www.mcsprogram.org
Community-Based Newborn Health 
Promotion in Pastoralist Ethiopia:
The Social Mobilization and Demand 
Creation Project
FINAL ASSESSMENT REPORT
Authors:
Alfonso Rosales
David Shanklin
Abiy Seifu
MCSP is a global USAID initiative to introduce and support high-impact health interventions in 24 priority 
countries with the ultimate goal of ending preventable child and maternal deaths (EPCMD) within a 
generation. MCSP supports programming in maternal, newborn, and child health, immunization, family 
planning and reproductive health, nutrition, health systems strengthening, water/sanitation/hygiene, malaria, 
prevention of mother-to-child transmission of HIV, and pediatric HIV care and treatment. MCSP will tackle 
these issues through approaches that also focus on household and community mobilization, gender 
integration, and eHealth, among others.
This study is made possible by the generous support of the American people through the United States 
Agency for International Development (USAID) under the terms of the Cooperative Agreement 
AID-OAA-A-14-00028. The contents are the responsibility of the Maternal and Child Survival Program and 
do not necessarily reflect the views of USAID or the United States Government.
Cover Photo : David Shanklin
August 2017
Community-Based Newborn Health Promotion in Pastoralist Ethiopia iii
Table of Contents 
Acknowledgments........................................................................................................................ iv
Abbreviations................................................................................................................................. v
Executive Summary.....................................................................................................................vii
Introduction................................................................................................................................... 1
Project Rationale........................................................................................................................... 1
Project Description....................................................................................................................... 2
Results ............................................................................................................................................ 7
Discussion..................................................................................................................................... 10
Conclusion.................................................................................................................................... 11
References.................................................................................................................................... 12
iv Community-Based Newborn Health Promotion in Pastoralist Ethiopia
Acknowledgments 
Many people and organizations contributed to the successful implementation of the study described in this 
document. In particular, the authors greatly appreciated the support and guidance we received from 
Dr. Hilena Kebede, Federal Ministry of Health (FMOH) Child Health Team Leader, and Dr. Lisanu Tadesse, 
FMOH Child Health Team Advisor.
Project planning and, later, project results were considered through a consultative process during meetings 
held in Addis Ababa with many of our colleagues. There are too many individuals and agencies to thank by 
name, but we thank them all for the valuable input they provided during both of these meetings (March 2015 
and August 2016).
Implementation was achieved thanks to the full support of Consortium of Christian Relief and Development 
Associations leadership and administrative staff, as well as the CORE Group Polio Project leadership, 
especially Dr. Filimona Bisrat, Director; Ms. Bezunesh Rogie, Technical Backstop; and Mr. Legesse Kidanne, 
Deputy Director. Thanks for the leadership of Haraghe Catholic Secretariat in Dire Dawa, and especially for 
the tireless work of the Project Manager, Mr. Nebiyu Maseresha. Dr. Fasil Tessema provided invaluable 
support during the project evaluation. We received constructive guidance and input throughout the project 
period from Dr. Abeba Bekele, MCSP Ethiopia Community-Based Newborn Care Project Director. Finally, 
this project was made possible by the interest and direct support of USAID/W Global Health Bureau staff, 
Ms. Lily Kak, and Ms. Nazo Kureshy.
This project is dedicated to the many women and families of the Ethiopia pastoralist communities. It is the 
authors’ sincere hope that these results will inform improved health coverage and outcomes for this 
important population group.
Community-Based Newborn Health Promotion in Pastoralist Ethiopia v 
Abbreviations 
ANC antenatal care
CCRDA Consortium of Christian Relief and Development Associations
CGPP CORE Group Polio Project
CV community volunteer
ENAP Every Newborn Action Plan
EPI essential practices in immunization
FGD focus group discussion
FHC family health card
FMOH Federal Ministry of Health
HCS Haraghe Catholic Secretariat
HDA health development army
HEW health extension worker
MCSP Maternal and Child Survival Program
MDG Millennium Development Goal
MNH maternal and neonatal health
NGO nongovernmental organization
PHCU primary health care unit
RHB regional health bureau
SMDC social mobilization and demand creation
TBA traditional birth attendant
UN United Nations
USAID US Agency for International Development
vi Community-Based Newborn Health Promotion in Pastoralist Ethiopia
Community-Based Newborn Health Promotion in Pastoralist Ethiopia vii
Executive Summary
The Ethiopian government’s commitment and resources have contributed to considerable progress in 
improving child health during the past 15 years. To further reduce newborn deaths (those in the first 28 days 
of life), Ethiopia has implemented a policy framework to support newborn health, reflecting the Every 
Newborn Action Plan approved by the World Health Assembly in May 2014. In this context, the CORE 
Group, through the Maternal and Child Survival Program (MCSP), recently designed and implemented the 
Social Mobilization and Demand Creation (SMDC) Project to test a set of community-based activities to 
further the goal of lowering neonatal mortality.
1
The Ethiopia CORE Group Polio Project (CGPP) developed the SMDC Project by (1) reviewing existing 
government documents on demand generation and social mobilization, (2) conducting a planning workshop, 
and (3) writing an implementation plan for the project by synthesizing the information from the review and 
planning workshop. Implementing partners—Hararghe Catholic Secretariat and Erer Woreda, located in Siti 
Zone, Ethiopia Somali Regional State—were selected for the pilot project based on the accessibility of the 
site, technical capacity, and flexibility of administrative systems to ensure adequate support and oversight of 
project activities.
To achieve outcomes in a short period of time (4 months), the project considered two overarching strategies: 
harmonization of newborn care messages and use of community-based structures, including CGPP 
structures, traditional community structures, and formal government structures, to share and promote the 
selected newborn care messages. 
The project developed a project logic model that identified inputs, processes, outputs, outcomes, and impact. 
HEWs and CVs were the two main cadres selected to deliver the newborn care messages to pregnant women 
and mothers. Project staff received training on a package of guides and materials related to the messages in
November 2015, and in November and December they provided 3-day cascade training to 23 health 
extension workers (HEWs) and 64 community volunteers (CVs). In addition, about 170 community
stakeholders received an orientation to the project and key messages. 
The project used a mixed method, quantitative and qualitative assessment to measure predetermined project 
outputs and outcomes before and after implementation. During the quantitative phase of the assessment, the 
team reviewed records from the three health centers and nine of the 14 health posts in the study and collected
data on seven maternal and newborn health indicators. Results of the quantitative assessment indicated that 
health facility delivery rates doubled (from 7.1% to 14.2%) and registration of pregnant women quadrupled 
(from 7.4% to 30.0%) compared to pre-intervention rates, although both remained lower than desired. Nearly
three in five women (59.3%) who were registered chose to receive antenatal care (ANC) services at a health 
facility. Further, the same proportion (60.0%) received newborn counseling by CVs who visited their 
communities. Impressively, more than half of all women (53.9%) who delivered in a health center during the 
intervention period had previously been registered by the project (and encouraged to use these services). 
Overall, one in four registered pregnant women delivered in a facility center, much higher than the average 
for the area, and probably a low estimate given the short intervention measurement period. Home-based 
newborn visits within 2 days of birth were quite limited, suggesting the many challenges of identifying and 
reaching newborns in a timely way among a low-literacy, remote, and mobile population.
During the qualitative phase, the project conducted focus groups discussions (FGDs) as well as in-depth 
interviews with HEWs, CVs, groups of mothers, and a group of community stakeholders. CVs said they 
considered their communication with HEWs, mostly through mobile phones and meetings, to be adequate.
HEWs stated that this communication should occur more frequently, with the support of a more formal 
system of mobile phone communication. The main barrier to institutional delivery, according to CVs and 
 
1
MCSP/CORE Group conducted the study in cooperation with the Consortium of Christian Relief and Development Associations (CCRDA) 
and Ethiopia’s FMOH. 
viii Community-Based Newborn Health Promotion in Pastoralist Ethiopia
HEWs, was limited transportation, in both directions. Especially mentioned was the lack of transportation 
from health center to home after delivery. 
Mothers who had recently delivered at home identified lack of transport and/or related costs, culturally based 
delivery traditions, perceived lack of quality of care at health facilities, and mother-grandmother influences as
enablers of choosing home delivery. These women viewed the main benefits of CVs’ activities as demand 
creation for ANC and child immunization, exclusive breastfeeding, and newborn danger sign recognition. 
Health providers viewed the identification and referral of pregnant women in the community as CVs’ most
important contribution. Interestingly, they also mentioned that female patients’ change in attitude toward
male providers was an important contribution. 
Discussion/Conclusions
This study shows that community-based pregnancy surveillance can be achieved rapidly and effectively when 
it is integrated into established community-based delivery platforms. Within a 4-month period, about three 
out of five pregnant women who were registered had received ANC and newborn health messages. On the 
other hand, identification and registration of pregnant women does not always translate into service demand. 
In this project, there seems to be a gap of about 40% between those initially identified and registered and 
those who later received ANC and counseling on newborns.
This project also demonstrated an immediate impact on the number of women who sought facility-based 
delivery services, with demand doubling during the pilot. Problems with estimating pastoralist population size 
make it difficult to measure the actual number and proportion of women who are pregnant and in need of 
maternal and neonatal services.
Research is needed to find ways to improve methods to track the changing population size and locations of 
pastoralist groups; to more consistently provide timely local transportation for women who wish to deliver in 
a facility; and to improve communications between health facilities and community-based CVs in order to 
increase postpartum care coverage in the home.
Community-Based Newborn Health Promotion in Pastoralist Ethiopia 1 
Introduction 
This report describes a study conducted by the CORE Group through the Maternal and Child Survival 
Program (MCSP), a global USAID Cooperative Agreement to introduce and support high-impact health 
interventions in 25 high-priority countries with the ultimate goal of ending preventable child and maternal 
deaths. The program focuses on ensuring that all women, newborns, and children have access to high-quality 
health care services.
MCSP/the CORE Group conducted this study in cooperation with the Consortium of Christian Relief and 
Development Associations (CCRDA) and Ethiopia’s Federal Ministry of Health (FMOH). The study 
supports MCSP’s Fiscal Year 2015–2016 work plans.
Project Rationale 
In 1990 an estimated 204 of every 1,000 children in Ethiopia died before the age of 5. In response to this 
appalling statistic, and in keeping with the United Nations Millennial Development Goal (MDG) Four, 
Ethiopia set out to reduce the child mortality rate by two-thirds between 1990 and 2015. The latest data 
confirm that by 2012 the rate had dropped to 68 deaths per 1,000, a massive 67% decline in the under-5 
mortality rate, achieved 3 years ahead of the self-imposed deadline. 
The Ethiopian government’s commitment and resources have contributed to considerable progress in 
improving child health. Bold and ambitious targets were set and backed with real domestic and international 
resources—in particular, through a new National Health Extension Program. This program added more than 
38,000 health extension workers (HEWs) to the government payroll and deployed them to more than 
15,000 health posts across Ethiopia. It has played an important role in reducing Ethiopia’s under-5 mortality 
rate, and it could help the country sustain these gains in the long term.
However, further progress on reducing newborn deaths (those in the first 28 days of life) requires an 
increased commitment and additional resources. To this end, Ethiopia implemented a policy framework to 
support newborn health, reflecting the Every Newborn Action Plan (ENAP) approved by the World Health 
Assembly in May 2014 (WHO 2014). Among other initiatives, a number of global development alliances have 
focused on improving the capacity of facility-based health workers to provide consistent, high-quality
maternal and newborn care services.
In this context, the CORE Group Polio Project (CGPP), through MCSP, recently tested a set of 
community-based activities to meet the goal of lowering neonatal mortality. MCSP designed the Social 
Mobilization and Demand Creation (SMDC) Project to expand high-quality newborn messages and actions at 
the woreda, kebele, and family levels in pastoralist areas of Ethiopia. The project supports the Government of 
Ethiopia’s Community-Based Newborn Care program and the FMOH Technical Working Group on Child 
Health. CGPP promoted implementation of maternal and newborn health (MNH) communication strategies 
and messages harmonized with government programs and implemented in tandem with existing public 
services. International and local nongovernmental organizations (NGOs) leveraged their 
program/community platforms to promote the adoption of healthy practices for women and newborns, 
tapping into the “parent power” advocated in the ENAP. 
The potential targeted populations included communities in the pastoralist areas of Ethiopia (Benishangul 
Gumuz, Dire Dawa, Gambella, Harari, SNNP, Somali, and parts of Oromia), where health extension 
programs and health development army (HDA) programs are almost nonexistent. The program strategy, 
developed during a 3-day planning workshop (described below), was a generic one that can be adapted 
according to the development platforms used by partner agencies in the region. The strategy included three 
main approaches: (1) facilitate collaboration and consensus across the Ethiopian FMOH and NGOs currently 
2 Community-Based Newborn Health Promotion in Pastoralist Ethiopia
implementing immunization activities related to the CGPP in pastoralist areas; (2) develop a consensus-based
package of newborn-related messages to be delivered by a local community-based organization using the 
CGPP platform, in coordination with local stakeholders; and (3) measure feasibility and effectiveness of the 
newborn package through a short-term intervention. 
Project Description
Planning Process and Design
Planning process
CGPP developed the SMDC pilot project to align with ENAP’s strategic objective to promote harnessing the 
power of parents, families, and communities. The planning process included: (1) reviewing existing 
government documents on demand generation and social mobilization; (2) conducting a planning workshop; 
and (3) writing an implementation plan by synthesizing the information from the review and the planning 
workshop. 
The Government of Ethiopia’s key newborn health promotion strategy is to include newborn care messages in 
the family health card (FHC). The card includes about 64 messages on health issues, which HEWs1 and HDA
volunteers2 share with mothers and caregivers during counseling, either individually or in groups. This strategy 
has been relatively effective in the agrarian regions of the country. However, there are no well-documented 
strategies or “best experiences” on promotion of newborn health in the more remote pastoralist and semi￾pastoralist areas.
The planning workshop brought together officials and experts representing the FMOH, regional health 
bureaus (RHBs), the USAID Ethiopia Mission Health Office, CCRDA, CORE CGPP, CGPP member 
organizations, key organizations working on newborn in Ethiopia and in pastoralist and semi-pastoralist areas, 
MCSP Ethiopia, health professional associations and societies, the School of Public Health of Addis Ababa 
University, and others. In addition to sensitizing key partners to the proposed project and getting government 
buy-in to pilot the project, the workshop mapped the experiences of CGPP members and other MNH
partners, reviewed the newborn care messages and associated pictures on the FHC, and proposed strategies
to promote newborn health among vulnerable populations in pastoralist and semi-pastoralist areas.
Following the planning workshop, a detailed implementation plan for the pilot project was prepared. A 
consultant (Abiy Seifu Estifanos) worked with MCSP/CORE Group (David Shanklin and Alfonso Rosales) 
and the CGPP Ethiopia team (Filimona Bisrat) to write the plan. The implementing partners—Hararghe 
Catholic Secretariat (HCS) and Erer Woreda, located in Siti Zone, Ethiopia Somali Regional State—were
selected for the pilot project based on agreed-upon criteria. These criteria included accessibility of the site, 
technical capacity of the implementing partners, and flexibility of the administrative system to ensure 
adequate support and oversight of project activities. The FMOH and Ethiopia Somali RHB also contributed 
to the development and review of the pilot project.
 
1 In most cases HEWs are women who completed grade 10 level education and received 1 year of training before deployment. Two HEWs are 
placed at a health post with an average catchment population of 5,000 to provide basic health promotion, and preventive and curative services. 
There are about 38,000 government-salaried HEWs providing services at more than 16,000 health posts throughout the country. The package 
of services provided by HEWs includes environmental health promotion, family planning, immunization, and maternal and child health services. 2 HDAs are female voluntary community workers who support HEWs through mobilizing the community and providing selected health 
messages. The HDA is organized in such a way that a woman who adopts recommended health behaviors (by implementing the 16-plus health 
packages of the Health Extension Program)—the network lead—is networked with five other women in the neighborhood to disseminate the 
healthy behaviors and practices. About five of these networks are grouped into a health development team that has a leader—Health 
Development Team Leader (HDTLs). The HDTLs meet with HEWs every 2 weeks to review the progress of their work and plan activities and 
logistics that they use to promote healthy behaviors among mothers in their teams.
Community-Based Newborn Health Promotion in Pastoralist Ethiopia 3 
Design 
The development of strategies for the SMDC project was guided by intended outcomes of Ethiopia’s 
Community-Based Newborn Care program, including improved early identification of pregnant women and 
increased coverage of antenatal care (ANC), facility-based assisted delivery, and timely postpartum care. 
Other intended outcomes were improved preparation for clean and safe delivery, improved immediate and 
exclusive breastfeeding, improved skills among women and families for recognizing neonatal danger signs, 
improved care seeking for sick newborns, and completion of treatment. To realize these outcomes, the 
project decided on two overarching strategies.
The first strategy was harmonization of newborn care messages. In the SMDC planning workshop, experts 
representing CORE Group members, other MNH implementing partners, the FMOH, RHB staff, and 
donors thoroughly reviewed the messages on the FHC. They confirmed that the messages represented
current recommendations for messages for women who are pregnant and have recently delivered, and that 
the language and pictures used to depict the messages were appropriate for most families and communities. 
The project kept most of the newborn care messages in the FHC as they were and added a few modifications 
to improve the language and pictures to better reflect the cultural and geographic contexts of pastoralist and 
semi-pastoralist areas. 
The second strategy was use of CGPP/community structures to share and promote the selected newborn care messages. In its 
polio program, CGPP uses community volunteers (CVs) to identify and register pregnant women and 
newborns.3 CGPP built on this structure to deliver messages about newborn care for pregnant women and 
mothers. Furthermore, in response to recommendations made during the planning workshop, the project 
considered coordinating with additional community resources for promoting the newborn care messages. The 
three main community-based structures and influential community groups that project staff and volunteers 
interacted with to implement the SMDC strategy are listed below. 
1. CGPP structures and groups: These groups included CVs, traditional birth attendants (TBAs), and pregnant 
women’s forums. 
2. Traditional community structures: These community structures included religious leaders, clan leaders, 
community-based organizations (idirs, hafoshas, etc.), traditional healers, and elders and males/husbands. 
Other influential community members, such as agricultural development agents, veterinary/animal health 
personnel, and schoolteachers were also included in orientation meetings and follow-up coordination. 
3. Formal government structures: The formal government community-based structures that the project worked
with included regional and woreda-level health officials and primary health care unit (PHCU) staff, 
including midwives and HEWs. 
For the pilot woreda, HEWs and CVs were the two main cadres selected to deliver the newborn care 
messages to pregnant women and mothers. Skills-based training on the selected newborn care 
messages and delivery method was planned for 28 HEWs (an average of two HEWs per kebele) and 
70 CVs (five per kebele) in the 14 kebeles of the pilot woreda. 
To regularly monitor the performance of the SMDC pilot project and improve the quality of 
implementation, the project used four different monitoring activities: 
1. Joint integrated supportive supervision: CGPP members conducted integrated supportive supervision jointly 
with the woreda health offices and health partners working in the woreda. Standard checklists and job aids 
were developed and used to guide supportive supervision. The members reported conducting 43 visits 
during the pilot period.
2. Regular integrated review meetings: The project conducted integrated review meetings to review the progress 
and challenges faced in SMDC for newborn care. The meetings were used to share lessons and best 
 
3 There was not a strong HDA structure in the pilot site. The project used CVs in place of HDAs to disseminate the newborn care messages.
4 Community-Based Newborn Health Promotion in Pastoralist Ethiopia
experiences from best-performing kebeles and to recognize best-performing cadres and local leaders to 
motivate them. The project conducted quarterly review meetings at the woreda level, with a defined agenda 
and participants, and monthly review meetings at the PHCU level.
3. Routine program reports: The project routinely collected program implementation data on SMDC activities, 
including the number of orientations/trainings and supportive supervision/review meetings conducted
and the number of checklists/job aids/training manuals distributed. The routine monitoring data were
used to track the project’s progress in implementing its activities. During regular review meetings, 
corrective measures were proposed to ensure that project implementation was kept on track.
1. When possible, the project took baseline measures of key indicators, including the number and 
proportion of newborns and pregnant women identified, the number provided with counseling on 
newborn care messages, and the number linked with health service facilities. The project logic model 
(Figure 1) summarizes these baseline measures, and the Results section defines and discusses the 
indicators. 
4. Assessment: The project conducted a participatory evaluation of pilot implementation. Project data were
reviewed, stakeholder interviews conducted, and data analyzed to draw conclusions and 
recommendations. The methods and results of the evaluation are detailed in the next section.
Figure 1: SMDC Project Logic Model
Note: CV=community volunteer; FMOH=Federal Ministry of Health; HEW=health extension worker; PHCU=primary health care unity; 
SMDC=social mobilization and demand creation
Implementation
Dr. Abiy Seifu, who led the development of the training guides and tools as well as the project’s detailed 
implementation plan, also conducted the initial training of trainers in November. The training centered on 
two guides, one for training facilitators and one for participants. Participants included staff from CGPP Addis 
Ababa who were responsible for project backstopping, local HCS project staff, staff from the Erer Woreda 
health office, and seven midwives from three health centers. These trained staff then provided 3-day cascade 
training in November and December to 23 HEWs and 64 CVs (Figure 2). Of the 64 CVs, 53 were new; and 
Community-Based Newborn Health Promotion in Pastoralist Ethiopia 5 
of these new CVs, 48 (90%) were TBAs and only seven were literate. Five CVs reportedly left the project 
during the implementation period, resulting in 59 CVs by project’s end. An additional four HEWs received 
on-the-job training by the HCS Project Manager to ensure that there were two trained HEWs per kebele
health post. 
About 170 community stakeholders, including kebele and religious leaders, received an orientation to the 
project and key messages (Figure 2). The orientation consisted of a series of local half-day meetings that 
began during the staff and volunteer training period and continued into the first portion of implementation. 
The purpose of these orientations was to explain project activities, gain the moral and vocal support of these 
leaders, and answer any questions or concerns raised. In turn, these stakeholders visibly supported the 
volunteers during the project, and in some cases, they actually participated in village meetings and home 
visits. Unfortunately, the project did not document this participation during the assessment.
Figure 2: Training Outputs Targeted and Achieved
Indicator Target Actual Result Percentage of 
Target Achieved
HEWs trained 28 23 82.1%
CVs trained 70 64 91.4%
Stakeholders oriented 187 170 90.9%
As described in the previous section, a series of supervisory activities supported project implementation. The 
project developed four supervisory checklists to assist in this process, one each for pregnant women, CVs, 
HEWs/health posts, and midwives/health centers. Again, the project did not document supervision 
sufficiently to report results during the assessment. A comment that project staff made consistently during 
this period was that HEWs moved frequently from health post to health post, making consistent follow-up 
with CVs difficult at times.
Assessment
CGPP implemented the SMDC pilot in one woreda to test its feasibility, confirm the acceptability of key 
messages, and improve pregnancy registration. The pilot phase was conducted during a 6-month period
(October 2015–March 2016) in Erer Woreda, in the Siti Zone of the Ethiopia Somali Regional State. In 
2015–2016, Erer Woreda had a population of 97,410 living in 14 kebeles. The annual expected numbers of 
pregnant women and births for the woreda were 3,078 and 3,020, respectively (based on disaggregated 
woreda-level census data). 
The project used a mixed method, quantitative and qualitative assessment to measure predetermined project 
outputs and outcomes. Output indicators included the number of HEWs trained, the number of CVs trained, 
the number of stakeholders oriented to the project, and others (see Figure 1 and Results section). 
Outcome indicators reported here include the numbers and percentages of women who: were added to the 
newborn registry system; were referred and received ANC visits; received community-based newborn care 
messages; delivered at health centers attended by a trained midwife; and received home-based postnatal visits 
within 2 days of birth.
Quantitative phase 
Assessment consultant Fasil Tessema extracted data from records of CVs, health posts, and health centers, 
with the assistance of participating health posts/HEWs and health centers/midwives. The quantitative data 
from health facility registries were summarized based on the indicators developed for project evaluation for 
each health center as well as at the woreda level.
6 Community-Based Newborn Health Promotion in Pastoralist Ethiopia
Because project implementation began in December, basic comparisons of project performance were possible 
by categorizing October and November as the pre-implementation period and December through mid-March 
as the implementation period. The implementation period started when training of CVs and other health staff
was completed. Thus, the project team extracted six months (October–March) of data from ANC, delivery, 
and immunization registries from the three health centers. Because data collection for the assessment 
occurred in late March, most measures ended in mid-March. Woreda-level population, pregnancy, and birth 
estimates were obtained from the 2015–2016 projected population for the woreda, and the 3.16% expected 
pregnant women and 3.1% expected birth rate were used to calculate target estimates. Figure 3 summarizes 
the 2015–2016 project estimates in the health center catchment areas. 
Figure 3: Project Area Population Estimates, Including Expected Pregnancies and Births
Health service
areas
Total
population Annual estimates Monthly estimates
Health center Pregnancies Births Pregnancies Births
Ayidora 30,629 968 949 81 79
Erer 42,805 1,353 1,327 113 111
Hurso 23,976 758 743 63 62
Total 97,410 3,078 3,020 257 252
Qualitative phase
During the qualitative phase of the assessment, the project conducted focus group discussions (FGDs) and 
in-depth interviews. These activities were conducted in Somali and then translated and transcribed into 
English. The main objective of this phase was to assess the appropriateness of training content, acceptability 
and usefulness of training and data-collection tools, and supervisory processes.
Three groups of two persons (facilitator and note taker), supported by assessment coordinators, were 
assigned to conduct FGDs in each health center catchment. Six individuals identified from within the Dire 
Dawa and Somali regions, who had at least a bachelor’s degree and who spoke fluent Somali, English, and 
Amharic, were assigned as FGD note takers and facilitators. All FGDs were recorded on audio recording. On 
a daily basis the notes taken and tape records were translated into/documented in English. The facilitators
and the assessment consultant, supported by the evaluation team, carried out the analysis. 
The consultant and evaluation team gave an orientation and briefing for FGD facilitators on the procedures 
and methods to follow during the FGDs. The consultant prepared a short summary of the general guidelines 
for FGD facilitation, and the orientation team gave it to the facilitators during the orientation, along with the 
topics and questions for each group’s discussion. Each FGD took place at one of the three health centers or a 
nearby public school and lasted no more than one hour.
There were nine FGDs: two groups of HEWs, three groups of CVs, three groups of mothers, and one group
of community stakeholders. Two groups of HEWs (one group of nine and another of eight HEWs) were 
held in the Erer and Hurso Health Center catchment areas. Three groups of CVs (two groups of 10 and one
group of nine) were held in the three health center catchment areas of Erer, Hurso, and Ayidora. One group 
of eight mothers who had delivered at the Erer Health Center and two groups of mothers who delivered at 
home (eight from Ayidora and nine from the Hurso Health Center catchment areas) were interviewed. 
Finally, one group of stakeholders (two women and 10 men) representing the Ayidora Health Center 
catchment area was gathered for a FGD.
Figure 4 summarizes the composition of the focus groups by health center:
Community-Based Newborn Health Promotion in Pastoralist Ethiopia 7 
Figure 4: Composition of Focus Groups 
Category Sex/Age Erer Health 
Center
Hurso 
Health 
Center
Ayidora 
Health 
Center
Total
Health extension 
workers
Female (male) 3 (6) 2 (6) No HEW 
group
17
Age range 22–42 25–45
Community 
stakeholders
Female (male) No community stakeholders group 2 (10) 12
Age range 26 - 60
Community 
volunteers
All are females 10 9 10 29
Age range 38–56 40–62 34–53
Recent mothers
All are females 8 9 8 25
Age range 20–35 19–37 17–40
Opinions regarding the feasibility of using the civil society organization development platform for newborn 
messages and actions were collected from three midwives, one from each health center. Interviewers 
conducted the in-depth interviews mainly in English but also in Amharic if necessary. One person conducted
in-depth interviews with midwives or health officers who attended births and were involved in ANC activities 
at the health centers. Each in-depth interview, conducted face-to-face at the office of the selected respondent,
took up to an hour. Interviewers recorded the interviews so that evaluators could extract responses. Before 
starting each interview, interviewers explained the purpose of the interview and the confidentiality of 
responses to the participant and obtained their informed consent. The qualitative data from the FGDs and in￾depth interviews were transcribed directly and analysis was done based on the topic. 
Results 
Quantitative Assessment
During data collection, the team reviewed records from the three health centers and nine of the 14 health 
posts. The data collected included seven MNH-related indicators (Figure 5).
Figure 5: Indicator Targets and Results
Indicator Definition Target Pre-intervention Intervention 
period result
Estimated percentage 
of pregnant women 
who delivered in a 
health facility
Number of women 
delivering at health 
facility / number 
expected to deliver in 
service area
N/A
7.1%
(36/504)
14.2%
(128/900)
Percentage of pregnant 
women registered by 
CGPP project
Number of pregnant 
women registered / 
number expected to be 
pregnant
50%
7.4%
(38/514)
30.0%
(270/900)
Percentage of pregnant 
women who were 
registered who also 
received antenatal care
Number of pregnant 
women who received 
facility-based antenatal 
care / number registered 
40%
59.3%
(160/270)
8 Community-Based Newborn Health Promotion in Pastoralist Ethiopia
Indicator Definition Target Pre-intervention Intervention 
period result
Percentage of pregnant 
women who were 
registered and who 
received newborn 
counseling from a 
community volunteer
Number of pregnant 
women who received 
community-based 
newborn counseling / 
number registered 
50%
0.0%
(0/38)
60.0%
(162/270)
Percentage of pregnant 
women who delivered 
at the health facility 
and also were 
registered 
Number of women who 
delivered who were 
registered and delivered 
at health facility / 
number who delivered 
at health facility
30%
53.9%
(69/128)
Percentage of pregnant 
women who were 
registered and who 
delivered at a health 
facility
Number of pregnant 
women who were 
registered and delivered 
at health facility / 
number registered
30%
25.6%
(69/270)
Percentage of
registered newborns 
who were visited 
within the first 2 days 
of life in their homes
Number registered 
newborns visited in their 
homes within 2 days / 
number registered 
newborns
30% N=15
In all, 270 pregnant women were registered during the implementation period (3.5 months), compared to 
38 during the previous 2 months. The results for the first two indicators indicate the measurement challenges 
in maternal and newborn health care in this population. The population estimate for this largely pastoralist 
area is subject to large fluctuations due to dynamic fertility rates and outmigration (which in turn is due to 
economic privation and an extended severe drought, not just seasonal migration of herds and families). As a 
result, census-based population estimates (used as the denominator) are subject to question. Estimates based 
on contemporary census status would be better but are technically difficult to achieve. As a result, we know 
only that the low rates of delivery in health facilities and of registration of pregnant women are due to a 
combination of actual limited access and use of services on the one hand and the population estimate 
problem just mentioned on the other hand. Impressively, delivery rates at health facilities doubled, and 
registration of all pregnant women quadrupled compared to pre-intervention rates, although both rates are 
lower than desired. 
Almost three in five registered women (59.3%) chose to receive ANC services at a health facility. Further, the 
same proportion (60.0%) received newborn counseling from CVs who visited their communities. More than 
half of all women (53.9%) who delivered in a health center during the intervention period had previously 
been registered by the project (and encouraged to use these services). Overall, one in four registered pregnant 
women delivered in a facility, much higher than the average for the area and probably a low estimate given 
the short intervention measurement period. Home-based newborn visits within 2 days of birth were quite
limited, suggesting the challenges of identifying and reaching newborns in a timely way among a low-literacy, 
remote, and mobile population. Indeed, it was not possible for the project to count the total number of births 
among registered women outside of health facilities—thus, the lack of a denominator to use to calculate a 
result against the target indicator (see Discussion section). 
Qualitative Assessment
Overall, participants in the FGDs considered the content of the training materials on pregnancy, problems 
related to home delivery, and demand creation for institutional delivery to be appropriate. Participants 
considered the training modules on pregnancy-related danger signs, assessment of a newborn, exclusive 
Community-Based Newborn Health Promotion in Pastoralist Ethiopia 9 
breastfeeding, and postpartum care to be most useful. They found the pictorial information included in the 
manual to be helpful. Nonetheless, since most health volunteers were illiterate, they still needed support from 
literate individuals (most often their own school-aged children) to understand and report on some topics. 
Pictorials most often mentioned by CVs were body edema, headache, and a woman fallen on the ground as a 
sign of hypertension. Among the messages that FGD participants found most interesting were how to 
identify the position of the fetus and how to recognize pregnancy- and delivery-related danger signs. During 
home visits, CVs registered pregnant women, provided messages, and advised mothers to go to the health 
center for delivery. One suggestion for improvements to the messages and training materials was to add 
information on genital mutilation. According to HEWs, one of the main benefits of collecting recording 
forms was that the forms identified pregnant women in the community, which enabled the CVs to encourage 
the women to start ANC and to consider a facility-based delivery.
On average, CVs reported that they spent 2–4 days per week and about 3 hours per day (usually 
8:00–11:00 a.m.) on home visits. CVs considered communication between CVs and HEWs, mostly done 
through mobile phones and meetings at the health post, to be adequate. Nonetheless, HEWs stated that 
communication should be more frequent, supported by a more formal system of mobile phone 
communication.
The main barrier to institutional delivery, according to CVs and HEWs, was limited transportation, in both 
directions. Ambulance services are only available for one-way transportation, from home to health center. 
The lack of transportation from health center to home after delivery is an issue that deters women. The main 
incentives recommended by CVs during FGDs were mobile phones, umbrellas, torches, bags, and frequent 
supervision and refresher training. Monetary incentives for CVs, as well as community and institutional 
recognition for the contribution CVs make to community health, were mentioned only by HEWs. HEWs also 
noted increased awareness of the project among community stakeholders as a means to promote
collaboration with and support from communities. Likewise, they noted that improving communication 
between CVs and the woreda level was important to raise awareness of and address local service bottlenecks 
that impede maternal and newborn service demand and utilization. Recent users of facility-based delivery
services (pregnant and postpartum women) noted that knowledge of obstetric complications (conveyed either 
by CVs or through ANC visits) was the main reason they sought institutional delivery. Most of the time 
women decide on the place for delivery. Likewise, participants reported mothers most frequently make the 
decision to seek care for newborns, with the father only supporting the decision with money.
Severity of newborn illness was associated with continuous crying, fever, and/or not feeding well. Traditional 
healers were the primary health provider sought for newborn illness. Among mothers who had recently 
delivered at home, the main reasons for choosing home delivery were lack of transport and/or the cost of 
transport, culturally based delivery traditions (family support, privacy, and receiving care from TBAs/female 
attendants), perceived lack of quality of care at health facilities, and mother-grandmother influences. Women 
viewed the main benefits of CVs’ activities as demand creation for ANC and child immunization, exclusive 
breastfeeding, and recognition of newborn danger signs. In addition, these women reported that CVs 
promoted community links with the health system.
Community leaders mentioned user-focused incentives, the availability of transportation, and attendance at 
delivery by female providers as the most important interventions to increase institutional delivery. They 
mentioned that their inclusion as social mobilizers within this project was an important vehicle for demand 
creation and service utilization.
Health providers noted that CVs’ most important contribution was the identification and referral of pregnant 
women in the community. Interestingly, they also mentioned the importance of the change in female patients’ 
attitudes toward male providers. They acknowledged that the increase in use of ANC and institutional 
delivery correlated directly with project implementation. Other benefits acknowledged were changes in the 
mothers’ knowledge of danger signs, women’s willingness to continue ANC without interruption, women’s 
willingness to deliver at health centers, and women coming to the health center for care and support when 
they identified a potential health problem. Before the project, mothers reported discarding colostrum and 
10 Community-Based Newborn Health Promotion in Pastoralist Ethiopia
giving camel milk to the newborn, but maternal practices seemed to change for the better during project 
implementation. 
The following quote from a facility-based trained midwife illustrates the comments received during the in￾depth interviews:
“One recently delivered mother brought her newborn with abdominal problems, and when we enquire[d] about the causes, she told 
us that she gave a water and sugar mixture to the child. Then after treating the baby, we provided her with key postnatal 
messages and she told us that she had received similar key messages from CVs but didn’t practice it accordingly and her child got 
sick. She now regrets not practicing what she was told by CVs.” 
Discussion 
Quantitative data from the study indicate that community-based pregnancy surveillance can be achieved
rapidly and effectively by integrating it into established community-based delivery platforms, such as an NGO 
immunization program. Within a 4-month period, about three out of five registered pregnant women received 
ANC and newborn health messages. On the other hand, identification and registration of pregnant women 
does not always translate into service demand. In this project, there seems to be a gap of about 40% between 
those initially identified and registered and those who later received ANC and counseling on newborns.
This project also demonstrated an immediate impact on the number of women who sought facility-based 
delivery services, with demand doubling during the pilot. Problems with estimating pastoralist population size 
make it difficult to measure the actual number and proportion of women who are pregnant and in need of 
maternal and neonatal health care services. Addressing this problem will require creative real-time methods to 
monitor population size and mobility among pastoralist groups. Likewise, it appears difficult for CVs to track 
individual births and provide timely home visits within the first 2 days of life. Perhaps cellular telephone 
communications with health facilities and/or transportation providers will help to resolve this challenge.
On the qualitative side, the most important findings relate to the identification of effective messages to 
increase demand for and use of delivery and newborn services. By targeting pregnant women with counseling 
on essential care for newborns, the project has the potential to improve care-seeking behavior among 
mothers for their newborns. However, solutions to the challenges of timely access and high cost of 
transportation must accompany the counseling intervention. Lack of transportation is a primary barrier to 
service demand, and the availability of local ambulance transportation contributed to women choosing a 
facility-based delivery in this project. A second program enabler was the use of cellular telephones by CVs 
and HEWs to share timely information and notify health facilities and transportation providers.
In this study, participants identified the mother, supported by her husband, as the main decision-maker in
seeking care for ill newborns. The rapid increase in the numbers and percentages of women registered, 
attending ANC, and seeking a facility-based delivery during this brief pilot study suggests that pregnant 
women are open to changing their health behaviors when they are encouraged and supported by local CVs in 
the community. This is an important lesson for the design of interventions to increase service demand and 
thereby decrease newborn mortality in pastoralist populations. 
The finding that women are the decision-makers in seeking care is not consistent with results found in other 
countries. For example, a study in rural Ghana (Hill et al. 2008) found that several people were involved in 
the care-seeking decision process, and that husbands were key in financial decision-making. Based on the 
care-seeking patterns of participating women, the strongest barrier to early skilled care seeking was not lack of 
danger sign recognition, but rather cultural beliefs related to the causes of newborn illness. Studies by Mesko 
et al. (2003) in Nepal and by Bazzano et al. (2008) in Ghana also found that care seeking in the newborn 
period was related to cultural beliefs, such as the need to “watch and wait” (Nepal) and the identification of 
some newborn illnesses as “not for hospital” (Ghana). Stewart et al. (1993) found the ascription of cultural 
Community-Based Newborn Health Promotion in Pastoralist Ethiopia 11 
beliefs (evil forces) to newborn disease, with consequential preference for traditional treatment. It seems that 
the first preference for newborn care is generally home treatment, as has been reported by multiple studies in 
Ghana (Bazzano et al. 2008) and Bangladesh (Winch et al. 2005), as well as for children under 5 in Ethiopia 
(Teka and Dagnew 1995). Most barriers to seeking care outside the home were related either to a sense of 
fatalism or to a lack of finances for transportation.
Conclusion 
The integration of a newborn-related intervention into an immunization delivery platform in a pastoralist area 
of Ethiopia is operationally feasible, and has the potential to rapidly expand coverage of essential newborn 
care, including ANC and facility-based delivery. Research is needed to improve methods to track the changing 
population size and locations of pastoralist groups; to more consistently provide timely local transportation 
for women who wish to deliver in a facility; and to improve communication between health facilities and 
community-based CVs in order to increase postpartum care coverage in the home. Finally, to ensure that
these mobile populations are part of an effective national maternal and neonatal health system, it is crucial to 
support integrated local solutions until an accessible, functioning facility-based system is in place. 
12 Community-Based Newborn Health Promotion in Pastoralist Ethiopia
References 
Bazzano A, Kirkwood BR, Tawiah-Agyemang C, Owusu-Agyei S, Adongo PB. 2008. Beyond symptom 
recognition: Care-seeking for ill newborns in rural Ghana. Trop Med Int Health 13(1): 123–128.
Hill Z, Manu A, Tawiah-Agyemang C, Gyan T, Turner K, Weobong B, Ten Asbroek AH, Kirwood BR. 2008. 
How did formative research inform the development of a home-based neonatal care interventions in rural 
Ghana? Journal of Perinatology 28: S38‒S45.
Mesko, N, Osrin, D, Tamang, S, Shrestha, BP, Manandhar, DS, Manandhar M, Standing H, and Costello, 
AM. 2003. Care for perinatal illness in rural Nepal: A descriptive study with cross-sectional and qualitative 
components. BMC Int Health and Hum Rights 3(12): 3.
Stewart MK, Parker B, Chakraborty J, Begum H. 1993. Acute respiratory Infection (ARI) in rural Bangladesh: 
Perceptions and practice. Medical Anthropology 15: 377–394. 
World Health Organization. 2014. Every Newborn: An action plan to end preventable deaths. Geneva: WHO.
Teka T. and Dagnew M. 1995. Health behaviour of rural mothers to acute respiratory infections in children in 
Gondar, Ethiopia. East African Medical Journal 72 (10): 623‒625. 
Winch PJ, Alam MA, Akther A, Afroz D, Ali NA, Ellis AA, Baqui AH, Darmstadt GL, Arifeen SE, Seraji 
MHR, and Bangladesh PROJAHNMO Study Group. 2005. Local understandings of vulnerability and 
protection during the neonatal period in Sylhet district, Bangladesh: A qualitative study. Lancet 366(9484):
478–485.