a volunteer and wanted to
more fully immerse myself in that context. The fellowship began with a week-long training and orientation
session in the U.S. (Washington, D.C. and Baltimore, MD) before departing to Nairobi, Kenya.
I spent the remainder of June in a training workshop on LiST, and being introduced to World Vision Kenya staff
in the national office and at the ADP. Once all of the fellows were situated, the head consultant on this mid-term
evaluation started sharing the evaluation protocol and instruments to be used with us. Although I did not develop
any of these instruments out right, I was able to provide input during revisions. Data collection started in July,
and I was immediately responsible for putting my graduate training to use by organizing a training of my own
with the hired focus group discussion moderators. I pulled together the resources I had and attempted to make it
an interactive session, but time constraints left us to focus on the essentials only. The ADP staff were really
diligent about getting me to the scheduled interviews and making sure I had the chance to experience other
projects being implemented in the area. As predicted in my classes, it soon became clear that data collection
would take slightly longer than expected, but I was prepared to work intensely on the analysis and write-up in
August.
I constantly reflected on why I felt “this” wasn’t right or “that” could be done better when developing my report
and recommendations, being careful to avoid the cultural barrier scapegoat. As Mark Nichter put it, “Cultural
barrier explanations are commonly based on (mis-) representations of culture as monolithic and the “local” as
both stagnant and somehow juxtaposed to the modern.”g I had to understand that things were done in a manner
that best reflects the complex reality in which health programs are situated. For this practicum, some challenges I
faced was that the quality of data to use for this evaluation (both collected by myself and by others) was poor at
times. Also, I learned a lot about Grounded Theory in my Qualitative Research series, but when faced time
constraints to conduct an analysis of my data, I opted to use the less well known FrameWork Approach.
Although English was the primary language between staff/interviewees and I, sometimes there
miscommunications where one party did not fully understand the meaning of what the other party was saying.
I really enjoyed my time in Kenya, and I am grateful for the opportunity. Although the security situation was
shaky during my time there (because of terrorist attacks on the coast), I met so many wonderful people, tried
new foods, and even got to explore Kitui county and Nairobi. When I pursued this fellowship, I sought a field
experience with an international NGO that tied together knowledge I had gained from the spectrum of my first
year courses. Through experiential learning and professional mentorship, I hoped to intimately improve my skills
in designing and evaluating behavior change interventions.
With this practicum, I learned how to manage expectations from multiple parties, as well as communicate
effectively with my superiors at different levels. I also felt that I was learning the intricacies of managing a large
project: from collection to analysis and changes in the protocol in between. I was able to understand on the
ground realities that surround health programs in Kenya. Overall, I wholeheartedly believe I achieved my
practicum goals, and I am excited to apply this experience to the next phase of my career.
g Nichter, M. (2008). Global health: Why cultural perceptions, social representations, and biopolitics matter. University of Arizona Press.
23
6.0 Annexure
References 1. United Nations Environment Programme. Chapter 2: Millennium Development Goals. In: Kenya: Atlas of Our
Changing Environment. Nairobi, Kenya; 2009:41–66. Available at: http://na.unep.net/atlas/datlas/sites/default/files/unepsiouxfalls/atlasbook_1135/Kenya_Screen_Chapter2.pdf.
2. Kenya National Bureau of Statistics, ICF Macro. Kenya Demographic and Health Survey 2008-09.; 2008. 3. Kenya National Bureau of Statistics. Kitui District Multiple Indicator Cluster Survey 2008.; 2009. 4. National Coordinating Agency for Population and Development, Ministry of Medical Services, Ministry of Public
Health and Sanitation, Kenya National Bureau of Statistics, ICF Macro. Kenya Service Provision Assessment Survey 2010.; 2011.
5. Government of the Republic of Kenya. Kenya Vision 2030.; 2007. Available at: http://www.vision2030.go.ke/cms/vds/Popular_Version.pdf.
6. Republic of Kenya. Health Sector Strategic and Investment Plan July 2012-June 2017. Ministry of Medical Services and Ministry of Public Health Sanitation; 2013.
7. Kenya Ministry of Health. Taking the Kenya Essential Package for Health to the Community: A Strategy for the Delivery of Level One Services.; 2006.
8. Ministry of Medical Services, Ministry of Public Health and Sanitation. Kenya Health Policy 2012-2030. 9. World Vision International. 7-11 Start-up Field Guide.; 2010. 10. Bhutta Z a, Darmstadt GL, Hasan BS, Haws RA. Community-based interventions for improving perinatal and
neonatal health outcomes in developing countries: A review of the evidence. Pediatrics. 2005;115(2):519–617. doi:10.1542/peds.2004-1441.
11. Darmstadt GL, Bhutta ZA, Cousens S, Adam T, Walker N, De Bernis L. Evidence-based, cost-effective interventions: How many newborn babies can we save? Lancet. 2005;365:977–988. doi:10.1016/S0140-6736(05)71088-6.
12. Haws RA, Thomas AL, Bhutta ZA, Darmstadt GL. Impact of packaged interventions on neonatal health: A review of the evidence. Health Policy Plan. 2007;22:193–215. doi:10.1093/heapol/czm009.
13. Jones G, Steketee RW, Black RE, Bhutta Z a., Morris SS. How many child deaths can we prevent this year? Lancet. 2003;362:65–71. doi:10.1016/S0140-6736(03)13811-1.
14. Freeman P, Perry HB, Gupta SK, Rassekh B. Accelerating progress in achieving the millennium development goal for children through community-based approaches. Glob Public Health. 2012;7(February 2015):400–419. doi:10.1080/17441690903330305.
15. Marston C, Renedo A, McGowan CR, Portela A. Effects of community participation on improving uptake of skilled care for maternal and newborn health: A systematic review. PLoS One. 2013;8(2):e55012. doi:10.1371/journal.pone.0055012.
16. Ricca J, Kureshy N, LeBan K, Prosnitz D, Ryan L. Community-based intervention packages facilitated by NGOs demonstrate plausible evidence for child mortality impact. Health Policy Plan. 2014;29(2):204–16. doi:10.1093/heapol/czt005.
17. Adam MB, Dillmann M, Chen M, et al. Improving maternal and newborn health: Effectiveness of a community health worker program in rural Kenya. PLoS One. 2014;9(8):e104027. doi:10.1371/journal.pone.0104027.
18. Olayo R, Wafula C, Aseyo E, Loum C, Kaseje D. A quasi-experimental assessment of the effectiveness of the Community Health Strategy on health outcomes in Kenya. BMC Health Serv Res. 2014;14(Suppl 1):S3. doi:10.1186/1472-6963-14-S1-S3.
19. Peters DH, Tran NT, Adam T. Implementation Research in Health: A Practical Guide. World Health Organization; 2013. Available at: http://who.int/alliance-hpsr/alliancehpsr_irpguide.pdf.
20. Onwuegbuzie AJ, Leech NL, Collins KMT. Innovative Data Collection Strategies in Qualitative Research. Qual Rep. 2010;15(3):696–726.
21. Rhoda D a., Fernandez S a., Fitch DJ, Lemeshow S. LQAS: User beware. Int J Epidemiol. 2010;39:60–68. doi:10.1093/ije/dyn366.
22. Robertson SE, Valadez JJ. Global review of health care surveys using lot quality assurance sampling (LQAS), 1984-2004. Soc Sci Med. 2006;63:1648–1660. doi:10.1016/j.socscimed.2006.04.011.
23. Walker N, Tam Y, Friberg IK. Overview of the Lives Saved Tool ( LiST ). BMC Public Health. 2013;13(Suppl 3):S1. doi:10.1186/1471-2458-13-S3-S1.
24. Walker N, Fischer-Walker C, Bryce J, Bahl R, Cousens S. Standards for CHERG reviews of intervention effects on child survival. Int J Epidemiol. 2010;39:i21–i31. doi:10.1093/ije/dyq036.
24
25. Stover J, McKinnon R, Winfrey B. Spectrum: A model platform for linking maternal and child survival interventions with AIDS, family planning and demographic projections. Int J Epidemiol. 2010;39:i7–i10. doi:10.1093/ije/dyq016.
26. Winfrey W, Mckinnon R, Stover J. Methods used in the Lives Saved Tool ( LiST ). BMC Public Health. 2011;11(Suppl 3):S32. doi:10.1186/1471-2458-11-S3-S32.
27. Friberg IK, Kinney M V, Lawn JE, et al. Sub-Saharan Africa’s mothers, newborns, and children: How many lives could be saved with targeted health interventions? PLoS Med. 2010;7(6):e1000295. doi:10.1371/journal.pmed.1000295.
28. Bryce J, Friberg IK, Kraushaar D, et al. LiST as a catalyst in program planning: experiences from Burkina Faso , Ghana and Malawi. Int J Epidemiol. 2010;39:i40–i47. doi:10.1093/ije/dyq020.
29. Komatsu R, Korenromp EL, Low-Beer D, et al. Lives saved by Global Fund-supported HIV/AIDS, tuberculosis and malaria programs: estimation approach and results between 2003 and end-2007. BMC Infect Dis. 2010.
30. Ricca J, Prosnitz D, Perry H, et al. Comparing estimates of child mortality reduction modelled in LiST with pregnancy history survey data for a community-based NGO project in Mozambique. BMC Public Health. 2011;11(Suppl 3(S35). doi:10.1186/1471-2458-11-S3-S35.
31. Hazel E, Gilroy K, Friberg I, Black RE, Bryce J, Jones G. Comparing modelled to measured mortality reductions: applying the Lives Saved Tool to evaluation data from the Accelerated Child Survival Programme in West Africa. Int J Epidemiol. 2010;39:i32–39. doi:10.1093/ije/dyq019.
32. Gale NK, Heath G, Cameron E, Rashid S, Redwood S. Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Med Res Methodol. 2013;13(1):117. doi:10.1186/1471-2288-13-117.
33. Bellows NM. Vouchers for reproductive health care services in Kenya and Uganda: Approaches supported by financial cooperation.; 2012.
34. Janisch CP, Albrecht M, Wolfschuetz A, Kundu F, Klein S. Vouchers for health: A demand side output-based aid approach to reproductive health services in Kenya. Glob Public Heal Heal. 2010;5(6):578–594. doi:10.1080/17441690903436573.
35. Arur A, Gitonga N, O’Hanlon B, Kundu F, Senkaali M, Ssemujju R. Insights from innovation: Lessons from designing and implementing FamilyPlanning/Reproductive Health Voucher Programs in Kenya and Uganda. Bethesday, MD: Abt Associates; 2009.
36. Plosky WD, Stover J, Winfrey B. The Lives Saved Tool: A Computer Program for Making Child Survival Projections. USAID; 2011. Available at: http://www.jhsph.edu/research/centers-and-institutes/institute-for-international-programs/_documents/manuals/list_manual.pdf.
37. Liu L, Johnson HL, Cousens S, et al. Global, regional, and national causes of child mortality: An updated systematic analysis for 2010 with time trends since 2000. Lancet. 2012;379(12):2151–2161. doi:10.1016/S0140-6736(12)60560-1.
25
Programmatic Context Figure 1: Map of Kitui County
Source: http://www.kitui.go.ke/
Figure 2: AIM Health MTR Timeline
26
Figure 3: LiST Modelling Examples
Key: Cause of death; Disease specific treatments; Risk factors; Disease specific preventions; Distant factors
27
Data Collection Summary
Table 2: Qualitative Data Collection
Respondent Category Location/Community Date
Focus Group Discussions
Mothers Kivani (N=9) July 8
Kiseveni (N=5) July 14
Husbands/male partners Musengo (N=10) July 10
Katutu (N=8) July 15
CHWs Mithini (N=10) July 10
Mutanda (N=7) July 11
COMM Kauwi (N=7) July 8
Katutu (N=7) July 11
CVA Katutu (N=4) July 15
In-Depth Interviews
PHU/HP in-charge Kivani July 3
District Medical Officer Kauwi July 17
District M&E officer Kauwi July 4
District Hospital (senior clinical person) Kauwi July 4
Liaison/Point person within DHMT Kauwi July 9
Local Council/District Chairman Mithini July 16
AIM Project Officer Kitui July 23
ADP Manager Mutonguni July 24
AIM Project Manager Kitui July 24
Base/regional manager Mutonguni July 22
National Technical Coordinator Kitui July 24
WV M&E officer - Base/region Nanyuki July 29
Grant Finance Manager – Base/region Mutonguni July 22
Grant Finance Manager – National Nairobi July 24
National MOH point person Kitui July 23
Table 3: Survey Results
Indicator 2011 Baseline Survey (%) 2014 LQAS (%)
6-59m stunted 47.4 38
6-23m underweight 20 12
0-23m initiated on breastmilk within 1 hour 43.6 71
0-5m exclusively breastfed 83.1 91
6-59 with adequate iron intake 71.9 53
6-23m minimum meal frequency 50 77
6-23m diverse diet 50 79
PW with adequate iron intake (iron-rich food) 64.7 67
PW who took diverse diet 50 90
PW took iron previous day 72.2 46
6-59m vitamin A supplement past 6m 69.2 80
12-59m fully immunized 92.6 64
0-59m with ARI 12 21
0-59m with fever 30.7 49
0-59m with diarrhea 15.3 20
0-59m with ARI taken to facility in 24h 50 73
0-59m with diarrhea given ORT 54.7 46
0-59m with fever taken to facility in 24h 58.2 43
28
0-59m slept under LLIN 70.5 71
caregivers knowing 3 MTCT 84.2 4
HH with unrestricted access to safe water 50.4 19
HH with access to sufficient water 50.4 46
HH with access to improved sanitation 91 90
caregivers washing hands 2/4 times 87.6 69
0-59m birth attended by SBA 53.6 76
mothers of children 0-59m who had 4 ANC 43.5 58
mothers of children 0-59m who had 3 PNC 20 30
caregivers with intent to use spacing 51.6 77
0-59m with birth certificates 18.5 16
29
LiST Data Collection
Table 4: Demographic Data
Demographic Information Value Source of data & year Comments
Population (ADP) 49,055 AIM Health Baseline (2011)
Women of reproductive age 23.9% 11,722
Spectrum/DemProj (2011)
Children under 5 17.1% 8,366
Spectrum/DemProj (2011)
Sex ratio at birth 103 Spectrum/DemProj (2011) Males to females
Total Fertility Rate (TFR) 4.81 Spectrum/DemProj (2011) Recalculated as trend from MICS (2008) TFR of 5.1 for Kitui District
Crude birth rate 37.6 Spectrum/DemProj (2011) Per 1000 population
Annual # of births 1,843 Spectrum/DemProj (2011)
Poverty 63.7% Kenya Integrated Household Budget Survey (2005-2006)
Kitui District – Population living below poverty line of 1,562 KES per month
Average household size 4 Kenya DHS (2008)
Table 5: National/Region Coverage Levels Update
(Only including indicators which differed from Spectrum/LiST defaults)
Indicator LiST Default Coverage (%)
Updated Coverage (%)
Source of updated data & year
Comments (proxy indicators, other)
Postnatal care (thermal and clean practices)
6.9% 15.5% Kenya DHS (2008-2009)
Recalculated to reflect the LiST definition as specified in LiST.
Iron folate supplementation
0% 2.5% Kenya DHS (2008-2009)
IPTp 0% 25.4% Kenya MIS (2010)
Policy recommendation in place, but not for project area.
Tetanus toxoid vaccination
73% 15.5% Kitui District MICS (2008)
Water connection in the home
20.2% 12.6 WHO/UNICEF Joint Monitoring Programme (2012)
IPTp 25.4% 57.8% AIM Health (2011)
Hygienic disposal of children's stools
78.4% 49.2% MICS (2008)
Zinc - for treatment of diarrhea
0% 0.9% DHS (2008-2009)
Antimalarials - Artemesinin compounds for malaria
10.6% 12.7% MICS (2008)
30
Table 6: AIM Health Indicators and Values
Indicator in LiST
Indicator in AIM Baseline/ other sources
2011 Baseline Value (%, n, 95% CI)
2014 LQAS Value (%, n, 95% CI)
Target for MTR (%)
Target for 2015 (%)
Any departure from LiST definitions
Assumptions, quality of indicator, and source of information
Antenatal care % of pregnant women with at least 4 antenatal care visits
% mothers of children 0-59m who had 4 ANC
43.5% 58%, 128, [65.6, 50.4]
56.6% 69.6% Population difference (mothers vs. pregnant women)
The postnatal and antenatal visits were subject to recall bias (source: baseline report.
Skilled birth attendance % of children born with a skilled attendant present, including doctors, nurses, or midwives, in a facility or at home.
% 0-59m births attended by SBA
53.6%, [57.8, 49.4]
76%, 128, [82.8, 69.2]
69.7% 85.7%
Facility delivery % of infants delivered in a facility
% 0-59m births attended by SBA
53.6%, [57.8, 49.4]
76%, 128, [82.8, 69.2]
69.7% 85.7% The baseline reported SBA and facility deliveries as equal under the assumption that all SBA were in facilities.
Promotion of breastfeeding % of mothers of children 0-11 months of age exposed to a breastfeeding promotion message
% 0-5m exclusively breastfed the day prior to interview
83.1%, [92.2, 74]
91%, 252, [94, 88]
70% 80% LiST recommends using “% of 1-5 month old children exclusively breastfed” as an appropriate indicator.
Clean postnatal practices % of infants with a postnatal health contact/visit within 2 days of birth
% mothers of children 0-59m who had 3 PNC within the first week of birth, with an SBA
20% 30%, 128, [37, 23]
26% 32% The postnatal and antenatal visits were subject to recall bias (source: baseline report.
Complementary feeding--education only
% 6-23m minimum acceptable
50% 65%, 504, [69, 61]
60% 80% LiST uses % of children 6-23 months of age
Baseline did not include value for “minimum
31
% of mothers intensively counseled on the importance of continued breast feeding after 6 months and appropriate complementary feeding practices
diet receiving all 3 age-appropriate IYCF practices, or % of children 6-9 months of age breastfed and receiving complementary foods.
acceptable diet”, just “minimum meal frequency” and diet diversity separately. Because both were equal, an assumption was made that “acceptable diet” would also be equal.
Vitamin A supplementation % of children 6-59 months of age receiving 2 doses of Vitamin A during the last 12 months
% 6-59m vitamin A supplement in the past 6 months (at least 1 dose)
69.2%, [73.3, 64.7]
80%, 772, [82, 78]
85% 90% Clinic cards were often not filled in by health facility staff – vitamin A information was often missing (source: baseline report). LiST recommends using 1 dose in past 6 months if necessary.
Improved water source % of households having an improved water source within 30 minutes
% HH with unrestricted access to safe water
50.4%, [55.3, 45.5]
19%, 128, [24, 14]
75% 85%
Improved sanitation - Utilization of latrines or toilets % of households using an improved sanitation facility
% HH with access to improved sanitation
91% 90%, 128, [95, 85]
75% 85% Slight difference in indicator meaning between “using” and “access”
ITN/IRS - Ownership of insecticide treated nets (ITN/LLIN) or household protected with indoor residual spraying % of households owning at least 1 insecticide treated bed net or protected by indoor residual
% 0-59m slept under LLIN
70.5%, [75.1, 69.5]
71%, 254, [76, 66]
90% 95% LiST acknowledges that "children sleeping under an ITN" is an acceptable substitute, but the estimate will be conservative.
32
spraying
Thermal care % of infants with a postnatal health contact/visit within 2 days of birth
% mothers of children 0-59m who had 3 PNC within the first week of birth, with an SBA
20% 30%, 128, [37, 23]
26% 32% The postnatal and antenatal visits were subject to recall bias (source: baseline report.
ORS - oral rehydration solution % of children with suspected diarrhea treated with oral rehydration solution, including sachets or pre-mixed solutions
% 0-59m with diarrhea given ORS
54.7%, [66, 43.4]
46%, 204, [52, 40]
60% 80% LiST does not consider ORS made using home-available solutions as acceptable, but survey did include these
Oral antibiotics: case management of pneumonia in children % of children 1-59 months with suspected pneumonia or ARI treated with antibiotics
% 0-59m with ARI taken to facility in 24h
54.7%, [66.9, 42.5]
73%, 97, [81, 65]
70% 95% LiST recommends care seeking for suspected pneumonia as a more appropriate indicator
Table 7: LiST Intervention Coverage Summary
Intervention 2011 (%) 2014 (%) Change between 2011-2014 (%)
2015 (%) Change between 2011-2014 (%)
Antenatal care 43.5 58 14.5 70 12
Skilled birth attendance (facility births)
54 76 22 86 10
Postnatal care 20 30 10 32 2
Exclusive breastfeeding 83 91 8 91* 0*
Complementary feeding 50 65 15 80 15
Vitamin A supplementation
69 80 11 90 10
Improved water source 50 19 -31 85 66
Improved sanitation 91 90 -1 90* 0*
ITN ownership 70.5 71 0.5 95 24
ORS for diarrhea 55 46 -9 80 34
Case management for pneumonia
55 73 18 95 22
* = Indicator value for 2014 met or exceeded 2015 target values
33
LiST Results
Table 8: Deaths Averted, Neonates and Children (Midterm)
Neonates (0-1m) Total (0-60m)
Deaths averted in 2014 Deaths averted in 2014
Total: 10 14**
By Cause:
*NN Sepsis NN Asphyxia NN Prematurity
2 5 4
NN Sepsis NN Asphyxia NN Prematurity Diarrhea Pneumonia
2 5 4
-2 5
By Intervention:
Clean birth practices Immediate assessment and stimulation Labor and delivery management Neonatal resuscitation Antenatal corticosteroids for preterm labor Clean postnatal practices
1 1 4 1 2
1
Improved water source ORS Oral antibiotics Clean birth practices Immediate assessment and stimulation Labor and delivery management Neonatal resuscitation Antenatal corticosteroids for preterm labor Clean postnatal practices
-1 -1 5 1 1 4 1 2
1
By Intervention by cause:
NN Sepsis Clean postnatal practices Clean birth practices NN Asphyxia Labor and delivery management Neonatal resuscitation NN Prematurity Antenatal corticosteroids for preterm labor Labor and delivery management
1 1 4 1
2
1
Diarrhea ORS Pneumonia Oral antibiotics NN Sepsis Clean postnatal practices Clean birth practices NN Asphyxia Labor and delivery management Neonatal resuscitation NN Prematurity Antenatal corticosteroids for preterm labor Labor and delivery management
-1
5
1 1
4 1
2
1
*NN = neonatal
** = Inclusive of neonates (0-1m)
34
Figure 4: Deaths Prevented by Cause (Midterm)
*Other includes: NN Diarrhea, NN Pneumonia, NN Tetanus, NN Congenital Anomalies, Diarrhea, Meningitis,
Measles, Malaria, Pertussis, AIDS, Injury, and other unspecified diseases.
Table 9: Maternal Deaths and Stillbirths Averted (Midterm)
Maternal Stillbirths
Deaths averted in 2014 Stillbirths averted in 2014
Total: 1 3
By Cause:
Zero (fractions) Intrapartum 3
By Intervention:
Labor and delivery management 3
By Intervention by cause: Intrapartum Labor and delivery management
3
35
Table 10: Changes in Nutritional Status (Midterm)
Per LiST Per Project Data
2011 2014 2011 (baseline) 2014 (LQAS)
Stunting: > -1 SD* -1 < -2 SD -2 < -3 SD < -3 > -1 SD -1 < -2 SD -2 < -3 SD < -3 SD > -1 SD -1 < -2 SD -2 < -3 SD < -3 SD
6-11m stunted (<-2SD): 26.6% 50.9% 22.4% 15.0% 11.6%
6-11m stunted (<-2SD): 25.6% 51.6% 22.8% 14.5% 11.1%
6-59m stunted (<-2Z): 47.4%
6-59m stunted (<-2Z): 38% 6-11m (<-2Z): 17.4% 6-11m (-2<-3Z): 8.7% 6-11m (<-3Z): 8.7%
12-23m stunted (<-2SD): 44.1% 30.3% 25.6% 25.7% 18.4%
12-23m stunted (<-2SD): 42.3% 31.2% 26.4% 24.7% 17.6%
12-23m (<-2Z): 36.4% 12-23m (-2<-3Z): 22.8% 12-23m (<-3Z): 13.6%
24-59m stunted (<-2SD): 38.6% 30.9% 31.6% 23.3% 14.3%
24-59m stunted (<-2SD): 37.2% 31.0% 31.7% 23.1% 14.1%
24-59m (<-2Z): 46.9% 24-59m (-2<-3Z): 33.6% 24-59m (<-3Z): 13.3%
Wasting > -1 SD -1 < -2 SD -2 < -3 SD < -3 SD > -1 SD -1 < -2 SD -2 < -3 SD < -3 SD
6-11m wasted (<-2SD): 10.9% 71.7% 17.5% 9.7% 1.2%
No change 6-59m wasted (<-2Z): N/A
6-59m wasted (<-2Z): 3% 6-11m (<-2Z): 4.5% 6-11m (<-3Z): 0%
12-23m wasted (<-2SD): 5.1% 81.6% 13.3% 3.6% 1.5%
12-23m (<-2Z): 3.6% 12-23m (<-3Z): 0%
36
> -1 SD -1 < -2 SD -2 < -3 SD < -3 SD
24-59m wasted (<-2SD): 5.9% 79.0% 15.2% 4.2% 1.7%
24-59m (<-2Z): 1.8% 24-59m (<-3Z): 0%
Preterm SGA**
1.98 No change
Term SGA 13.51 No change
Table 11: Cases of Illness Averted (Midterm)
Year: 2014 Diarrhea Pneumonia Meningitis
<1m 5 0 0
1-23m -252 -2 0
24-59m -158 -2 0
Total (0-60m) By intervention
Improved water source Improved sanitation Vitamin A supplementation Breastfeeding
-1,372 -126
1,077 55
<1m by intervention Breastfeeding 5
Table 12: Deaths Averted, Neonates and Children (Endline)
Neonates (0-1m) Total (0-60m)
Deaths averted in 2015 Deaths averted in 2015
Total: 14 31**
By Cause:
*NN Sepsis NN Asphyxia NN Prematurity
2 6 5
Diarrhea Pneumonia Malaria NN Sepsis NN Asphyxia NN Prematurity
4 11
1 2 6 5
By Intervention:
Clean birth practices Immediate assessment and stimulation Labor and delivery management Neonatal resuscitation Antenatal corticosteroids for preterm labor Clean postnatal practices Thermal care
1 1
6 1 3
1 1
Improved water source ORS Oral antibiotics Appropriate complementary feeding ITN/IRS Clean birth practices Immediate assessment and stimulation Labor and delivery management Neonatal resuscitation Antenatal corticosteroids for preterm labor Clean postnatal practices Thermal care
1 4
11 1 1 1 1
6 1 3
1 1
By Diarrhea
37
Intervention by cause:
NN Sepsis Clean postnatal practices Clean birth practices NN Asphyxia Labor and delivery management Neonatal resuscitation NN Prematurity Corticosteroids for preterm labor Labor and delivery management Thermal care
1 1
5 1
3 1 1
ORS Improved water source Pneumonia Oral antibiotics Malaria ITN/IRS NN Sepsis Clean postnatal practices Clean birth practices NN Asphyxia Labor and delivery management Neonatal resuscitation NN Prematurity Corticosteroids for preterm labor Labor and delivery management Thermal care
3 1
11
1
1 1
5 1
3 1 1
*NN = neonatal
** = Inclusive of neonates (0-1m)
Figure 5: Deaths Prevented by Cause (Endline)
*Other includes: NN Diarrhea, NN Pneumonia, NN Tetanus, NN Congenital Anomalies, Diarrhea, Meningitis,
Measles, Malaria, Pertussis, AIDS, Injury, and other unspecified diseases.
38
Table 13: Maternal Deaths and Stillbirths Averted (Endline)
Maternal Stillbirths
Deaths averted in 2015 Stillbirths averted in 2015
Total: 1 4
By Cause:
Zero (fractions Intrapartum Antepartum
4 1
By Intervention:
Labor and delivery management Syphilis detection and treatment
4 1
By Intervention by cause:
Intrapartum Labor and delivery management Antepartum Syphilis detection and treatment
4
1
Table 14: Cases of Illness Averted (Endline)
Year: 2015 Diarrhea Pneumonia Meningitis
<1m 6 0 0
1-23m 1700 -4 0
24-59m 1682 -6 0
Total (0-60m) By intervention
Improved water source Improved sanitation Vitamin A supplementation Breastfeeding
1,534 -617
2,023 52
<1m by intervention Breastfeeding 5
Table 15: Changes in Mortality
2011 2014 (LiST projected from baseline)
2014 (LiST projected with project values)
2015 (LiST projected from baseline)
2015 (LiST projected with target values)
MMR (ratio) 56.6 (360) 53.5 (360) 45.9 (309) 52.5 (360) 42.1 (288)
NMR 36.6 36.6 31.3 36.6 29.1
U5MR 87.8 87.2 80.2 87.1 71.3
Stillbirth rate 19.4 19.4 17.8 19.4 17.1
39
Figure 6: Reduction in Mortality by Intervention (Midterm)
Table 16: Lives Saved, Neonates
Neonates (0-1m)
Deaths in 2014 in no-change scenario (A)
Deaths in 2014 in midterm scenario (B)
Lives Saved (A-B)
Deaths in 2015 in no-change scenario (C)
Deaths in 2015 in endline scenario (D)
Lives Saved (C-D)
Total: 70 60 10 71 56 15
By Cause:
Diarrhea 1 Sepsis 13 Pneumonia 3 Asphyxia 18 Prematurity 25 Tetanus 1 Congenital anomalies 5 Other 4
Diarrhea 1 Sepsis 12 Pneumonia 3 Asphyxia 13 Prematurity 21 Tetanus 1 Congenital anomalies 5 Other 4
Sepsis 1 Asphyxia 5 Prematurity 4
Diarrhea 1 Sepsis 14 Pneumonia 3 Asphyxia 18 Prematurity 25 Tetanus 1 Congenital anomalies 5 Other 4
Diarrhea 0 Sepsis 11 Pneumonia 3 Asphyxia 12 Prematurity 20 Tetanus 1 Congenital anomalies 5 Other 4
Diarrhea 1 Sepsis 3 Asphyxia 6 Prematurity 5
Comments:
14.3% reduction in mortality. 21.1% reduction in mortality.
40
Table 17: Lives Saved, Children under 5
Total (0-60m)*
Deaths in 2014 in no-change scenario (A)
Deaths in 2014 in midterm scenario (B)
Lives Saved (A-B)
Deaths in 2015 in no-change scenario (C)
Deaths in 2015 in endline scenario (D)
Lives Saved (C-D)
Total: 165 152 13 167 136 31
By Cause:
Diarrhea 8 Pneumonia 24 Meningitis 4 Measles 0 Malaria 4 Pertussis 1 AIDS 5 Injury 9 Other 40
Diarrhea 10 Pneumonia 19 Meningitis 4 Measles 0 Malaria 4 Pertussis 1 AIDS 5 Injury 9 Other 40
Diarrhea -2 Pneumonia 5
Diarrhea 8 Pneumonia 24 Meningitis 4 Measles 0 Malaria 4 Pertussis 1 AIDS 4 Injury 10 Other 41
Diarrhea 4 Pneumonia 13 Meningitis 4 Measles 0 Malaria 3 Pertussis 1 AIDS 4 Injury 10 Other 41
Diarrhea 4 Pneumonia 11
Comments:
7.9% reduction in mortality. If all MTR targets had been met 18 deaths would have been averted (4 additional in children 1-59 months).
18.6% reduction in mortality.
* = Inclusive of neonates (0-1m)
Table 18: Lives Saved, Maternal
Maternal
Deaths in 2014 in no-change scenario (A)
Deaths in 2014 in midterm scenario (B)
Lives Saved (A-B)
Deaths in 2015 in no-change scenario (C)
Deaths in 2015 in endline scenario (D)
Lives Saved (C-D)
Total: 7 6 1 7 6 1
By Cause:
Antepartum hemorrhage 1 Postpartum hemorrhage 2 Hypertensive diseases of pregnancy 1 Sepsis 1 Other indirect 2
Antepartum hemorrhage 1 Postpartum hemorrhage 1 Hypertensive diseases of pregnancy 1 Sepsis 1 Other indirect 2
Postpartum hemorrhage 1
Antepartum hemorrhage 1 Postpartum hemorrhage 2 Hypertensive diseases of pregnancy 1 Sepsis 1 Other indirect 2
Antepartum hemorrhage 1 Postpartum hemorrhage 1 Hypertensive diseases of pregnancy 1 Sepsis 1 Other indirect 2
Postpartum hemorrhage 1
Comments:
14.3% reduction in mortality. 14.3% reduction in mortality.
41
Table 19: Lives Saved, Stillbirths
Stillbirths
Deaths in 2014 in no-change scenario (A)
Deaths in 2014 in midterm scenario (B)
Lives Saved (A-B)
Deaths in 2015 in no-change scenario (C)
Deaths in 2015 in endline scenario (D)
Lives Saved (C-D)
Total: 37 34 3 38 33 5
By Cause:
Antepartum 21 Intrapartum 16
Antepartum 21 Intrapartum 13
Intrapartum 3
Antepartum 22 Intrapartum 16
Antepartum 21 Intrapartum 12
Antepartum 1 Intrapartum 4
Comments
8.1% reduction in mortality. 13.2% reduction in mortality.
Figure 7: Changes in Mortality, Children under 5
42
Figure 8: Changes in Mortality, Neonates
Table 20: Lives Saved Summary
Lives Saved Summary
Children under 5 years*
Children under 1 month
Maternal Stillbirths Total
2012 5 (3) 1 1 7
2013 9 (7) 1 2 12
2014 13 (10) 1 3 17
2015 31 (15) 1 5 37
Total 58 (35) 4 11 73**
*Inclusive of children under one month
**Total up until 2014 only: 36