External 
Evaluation of
March 2013
Saving Mothers, 
Giving Life
INTERIM REPORT
PHASE 1
External Evaluation of 
Saving Mothers, Giving Life: 
Interim Report
Principal Investigators
Margaret E. Kruk, MD, MPH 
Sandro Galea, MD, DrPH
Co-investigators
Karen Grépin, PhD 
Miriam Rabkin, MD, MPH
Study Team
Tsitsi Masvawure, DPhil 
Katherine Austin-Evelyn, MSc, MPH 
Dana Greeson, MPH 
Emma Sacks, PhD (candidate) 
Daniel Vail, BA
Design: Jon Kalish LLC
© 2013
3
Executive Summary .................................................................. 5
Acronyms & Abbreviations .......................................................... 9
1 Introduction...................................................................... 11
1.1 SMGL background.............................................................. 11
1.2 Aims of the external evaluation and content of this report........................ 11
1.3 Methods ....................................................................... 12
2 Context........................................................................... 13
2.1 Global context.................................................................. 13
2.2 Country setting................................................................. 13
2.3 SMGL districts ................................................................. 18
3 SMGL program theory........................................................... 22
3.1 SMGL in the maternal and newborn care continuum ............................. 22
3.2 SMGL logic model .............................................................. 22
3.3 SMGL theory of change ......................................................... 25
4 Implementation partners and process......................................... 27
4.1 Partners ....................................................................... 27
4.2 Planning and implementation................................................... 29
4.3 Use of PEPFAR platform ........................................................ 30
4.4 Funding ........................................................................ 31
5 Mid-point results ................................................................ 34
5.1 Methods ....................................................................... 34
5.2 Results ........................................................................ 34
6 Successes, challenges, lessons learned ...................................... 39
6.1 Successes ..................................................................... 39
6.2 Innovations .................................................................... 42
6.3 Challenges ..................................................................... 44
6.4 Synergies and unintended consequences........................................ 47
6.5 Country ownership and sustainability ........................................... 49
6.6 Role of the private sector ....................................................... 53
7 Preliminary assessment and recommendations for Phase 2 ................ 55
7.1 Assessment of SMGL’s first six months ......................................... 55
7.2 Recommendations for Phase 2.................................................. 56
7.3 Next steps ..................................................................... 59
8 Appendix ......................................................................... 61
A National health and health system indicators.................................... 62
B Methods ....................................................................... 64
C Acknowledgements ............................................................ 67
D References..................................................................... 70
Contents

Saving Mothers, Giving Life External Evaluation / Interim Report 5
Executive 
Summary
Maternal mortality reduction is a major unfinished agenda in 
global health. More than 250,000 women die every year during 
pregnancy, childbirth and the postpartum period. Saving 
Mothers, Giving Life (SMGL) is a $200 million, five-year initia￾tive of the United States Government (USG), the Government 
of Norway, Merck for Mothers, Every Mother Counts, and 
the American College of Obstetricians and Gynecologists, in 
collaboration with the Governments of Uganda and Zambia 
which seeks to dramatically reduce maternal and newborn 
mortality by simultaneously addressing health system and 
community-level constraints to survival.
SMGL focuses on improving women’s access to high 
quality obstetric care during the highest risk period: the 24 
hours around delivery. In its first year, the initiative aims to 
reduce maternal mortality by 50% in eight rural pilot districts 
in Zambia and Uganda. SMGL activities focus on increasing 
demand for care (e.g. through community mobilization and 
birth planning), access to care (e.g. transport, expanding 
obstetric services), quality of care (e.g. hiring, training health 
workers, ensuring drug supply), and general health system 
strengthening (e.g. improving health records and maternal 
death audits). To support this ambitious program, the US 
government provided substantial funding. The approximately 
$31.5 million USD budgeted for SMGL implementation 
translates to $7.4 and $20.8 per person in the Ugandan and 
Zambian SMGL districts, which represent substantial propor￾tions of existing health budgets in the two countries. 
The external evaluation of SMGL was commissioned by 
the SMGL Global Secretariat and is led by a team of health 
system researchers from Columbia University. The evalua￾tion is a strategic implementation evaluation that aims to: 
1) Measure the extent and fidelity of implementation of SMGL 
interventions; 2) Assess functioning of the partnership and 
engagement of stakeholders; and 3) Identify lessons learned 
to inform the scale-up of SMGL to other districts and coun￾tries. This report summarizes progress in implementation 
and lessons learned from the first six months of the initiative.
Planning for SMGL implementation in Zambia and Uganda 
began in June 2011, spearheaded by the US Department of 
State, the US Agency for International Development (USAID), 
and the US Centers for Disease Control and Prevention 
(CDC). Working on an accelerated timeline, implementation 
of activities began in eight districts in Zambia and Uganda in 
early 2012. Such rapid implementation was made possible by 
the existence of a robust platform for health system support 
established for the President’s Emergency Plan for AIDS Relief 
(PEPFAR). This platform included a range of non-governmental 
organizations with expertise in service delivery and community 
mobilization and existing contractual relationships with USG 
agencies; strong relationships with Ministry of Health plan￾ners; and a bevy of tried and tested approaches for improving 
training, supply chain, and other health system inputs.
Data for this report were collected between October 
2012 and January 2013. We used a mixed methods approach: 
we conducted 143 interviews with US Government, Ministry 
of Health, implementing partners, and district health officials 
and collected quantitative data regarding program startup and 
information on 202 indicators to measure the “dose” of the 
intervention that had been delivered to SMGL districts as of 
December 1, 2012. The evaluation team visited 50 sites, trav￾eling to all eight districts in the two countries, and reviewed 
over 500 documents collected in country. Preliminary findings 
were shared with in-country partners for review and correc￾tions. This report will be followed by a second wave of data 
collection that will focus on the response of communities and 
health providers to SMGL. These data will be presented in a 
final report in fall of 2013.
Mid-point implementation results
Both countries made important progress on implementing 
core SMGL activities, with a slower start for procurement 
of transport vehicles and infrastructure improvements. 
Both Uganda and Zambia trained large numbers of commu￾nity mobilizers (Village Health Teams in Uganda and Safe 
Motherhood Action Groups in Zambia) to promote facility deliv￾eries and birth planning. These activities were supplemented 
with media campaigns and in Zambia through advocacy of 
community “Change Champions.” Both countries extensively 
distributed “mama kits:” packages of basic newborn supplies, 
such as soap and blankets, to encourage facility deliveries.
Activities to increase access focused on improving transpor￾tation and upgrading health facilities to handle emergencies. 
Both countries purchased ambulances—motorized vehicles 
as well as bicycle ambulances—although procurement of 
ambulances and utility vehicles was delayed in Zambia. The 
innovative use of vouchers for transport and private delivery 
care in Uganda, which involved collaboration with private 
health providers and motorcycle drivers’ associations, was 
well-received by communities and providers with demand 
outstripping supply of available vouchers. 
6 MARCH 2013
Considerable efforts were made to improve quality in 
obstetric care in both countries. Notably, SMGL imple￾menters in Uganda hired over 130 new health workers to 
ease the human resource shortages in districts. In the two 
countries, over 400 health workers were trained in clinical 
skills, particularly emergency obstetric and neonatal care 
(EmONC). Supervision and mentoring on the wards by senior 
clinicians was introduced to sustain high quality of care 
and staff morale. 
Some SMGL activities were intended to strengthen 
the health system and assist in coordination of this “big 
push” initiative. Dedicated SMGL coordinators were hired 
in districts to assist District Medical Officers in managing 
logistics and daily operations. Personnel were also trained in 
supply chain management, health worker supervision, and 
management of health information systems to promote more 
accurate and timely information feedback to facilities on 
health indicators and outcomes.
Successes and challenges
Although the health impact of SMGL is not yet known, there 
was a unanimous perception among the over 150 respondents 
interviewed that SMGL has already succeeded on multiple 
fronts. Specifically, respondents perceived an increase in facil￾ity deliveries, increased demand for obstetric services, and 
an improvement in the quality of services at health facilities. 
Community engagement, health worker trainings, improved 
availability of transport, improved data management, provi￾sion of equipment and supplies and facility refurbishment 
were also considered the major early successes. 
Two key success factors cited by Zambian and Ugandan 
stakeholders were the strong alignment between SMGL 
and national health policies and well-functioning national 
partnerships—between Ministry of Health, US Government, 
and a range of implementing partner organizations—that 
formed to implement SMGL. Early support from the Ministry 
of Health was essential in selecting districts and motivating 
district health staff to take on SMGL. Although SMGL funds 
were “off-budget” (not channeled through national govern￾ments), SMGL worked largely through the existing health 
system, rather than pursuing stand-alone solutions. While 
it was noted that some district health and medical officers 
were overwhelmed with the speed and intensity of SMGL 
implementation, all agreed that the buy-in, commitment 
and dedication of district officers was crucial to successful 
launch of the initiative.
Several SMGL activities emerged as particularly innova￾tive. In Uganda, the involvement of the private sector allowed 
for rapid expansion of transport and obstetric services for 
pregnant women. The high level of community participation 
and the connection between community groups and the 
health system were credited with building demand and 
supporting many expectant mothers in reaching health care. 
Finally, innovative technologies were deployed to support 
quality care and link women to the health system, including, 
for example, obstetric “tackle boxes” to ensure availability 
of supplies by every delivery bed and mobile phones to 
promote communication between community health work￾ers, transport providers, and women.
SMGL also encountered some major challenges, several 
related to the short time frame for the first phase. Launching 
so many activities in a short time required a huge organi￾zational effort and participation of many partners—driving 
up transaction costs and making coordination difficult. As a 
result there was duplication of some activities, and delays 
in funding and logistics that prevented timely delivery of 
some inputs. The pressure to deliver results in 12 months 
prevented implementers from tackling deep and urgent 
health system constraints, such as human resource short￾ages and poor transport, beyond implementing quick fixes. 
Respondents recognized that while the “sprint” nature of 
the first phase of SMGL generated excitement, this level of 
energy cannot be maintained over the long run and cannot 
be sustained by USG and its implementing partners alone. 
Crucially, the rapid pace hindered meaningful involvement 
of the Ministries of Health and their few senior reproduc￾tive health staff. Respondents from both USG agencies and 
Ministries of Health emphasized that government support of 
the program was not the same as government ownership: 
in both countries, governments were strongly supportive 
of SMGL but were not in the driver’s seat. Many agreed that 
ownership would have been greater had national govern￾ments made an explicit commitment to SMGL at the outset, for 
example, tackling issues such as roads, human resources, and 
infrastructure that cannot be adequately addressed by donors.
Positive and negative synergies
The introduction of complex new programs is often accom￾panied by a host of secondary effects—positive and negative. 
SMGL clearly energized national efforts in maternal and 
newborn health, nationally and at district levels. Somewhat 
more unexpectedly, it also spurred important actions to 
strengthen health systems. A prominent example is the 
government decision to double the salaries for physicians 
working in health centers in Uganda, for which a key cham￾pion was a parliamentarian from an SMGL district who cited 
the dearth of qualified health workers and the success SMGL 
had in recruiting doctors by paying higher salaries. 
At the district level, “positive spillover” was noted in the 
diffusion of improved clinical skills, record-keeping and 
data management from SMGL to non-SMGL staff. Increased 
demand for facility services led to reported greater uptake 
of HIV and syphilis testing. 
Several negative unintended consequences were noted, 
such as the “crowding out” of other health priorities, namely 
family planning and postnatal care on national agendas, 
EXECUTIVE SUMMARY
Saving Mothers, Giving Life External Evaluation / Interim Report 7
the redirection of USG resources to the focus districts, and 
the diversion of district health team time from non-SMGL 
activities. Over time, these diversions may have serious 
implications for other national health priorities.
The short timeframe also contributed to difficulties in 
sequencing activities; for example, demand creation inter￾ventions began to generate new patients before facilities 
were expanded and new health workers hired. As a result, 
health workers reported report being overloaded with 
patients, and complained of water and electricity shortages, 
and stock-outs of medications. The many off-site trainings 
took health workers away from facilities, exacerbating 
persistent health worker shortages. Finally, some community 
leaders began imposing fines on women who delivered at 
home. Such unsanctioned actions are likely to affect the 
most vulnerable women—the poor and women living farthest 
from facilities—disproportionately.
Assessment and recommendations
Overall, our interim evaluation found a tremendous amount 
of enthusiasm for SMGL in both countries. In both Uganda 
and Zambia, the support for continuing and expanding SMGL 
was strong across the board and largely undiminished by the 
assortment of challenges encountered in the start-up phase. 
The major challenges for SMGL as it scales up are how 
to sustain the early implementation wins, how to address 
endemic health system weaknesses, and how to plan for a 
transition to national ownership. Our primary recommenda￾tions address the content and implementation of SMGL. 
We recommend that SMGL stay the course on 
the following core interventions:
1. Engage communities as change agents: Working 
with community-level groups to address birth planning 
and encourage facility deliveries and timely referrals has 
been a game-changer in both countries. To make this 
sustainable, community workers should be compensated 
or supported in finding income generating activities.
2. Invest in the health system triad: Training staff, 
equipping and improving facilities. This combination is 
highly synergistic in promoting quality care and staff 
retention. These activities represent core health system 
strengthening that will require a greater contribution 
from national governments.
We recommend the following modifications 
to the SMGL package: 
3. Expand the focus beyond 24 hours: SMGL is missing 
a crucial opportunity to save lives by failing to include 
family planning, postnatal care, and HIV health services. 
These activities not only make good health sense, they 
are also efficient to deliver using the same health system 
platform, e.g. facilities, staff, and health records, as delivery 
services. In particular, there is a large unmet need in both 
countries for family planning, which reduces women’s risk 
of maternal death.
4. Develop and test solutions to address the bottlenecks 
of human resource shortages and transport: Health worker 
shortages and poor geographic access to facilities are 
structural constraints for improving maternal and newborn 
survival. These will require innovative solutions: ideas such 
as rural incentives, task shifting, telemedicine, and maternity 
waiting homes should be tested in Phase 2.
Finally, we suggest several ways to improve 
the implementation of SMGL going forward:
5. Coordinate around a single plan: For Phase 2, SMGL 
should strive for greater harmonization of program activ￾ities in each country. One useful model is the “Three Ones 
approach used in HIV:” one action framework, one coordinat￾ing authority and one monitoring and evaluation system.
6. Commit to five years—with a clear transition plan:
Future SMGL commitments should be for three to five years 
to permit more coherent and efficient implementation and 
better integration with the health system. Future commit￾ments should include a clear plan to transition responsibility 
for implementation to countries, clearly spelling out the 
continuing role of external funders. 
7. Collect (a few) standard measures to optimize 
implementation: Reliable and timely data is essential to 
manage the program and make adjustments. A few “tracer” 
indicators, collected as part of countries’ health informa￾tion systems, can be augmented by special studies into 
mission-critical issues such as care quality and the function￾ing of the rest of the health system.
8. Harness the private sector: Reducing maternal death 
requires an “all hands on deck” approach. The private sector 
can contribute expertise in management, marketing, logistics 
and procurement, and human resource management as 
well as raise funds for and awareness of the maternal and 
child death.
Next steps
In the next stage of the external evaluation we will work 
with research partners in Uganda and Zambia to explore: 
1) response of the community and providers to SMGL 2) extent 
of implementation at endline and 3) functioning of the SGML 
partnership. We will produce a final report in fall of 2013.
EXECUTIVE SUMMARY

9
ANC ................. Antenatal Care 
ANC4 ................ Four or more ANC visits
BEmONC ............ Basic Emergency Obstetric and Newborn Care 
CDC ................. Centers for Disease Control and Prevention
CEmONC ............ Comprehensive Emergency Obstetric and Newborn Care 
CIDRZ ............... Center for Infectious Disease Research in Zambia 
CKIs ................. Community Key Informants
CSH ................. Communications Support for Health (Chemonics, Zambia)
DAQS ................ Demand, Access, Quality and Health Systems Strengthening
DHO ................. District Health Office 
DHS ................. Demographic and Health Survey
EGPAF ............... The Elizabeth Glaser Pediatric AIDS Foundation
EmONC .............. Emergency Obstetric and Newborn Care
GDP ................. Gross Domestic Product
HIPS................. Health Initiatives for the Private Sector
HIV .................. Human Immunodeficiency Virus
IDI .................. Infectious Diseases Institute (Makerere University, Uganda)
IHME ................ Institute for Health Metrics and Evaluation
IP.................... Implementing Partner
JICA ................. Japan International Cooperation Agency
JSI .................. John Snow, Inc.
JTI................... Japan Tobacco International
MCDMCH ............ Ministry of Community Development, Mother and Child Health
MCHIP ............... Maternal and Child Health Integrated Program 
MDG ................ Millennium Development Goal 
MMR ................ Maternal Mortality Ratio
MNCH................ Maternal, Newborn and Child Health
MoH ................. Ministry of Health 
MTN ................. Mobile Telephone Network
NESH................ National Expansion and Strengthening of HIV/TB Services (Uganda Protestant Medical Bureau)
PACE ................ Program for Accessible Health, Communication and Education (PSI, Uganda)
PEPFAR ............. President’s Emergency Plan for AIDS Relief 
PMTCT .............. Prevention of Mother to Child Transmission (of HIV)
PPP ................. Purchasing Power Parity
SDS ................. Strengthening Decentralization for Sustainability
SMAGs .............. Safe Motherhood Action Groups (Zambia)
SMGL ............... Saving Mothers, Giving Life
SPEAR............... Supporting Public Sector Workplaces to Expand Action and Responses Against HIV/AIDS
U5MR ............... Under Five Mortality Rate
UCP ................. Uganda Community Project
UHMG ............... Uganda Health Marketing Group
UNAIDS.............. Joint United Nations Program on HIV/AIDS
UNFPA............... United Nations Population Fund
UNICEF .............. United Nations Children’s Fund 
USAID ............... United States Agency for International Development 
USG ................. United States Government 
VHT ................. Village Health Team (Uganda)
WHO ................ World Health Organization 
ZCAHRD ............. Zambia Center for Applied Health Research and Development (Boston University)
ZHECT ............... Zambia Health Education and Communications Trust
ZISSP................ Zambia Integrated Systems Strengthening Project (Abt Associates)
Acronyms & 
Abbreviations

Saving Mothers, Giving Life External Evaluation / Interim Report 11
1
Introduction
1.1 SMGL background
Saving Mothers, Giving Life (SMGL) is a $200 million, five-year initiative of the 
United States Government (USG), in collaboration with the Governments of 
Uganda, Zambia, and Norway, Merck for Mothers, Every Mother Counts, and the 
American College of Obstetricians and Gynecologists. The first phase—the proof 
of concept—aims to reduce maternal mortality by 50% within one year in eight 
focus districts in Uganda and Zambia. In its second phase, the program will 
expand to up to 10 additional countries with high maternal mortality.1
 SMGL’s 
focus is on reducing maternal and newborn deaths in the 24 hours around 
delivery by simultaneously increasing community demand for obstetric care 
and improving the availability and quality of health services. As a public-private 
partnership, SMGL engages governments, non-governmental groups, and private 
sector organizations.1
1.2 Aims of the external evaluation and content of this report
The external evaluation is a strategic implementation evaluation, commissioned 
by the SMGL Global Secretariat to enable course corrections in Phase 1 and to 
inform design of Phase 2. The specific aims of the external evaluation are to: 1) 
Assess the extent and fidelity of implementation of SMGL interventions; 2) Assess 
functioning of the partnership and engagement of stakeholders; and 3) Identify 
best practices and barriers to success in order to improve effectiveness in Phase 
2. The external evaluation complements the internal evaluation led by CDC and 
USAID, which will measure program outcomes, particularly changes in maternal 
deaths and program costs.
The external evaluation consists of two stages and will result in two reports. 
This report summarizes our findings from the first stage in which we: 1) 
Assessed the planning and first six months of SMGL implementation in Zambia 
and Uganda; and 2) Identified early lessons learned in order to provide feedback 
to stakeholders in country and at the global level. It largely reflects the voices 
and experiences of SMGL implementers in Uganda and Zambia. Specifically, we 
discuss program context, program theory, the implementation process to date, 
mid-point results, emerging lessons, and assessment and recommendations for 
Phase 2 of SMGL.
The second stage of the external evaluation will focus on the response of the 
community and health providers to SMGL. We will further assess implementation 
at endline in country and explore the functioning of the global and national part￾nerships. In the second stage of our work, we will collaborate with local research 
partners in Uganda and Zambia. This will culminate in a final report in fall of 2013.
Family and motorcycle driver in 
Nyimba District, Zambia
12 MARCH 2013
1.3 Methods
The external evaluation team is led by Dr. Margaret E. Kruk from the Department 
of Health Policy and Management and Dr. Sandro Galea of the Department of 
Epidemiology at Columbia University’s Mailman School of Public Health. They 
work closely with faculty co-investigators, Dr. Miriam Rabkin (Medicine and 
Epidemiology) and Dr. Karen Grépin (New York University, Health Economics), 
and a dedicated team of five researchers with expertise in public health, health 
systems, anthropology, and quantitative and qualitative methods. 
Data for this report were collected over a total of 134 person-days in country 
between October 2012 and January 2013. The majority of the time was spent in 
the eight SMGL districts. Our team conducted 143 interviews with 152 individu￾als, including 89 in-depth interviews focused on “lessons learned.” Respondents 
included all United States Government (USG) SMGL leaders and key Ministry of 
Health focal points, and a wide range of implementing partners. We interviewed 
all eight SMGL district medical officers, all district SMGL coordinators and a 
number of clinical staff involved in SMGL service delivery. We visited over 50 
sites—including district health offices, implementing partner program offices, 
US Embassies and USG offices, statistical offices, public and private hospitals 
and clinics, and mothers’ shelters—and reviewed over 500 documents collected 
in country. The data collected for the interim report were exempted from 
human subjects review by Columbia University as they concerned program 
implementation activities.
Qualitative data were entered, cleaned, and analyzed using the NVivo Software 
Package (Version 10). Transcripts of all interviews and notes were independently 
coded by two analysts using pre-existing codes and a process of constant 
comparison as described by Glaser.2
 Discrepancies were reconciled through 
a practice of consensus building between the two coders. A second level of 
open, iterative coding was then conducted, creating new codes around major 
themes, which were then reviewed by multiple team members to synthesize 
and summarize. These summaries were the basis of the emerging lessons and 
recommendations in this report. The summaries also provided data for several 
graphics in this report, including the logic model, partner network, and imple￾mentation calendar.
For our quantitative analysis, we assessed the “dose delivered” of SMGL activi￾ties at the mid-point of Phase 1. Dose delivered refers to the activities and inputs 
actually implemented in the districts, as reported by the district medical officers 
and implementing partners in the districts. Using the logic model we created 
with SMGL partner input in country (see Exhibit 5 on page 24), we collected over 
200 data elements on the extent of implementation of 28 SMGL activities by 
December 1, 2012, the SMGL program mid-point. We identified a subset of tracer 
indicators that reflect performance of core SMGL activities for presentation in 
the report. The data were cross-checked against program documents provided to 
the evaluation team by the Ministries of Health and implementing partners and 
shared with in-country partners at district level for their review and additional 
input. Detailed methods used for this evaluation are outlined in Appendix B.
Dose delivered data collection tool
1 INTRODUCTION
Interview at Ntara HC IV, 
Kamwenge District, Uganda
Saving Mothers, Giving Life External Evaluation / Interim Report 13
2.1 Global context
High rates of maternal mortality remain a major challenge in global health. 
More than 250,000 women die every year during pregnancy, childbirth, and the 
postpartum period,3, 4 with the vast majority occurring in low-resource settings 
in sub-Saharan Africa and South Asia. Many more women experience ill-health 
and disability after delivery, ranging from postpartum depression to obstetric 
fistula. In addition, approximately 3 million infants die within the first 28 days 
of life each year,3, 5 representing approximately 40% of deaths of children under 
the age of five.5
 Most of these newborn deaths occur during the early neonatal 
period (0 to 6 days).3
The primary causes of maternal deaths are postpartum hemorrhage 
(i.e. excessive bleeding during delivery and the immediate postpartum period) 
and hypertensive disorders of pregnancy.6
 These conditions are treatable with 
medical interventions, such as the use of uterotonic drugs, magnesium sulfate, 
blood transfusions and Caesarean section. Unfortunately, many women in need 
never receive the appropriate medical services, barred by lack of information 
and social status, difficulties reaching health facilities due to poor roads and 
inadequate transport, and the shortage of quality emergency obstetric services. 
Delays in deciding to seek medical care, in accessing health facilities, and in 
receiving the appropriate care and treatment once at a health facility are the 
three delays that drive high rates of maternal and newborn mortality in low￾and middle-income countries.
In order to avert maternal deaths, all pregnant women must have available and 
affordable emergency obstetric care since up to 15% of deliveries are expected 
to have some complication and most cannot be predicted during antenatal care 
screening.7
 Quality obstetric care and newborn interventions provided after deliv￾ery can also save many newborn lives. SMGL tackles the 24-hour period around 
delivery—the riskiest period for mother and baby.1
2.2 Country setting
Uganda
Uganda is a landlocked country in East Africa with a history of political strife 
and continuing conflict in the northern region and along the Congolese border 
(see Exhibit 1). It has maintained relative political stability since the election of 
President Yoweri Museveni in 1986. With an economic growth rate of 6.7% per 
annum, Uganda received approximately $1.7 billion in net official development 
assistance in 2010.8
 The country has a gross national income of $510 USD 
per capita and the government spends $10 per person on health in current 
US dollars, with 50% of total expenditure on health being paid out-of-pocket.9, 
10 External resources for health are equivalent to 26% of total expenditure on 
2
Context
Staff housing in Nyimba District, Zambia
PHOTO: Dr. Christopher Nyambe Sinyange, 
Nyimba District DMO
14 MARCH 2013
health.10 The country has a crucial shortage of health workers with 1.4 doctors, 
nurses, and midwives per 1,000 population in 2005, far below the international 
recommendation of 2.5 per 1,000.10, 11
Recent estimates of the maternal mortality ratio*
 in Uganda range from 
274 (CI: 206.2-352.3 [IHME]) to 438 (368-507 [DHS]) deaths per 100,000 live 
births.3, 4, 12 The total fertility rate is 6.1.12 Fewer than 20% of Ugandan women 
have more than a primary school education and the median age of women at 
first marriage is 17.9 years.12
Access to health facilities for obstetric care is limited by cost, transport, and 
distance. According to the 2011 Uganda Demographic and Health Survey (DHS), 
57% of deliveries were attended by a skilled birth attendant (doctor, nurse 
or midwife) and only 5% of births were Caesarean sections (compared to the 
expected rate of 10-15%).12 In the same survey, 65% of respondents reported 
difficulties accessing care when needed.12
Uganda faces other pressing health challenges. It has one of the highest 
recorded malaria transmission rates in the world and 33% of children are nutri￾tionally stunted.12 HIV continues to be a major challenge: the HIV prevalence 
among adults 15 to 49 in Uganda is 7.2%, although this has declined substantially 
since the peak of 15% in 1991.12-14 Additionally, Uganda has had outbreaks of viral 
hemorrhagic fevers (Ebola and Marburg Fever) in the past three years, including 
an Ebola outbreak in Kibaale in July 2012 and a Marburg outbreak in October 
2012 in Kabarole, both SMGL districts.
In 2007, Uganda launched a Road Map for accelerating the reduction of mater￾nal and neonatal mortality, in order to meet the Millennium Development Goals 
for health. The stated goals for 2007 to 2015 are:
■■ To increase the availability, accessibility, and utilization of quality skilled 
care during pregnancy, childbirth, and the postnatal period at all levels of 
the health care delivery system;
■■ To promote and support appropriate health-seeking behavior among 
pregnant women, their families and the community; and
■■ To strengthen family planning information and service provision for 
women, men, and couples who want to space or limit their childbearing 
thus preventing unwanted and/or untimely pregnancies that increase 
the risk of maternal death.15
The first two of these goals mirror the focus of SMGL, suggesting that the 
initiative is consistent with national policy.
Zambia
Zambia is a large, sparsely populated, landlocked country in Southern Africa 
(see Exhibit 1). President Michael Sata assumed office after a peaceful transfer 
of power in 2011. Copper, Zambia’s major export, suffered a price decline in 
the 1970s, which made Zambia more reliant on foreign aid. Since the 2000s, 
Zambia’s economy has stabilized and now enjoys a 6.5% economic growth 
rate per year (2011). Zambia received $900 million in net official development 
assistance in 2010.8
* The maternal mortality ratio measures the deaths of women in pregnancy and up to 42 days 
postpartum due to issues related to pregnancy per 100,000 live births. Given that most deaths occur 
around delivery, it is primarily a measure of the safety of childbirth.
2 CONTEXT
Cover of Uganda’s MNCH Road Map
Cover of Zambia’s MNCH Road Map
Saving Mothers, Giving Life External Evaluation / Interim Report 15
Sources
1 WHO, Global Health Observatory Data Repository, 2013
2 World Bank Data , 2011
3 DEMOGRAPHIC AND HEALTH SURVEY, UGANDA: 2011, ZAMBIA: 2007
Uganda
PopulatioN (IN MILLIONS)1 33.4
Total fertility Rate1 6.1
Gross natIONAL income $510
per capita (USD)2
Total health expENDITURE $47
per capITA (USD)1
HIV prevalence rate 7.2
among adults 15-491
Percentage of live births 57%
in health facilities3
ZAMBIA
PopulatioN (IN MILLIONS)1 13.1
Total fertility Rate1 6.3
Gross natIONAL income $1,160
per capita (USD)2
Total health expENDITURE $73
per capITA (USD)1
HIV prevalence rate 12.5
among adults 15-491
Percentage of live births 48%
in health facilities3
EXHIBIT 1 
SMGL focus countries
Mpika
Solwezi
Serenje
Chama
Mkushi
Mumbwa
Kasempa
Chinsali
Kazungula
Namwala
Isoka
Kasama
Mungwi
Samfya
Kapiri Mposhi
Chibombo
Mbala
Chongwe
Mporokoso
Mpongwe
Luwingu
Choma
Kafue
Milenge
Mwense
Lufwanyama
Kawambwa
Chilubi
Chipata
Mambwe
Monze
Katete
Siavonga
Mazabuka
Gwembe
Nakonde
Luangwa
Itezhi-Tezhi
Sinazongwe
Nchelenge
Masaiti
Chadiza
Kabwe
Kitwe
Chingola
Kalulushi
Kafue
Ndola
Mufulira
Luanshya
Chililabombwe
Petauke
Lusaka
MALAWI
ZIMBABWE
DEMOCRATIC 
REPUBLIC
OF THE CONGO
MOZAMBIQUE
Lundazi
Kalomo
Mansa
Nyimba
ZAMBIA
DEMOCRATIC 
REPUBLIC
OF THE CONGO
Kibaale
Kyenjojo
Kamwenge
Kabarole
Lake
George
Lake
Albert
Mbarara Rakai
Mubende
Kiboga
Hoima
Kasese
Bushenyi
Sembabule
Bundibugyo
UGANDA
20 miles 20 miles
ZAMBIA
Lusaka
UGANDA
Kampala
EXHIBIT 2 
Map of health 
facilities in SMGL
focus districts
Mpika
Solwezi
Serenje
Chama
Mkushi
Mumbwa
Kasempa
Chinsali
Kazungula
Namwala
Isoka
Kasama
Mungwi
Samfya
Kapiri Mposhi
Chibombo
Mbala
Chongwe
Mporokoso
Mpongwe
Luwingu
Choma
Kafue
Milenge
Mwense
Lufwanyama
Kawambwa
Chilubi
Chipata
Mambwe
Monze
Katete
Siavonga
Mazabuka
Gwembe
Nakonde
Luangwa
Itezhi-Tezhi
Sinazongwe
Nchelenge
Masaiti
Chadiza
Kabwe
Kitwe
Chingola
Kalulushi
Kafue
Ndola
Mufulira
Luanshya
Chililabombwe
Petauke
Lusaka
MALAWI
ZIMBABWE
DEMOCRATIC 
REPUBLIC
OF THE CONGO
MOZAMBIQUE
Lundazi
Kalomo
Mansa
Nyimba
ZAMBIA
DEMOCRATIC 
REPUBLIC
OF THE CONGO
Kibaale
Kyenjojo
Kamwenge
Kabarole
Lake
George
Lake
Albert
Mbarara Rakai
Mubende
Kiboga
Hoima
Kasese
Bushenyi
Sembabule
Bundibugyo
UGANDA
20 miles 20 miles
ZAMBIA
Lusaka
UGANDA
Kampala
SUMMER FALL
1-100
0
RAINFALL 
IN MM
AVERAGE SEASONAL RAINFALL, 1983-2013
101-200
201-300
301-400
401-500
501-600
WINTER SPRING
Mansa
Nyimba
Lundazi
ZAMBIA
Kalomo
UGANDA
Kibaale
Kyenjojo
Kamwenge
Kabarole
Provincial or 
regional hospitals 
Receive referrals 
from district hospitals
District hospitals 
Receive referrals from 
health centers and 
provide inpatient care
Health centers 
Capable of handling 
vaginal deliveries
Health posts 
Not capable of 
handling deliveries
District health 
offices
Roads
Note: Kalomo district boundaries were recently redrawn—some of the health facilities that are included in the Kalomo SMGL intervention now lie outside the district.
18 MARCH 2013
Zambia has a gross national income of $1,160 USD per capita and the govern￾ment spends $44 per person on health in current US dollars, with 27% of total 
expenditure on health being paid out-of-pocket.10 External resources for health are 
equivalent to almost 40% of total expenditure on health.10 Zambia faces a severe 
health worker shortage with 0.8 doctors, nurses, and midwives per 1,000 in 2006, 
far below the level of 2.5 required to provide near-universal delivery coverage to 
meet the child and maternal health Millennium Development Goals.10, 11
Zambia’s maternal mortality ratio is variously estimated at 293 (223.7-386.7 
[IHME])
3
 and 440 (CI:220-790 [UNFPA])
4
 deaths per 100,000 live births. Despite 
the wide range, these estimates are both lower than the 2007 Zambian DHS 
estimate of 591 deaths per 100,000 live births.16 The total fertility rate is 6.3.16
According to the DHS, only 16% of Zambian women have more than a primary 
school education and the median age at first marriage for women is 18.2 years.17
Many Zambian women have limited access to formal health care, especially in 
rural areas where the population is sparser and transport is more difficult. The 
2007 DHS reports that 47% of deliveries are attended by a skilled birth attendant 
and only 3% of women receive Caesarean sections. Over 77% of DHS respondents 
reported difficulties accessing care when needed.16
As in Uganda, Zambia faces many urgent health challenges. Approximately 
45% of children are considered nutritionally stunted and 18% of children under 
five years reported fever (suspected malaria) in the two weeks preceding the 
survey.16 HIV is a major health threat in Zambia, although HIV prevalence among 
adults has declined from 14% in 2007 to 12.5% in 2011.16-18
Zambia has recently published its Maternal and Newborn Road Map. The 
aim of the policy is to accelerate the reduction of maternal, newborn and child 
mortality to enable attainment of the MDGs in Zambia. The specific objectives in 
the Road Map are to: 1) Provide skilled attendance during pregnancy, childbirth, 
and the postnatal period, at all levels of the health care delivery system; and 
2) Strengthen the capacity of individuals, families, and communities to improve 
maternal, newborn and child health (MNCH).19 This strategy influenced the focus 
and choice of interventions for SMGL in Zambia.
2.3 SMGL districts
The SMGL districts selected for the SMGL pilot phase in Uganda—Kibaale, 
Kabarole, Kyenjojo and Kamwenge—are contiguous districts located in the 
Western part of the country. Their combined population of 1,750,000 people is 
roughly 6% of the national total.9
 The western region is rural and includes the 
Rwenzori Mountains, Kibaale National Forest, and the Crater Lakes. The major 
languages spoken in the region are Rutooro and Runyoro, although many resi￾dents also speak English and Kinyarwanda.
In contrast, the Zambian SMGL districts—Kalomo, Lundazi, Mansa and 
Nyimba—were selected from different regions of the country.20 While Nyanja 
and Bemba are spoken throughout the country and English is the official language 
of business, Zambia recognizes 73 languages. Residents of Mansa, in the north, 
are largely Bemba-speakers. Kalomo residents, in the South, are mostly Chitonga￾speakers, while those in the Eastern Province districts of Lundazi and Nyimba 
speak Chitumbuka, a language also spoken by their Malawian neighbors. Each 
of these districts share borders with different countries.
Exhibit 2 is a map of the SMGL districts in Uganda and Zambia, indicating 
health facilities capable of providing basic and emergency obstetric and neonatal 
care. The map demonstrates the paucity of comprehensive obstetric care (i.e. 
Caesarean section and blood transfusion) in SMGL districts. It also highlights the 
logistical challenges in delivering SMGL interventions to remote areas in Zambia.
“Women walk 
many kilometers—
like 7 to 20—to 
access health 
centers, so 
we thought 
[transport] was 
a very important 
intervention…”
— Zambia District Level
MoH Official
Typical housing in Lundazi District, Zambia 
Flooded road in Kamwenge District, Uganda
2 CONTEXT
Saving Mothers, Giving Life External Evaluation / Interim Report 19
In both countries, the Ministries of Health selected the districts with input from 
the United States Government (USG) development partners. Selection was based 
on high levels of maternal mortality, low socio-economic status, the absence of 
other large maternal health programs, and the presence of strong district lead￾ership and commitment (i.e. from local government and district health offices). 
In Uganda, the Ministry of Health, US Government, and implementing partners 
decided to work in contiguous districts to promote regional health system capac￾ity development and ease implementation. Existing US Government relationships 
with two strong PEPFAR partners with experience in programming for complex 
interventions (i.e. the Infectious Disease Institute [IDI] and Baylor College of 
Medicine Children’s Foundation Uganda [Baylor]) in the western region also influ￾enced the selection of SMGL districts. In contrast, the Zambian Ministry of Health 
focused on areas that were difficult to reach with routine programming and had 
high unmet health needs, selecting districts that were geographically dispersed 
throughout the country. 
The SMGL districts in Uganda and Zambia are predominantly rural with poor 
access to obstetric facilities. Long distances, dirt roads that become impassable 
during the rainy season, limited transport options, and inconsistent or unreliable 
communication, were cited as factors impeding access to delivery and newborn 
care. More generally, the agrarian economies of the districts provide relatively 
few income generating opportunities for communities. Poverty was frequently 
cited in interviews as a determinant of the poor health of the population.
Exhibit 3 summarizes the socioeconomic and health context of the eight SMGL 
districts using data from recent DHS surveys,†
 SMGL country operational plans, 
and district health management information systems. In Uganda, between 72 
and 89% of SMGL district residents are farmers. In Zambia, farmers comprise 
between 14 and 56% of inhabitants. In Uganda SMGL districts, between 2 and 
15% of residents have electricity. In Zambia, less than 4% of residents have elec￾tricity SMGL district, except for Mansa, where 26% of the people have access 
to electricity.
The HIV prevalence in SMGL districts ranges from 8% to 15%. Health systems 
suffer from workforce shortages and poor infrastructure. In Kibaale, the largest 
of the SMGL districts (population 646,500), there are 5 medical officers, 32 clinical 
officers, and 34 nurse/midwives. Kibaale reports a staff vacancy rate of over 50%. 
Similarly, Lundazi, the largest Zambian SMGL district (population 314,300), has 
5 medical officers and 12 clinical officers; however it has 119 nurse/midwives.
There are relatively few obstetric facilities in the SMGL districts. Within the 
focus districts, Uganda has 17 facilities that are expected to provide compre￾hensive emergency obstetric and neonatal care‡
 (CEmONC), however a baseline 
survey found that only 5 of these facilities were able to provide CEmONC; Zambia 
has only 4. Uganda’s SMGL districts have 58 basic emergency obstetric and 
neonatal care (BEmONC) facilities and those in Zambia have 91. Coverage of 
facility deliveries is low in all districts. In the western region of Uganda, approx￾imately 56% of deliveries occur in facilities.21 In Zambia, coverage of facility 
deliveries in SMGL districts ranges from 20% in Nyimba to 46% in Lundazi.16
† NB: DHS surveys are not sampled to be representative at the district level. DHS data should 
thus be interpreted as illustrative.
‡ Basic emergency obstetric care is comprised of: administration of antibiotics, oxytocics, and 
anticonvulsants, manual removal of the placenta, removal of retained products following miscarriage. 
or abortion, assisted vaginal delivery, preferably with vacuum extractor, and newborn care. 
Comprehensive emergency obstetric and neonatal care also includes Caesarean section, blood 
transfusion, and care for sick and low-birth weight newborns.
SMGL billboard in Kamwenge District, Uganda.
2 CONTEXT
“Infrastructure 
is too much of 
a barrier in our 
country. This 
ranges from the 
road the mother 
has to travel, to 
the waiting rooms, 
the buildings etc. 
Deliveries are 
even happening in 
clinics which are 
not equipped.”
— Uganda District Level
MoH Official
UGANDA ZAMBIA
Kabarole Kamwenge Kibaale Kyenjojo Year
SOURCE
Kalomo Lundazi Mansa Nyimba Year
SOURCE
Demographic
Population 409,400 324,400 646,500 369,700 2011 1 254,211 314,281 217,603 101,616 2010 1
Literacy rate (%) 69.2 47.1 61.2 55.2 2011 2 62.8 41.5 62.5 31.0 2007 2
Secondary education or higher (%) 24.0 5.8 32.9 14.9 2011 2 29.5 15.5 44.2 0.0 2007 2
Farmers (%) 71.9 86.6 78.3 88.5 2011 2 56.4 38.5 14.5 37.9 2007 2
Own a telephone (landline) (%) 0.0 2.3 1.5 3.5 2011 2 0.0 0.0 1.9 6.9 2007 2
Have electricity (%) 12.0 2.3 15.3 3.7 2011 2 1.4 3.6 25.9 0.0 2007 2
Improved water supply (%) 57.3 51.7 83.1 27.6 2011 2 38.6 20.2 60.5 10.3 2007 2
Geographic
Annual rainfall, 1983-2012 (mm) 942 1069 1072 1033 AVG 3 710 830 919 818 AVG 3
Health
HIV prevalence (%)* 8.2 2011 4 15.2 15.0 10.0 7.7 2010 6
Maternal deaths, annual** 80 63 126 72 2010 1 58 72 50 23 2010 5
Newborn deaths, annual*** 550 436 869 497 2010 5 1,136 1,460 973 472 2010 5
Deliveries, annual 18,341 14,533 28,963 16,563 2011 1 13,219 16,343 11,315 5,280 2010 1
Health System
Medical officers 21 0 5 3 2011 1 n/a 5 1 3 2011 7
Clinical officers 51 17 32 22 2011 1 n/a 12 7 9 2011 7
Anaesthetic officers 3 0 1 1 2011 1 n/a 1 0 0 2011 7
Nurse/midwives 168 37 34 37 2011 1 n/a 119 98 50 2011 7
Staff vacancy rate (%) 37.0 22.0 > 50.0 32.0 2011 1 n/a 17.5 61.0 n/a 2011 7
Provincial/regional/referral hospitals 4 0 1 0 2011 1 0 0 1 0 2011 1
District hospitals 2 4 5 2 2011 1 1 1 0 1 2011 1
Health centers 17 9 17 15 2011 1 29 23 27 12 2011 1
Health posts 24 14 20 15 2011 1 4 19 1 5 2011 1
Total CEmONC facilities **** 6 4 5 2 2011 1 1 1 1 1 2011 1
Total BEmONC facilities 17 9 17 15 2011 1 29 23 27 12 2011 1
Utilization
Women who completed at least 
4 ANC visits during pregnancy (%) 51.7 43.6 40.0 33.3 2011 2 72.4 61.9 66.0 72.7 2007 2
Women who gave birth 
by caesarean section (%) 6.9 3.6 6.3 1.8 2011 2 3.5 2.2 11.4 0.0 2007 2
Households with bednets (%) 82.7 75.9 77.1 88.5 2011 2 79.5 73.7 89.1 79.3 2007 2
Women who received a vitamin A 
dose within 2 months of delivery (%) 48.3 21.8 20.8 26.3 2011 2 44.8 55.6 60.7 36.4 2007 2
Women who use a modern 
contraception method (%) 37.9 29.1 29.2 22.8 2011 2 29.5 39.2 17.0 44.8 2007 2
Health facility delivery 
coverage (%)***** 55.9 2011 2 27.0 46.0 42.0 20.0 2011 1
* HIV Prevalence for Uganda reported for 
the Western Region as a whole
** Maternal deaths in Zambia calculated based on the 
WHO 2010 Global Health Observatory point estimate of 
440 maternal deaths per 100,000 births
*** Newborn deaths calculated based on the WHO 2010 
Global Health Observatory point estimate of 30 newborn 
deaths per 1,000 births in Uganda, and 86 newborn 
deaths per 1,000 births in Zambia
**** While there are a total of 17 health facilities in the 
Ugandan SMGL districts that are designated as ‘CEmONC,’ 
the baseline SMGL survey found that only five of these 
were actually capable of providing CEmONC services.
***** Facility delivery rates for Uganda reported for the 
Western Region as a whole
Sources
1 SMGL Country Operational Plans
2 Demographic and Health Surveys
3 TAMSAT Research Group
4 Uganda AIDS Indicator Survey
5 WHO Global Health Observatory
6 Zambia National AIDS Council
7 Dose Delivered Interviews with
District Staff
Note: Data from the 
Demographic and 
Health Surveys are 
not representative 
at the district level, 
and are limited by 
small sample sizes. 
Total sample size 
for Uganda across 
all four SMGL 
districts is 418. 
Total sample size 
for Zambian SMGL 
districts is 566.
EXHIBIT 3
District-level 
demographic 
and health 
indicators
Saving Mothers, Giving Life External Evaluation / Interim Report 21
The cultural context also influences care seeking for maternal and newborn 
health and, in turn, the success of initiatives such as SMGL. Several cultural 
norms, common in both countries, were cited as barriers to reducing maternal 
and newborn mortality. Early marriage puts young women at risk as maternal 
deaths are more common among adolescents. In Zambia a district health official 
noted that it is not uncommon to “see girls getting married at 13 or 14.” District 
managers and health workers noted that many women have a strong allegiance 
to traditional birth attendants (TBAs) who conduct deliveries in the home. The 
choice of home delivery was in some cases reinforced by poor health worker 
attitudes in health facilities that women encountered in antenatal care and 
other visits. Interview participants in both countries noted that men rarely take 
an active role in maternal health, especially with regard to birth planning. This 
leaves many pregnant women without financial or logistical support when emer￾gencies arise. 
Some respondents in Zambia reported a local belief that disclosing a preg￾nancy early on will result in a miscarriage, which prevents them from obtaining 
early antenatal care. Finally, participants in both countries mentioned that 
cultural and church groups sometimes speak out against family planning and 
in some cases against other modern maternal health care, including facility 
delivery. The existence of cultural traditions that conflict with efforts to promote 
maternal and newborn survival point to the importance of involving communities 
in SMGL, and the need for explicit communication and advocacy strategies.
2 CONTEXT
“Early marriage 
is not addressed 
and is a key factor 
which contributes 
to high risk 
pregnancies.”
— Uganda District Level
Implementing Partner
Family traveling by canoe to Luembe 
Clinic, which is not accessible by road, 
Nyimba District, Zambia
PHOTO: Dr. Christopher Nyambe Sinyange, 
Nyimba District DMO
22 MARCH 2013
“…without linked 
programs, this 
cannot [continue 
to] work. It is too 
narrow a point 
of view. Are we 
giving the message: 
‘it’s ok to just 
come for delivery’ 
by focusing on 
24 hours?”
— USG Uganda OFFICIAL
3.1 SMGL in the maternal and newborn care continuum
Maternal and neonatal survival is achieved through a chain of individual and 
household actions and health service interventions that begin before pregnancy 
and end one month post-delivery. SMGL interventions tackle the highest risk period: 
the 24 hours around delivery when 60% of maternal deaths and 50% of newborn 
deaths occur.4, 6, 21, 22 Exhibit 4 situates SMGL activities within the larger continuum 
of maternal and newborn care.
As Exhibit 4 shows, improving chances of maternal and newborn survival 
begins before pregnancy. Avoiding unwanted pregnancies reduces the number 
of maternal deaths. A recent study estimated that providing contraception for all 
women who want it but lack access would reduce maternal deaths globally by 
104,000, or 29% of the current total.23 Once women are pregnant, antenatal care 
focused on high-impact services can improve newborn birth weight and provide 
other essential protective functions, such as tetanus prophylaxis and prevention 
of HIV infection.6
While most deliveries are a normal life event, delivery and the postpartum 
period remains a high-risk period for mother and baby as complications can arise 
suddenly and kill quickly. The majority of these can be treated with basic and 
comprehensive emergency obstetric and neonatal care—a package of services 
long known and implemented in developed and many middle-income countries. 
Provision of emergency obstetric and neonatal care in the 24-hour period around 
birth is the focus of SMGL.
SMGL’s concentration on the peripartum period is both a strength and a weak￾ness. It focuses scarce resources on the most critical period for maternal and 
newborn mortality and seeks to strengthen a key weakness in many low-income 
country health systems: poor quality of obstetric care. However, maternal complica￾tions and many neonatal infections also occur after the first 24 hours. Additionally, 
the reduction of HIV-related maternal deaths—an important cause of maternal 
mortality in both countries—requires high quality antenatal care. The omission of 
family planning is also a missed opportunity to avert maternal deaths of women 
who did not want to be pregnant. The latter point in particular was echoed through￾out interviews with SMGL leads, implementing partners, health system managers, 
and facility-level providers, most notably in Uganda.
3.2 SMGL logic model
SMGL simultaneously tackles both supply-side (health system) and demand-side 
(community and user) barriers to the provision of effective obstetric and neonatal 
care, in what can be characterized as a whole-of-health systems approach. SMGL 
activities explicitly address all three of the delays that contribute to maternal 
mortality: delay in seeking care (through community activation and birth planning), 
delay in reaching care (through transport, more obstetric facilities closer to 
3
SMGL program theory
Tetanus toxoid immunization
Screening and treatment for 
HIV and sexually transmitted 
infections (STIs)
Intermittent preventive 
treatment (IPT) for malaria
Treatment of asymptomatic 
bacteriuria
Antiretroviral therapy for 
HIV-infected, eligible women
Emergency care 
(ectopic pregnancy, post￾abortion care)
Folic acid/iron and vitamin 
supplementation
Prevention of mother-to-child 
HIV transmission (PMTCT) 
Family planning
Birth spacing
Folic acid/iron and vitamin 
supplementation
Screening and treatment for 
HIV and sexually transmitted 
infections (STIs)
Prevention of mother-to-child 
HIV transmission (PMTCT)
Corticosteroids for 
preterm labor
Labor surveillance to detect 
complications (partograph)
Identification of breech, 
transverse lie, or multiple 
gestation pregnancy
Basic emergency 
obstetric care:
■■ Parenteral antibiotics 
for infection/premature 
rupture of membranes 
(PROM)
■■ Parenteral anticonvulsants 
for pre-eclampsia/
eclampsia
■■ Uterotonic drugs for 
treatment of hemorrhage
■■ Active management of 
labor and clean delivery
■■ Assisted vaginal delivery 
(vacuum extraction) 
■■ Manual removal of the 
placenta
■■ Manual removal of 
retained products
Comprehensive emergency 
obstetric care:
■■ Caesarean section for 
obstructed labor, etc.
■■ Blood transfusion for 
severe hemorrhage
MATERNAL
Antiretroviral therapy for 
HIV-infected, eligible women
Infection prevention at 
surgical sites (Caesarean 
incision, episiotomy)
Identification and treatment 
of sepsis
NEONATAL
Basic neonatal resuscitation 
(bag and mask)
Management of hypothermia
Promotion of exclusive 
breastfeeding
Detection/treatment of 
pneumonia/sepsis
Antiretroviral therapy for 
HIV-exposed infants
Extra care for low birth 
weight and preterm infants
Skin to skin contact
Early detection and referral 
of complications
MATERNAL
Postpartum family planning
Screening for postpartum 
depression
General nutritional 
counseling
Antiretroviral therapy for 
HIV-infected, eligible women
NEONATAL
Case management for 
pneumonia and tuberculosis
Treatment for malaria, 
use of bednets
Testing and treatment for 
pediatric HIV
Basic childhood 
immunizations
Treatment for diarrhea (oral 
rehydration therapy/zinc)
Growth monitoring 
and micronutrient 
supplementation
Integrated management of 
childhood illness (IMCI)
SMGL FOCUS
PRE-PREGNANCY PREGNANCY BIRTH POSTPARTUM / EARLY NEONATAL LATE POSTPARTUM / LATE NEONATAL
EXHIBIT 4 
Evidence-based clinical interventions to avert maternal 
and neonatal deaths along a continuum of care
OUTPUTS
Utilization
Increase in:
■■ facility deliveries
■■ four antenatal care visits
■■ postnatal care visits
■■ referrals for complications
■■ EmONC
■■ neonatal resuscitation
Quality
Reduction in:
■■ maternal facility case 
fatality rate
■■ perinatal mortality in 
facilities
System strengthening
■■ Stronger supply 
chains and health 
information systems
■■ New operating theaters
■■ More mothers’ shelters
■■ Improved transport and 
communication
OUTCOMES
■■ Reduction in 
maternal mortality
■■ Reduction in 
newborn mortality
■■ SMGL program plan
■■ Funding
■■ USG, MoH, and 
implementing partners
■■ Health workers
■■ Trainers and training 
curricula
■■ Equipment (e.g. laboratory, 
vehicles, scales, delivery 
equipment, etc.)
■■ Supplies (e.g. essential 
medicines, blood supply)
■■ Basic infrastructure (e.g. 
water, electricity)
■■ Guidelines (clinical, data, 
supply chain)
■■ Information and 
communication technology
INPUTS ACTIVITIES
Demand
■■ Train and supervise community groups to promote 
facility delivery and birth preparedness (Village 
Health Teams [VHTs], Safe Motherhood Action 
Groups [SMAGs])
■■ Identify and engage community influencers in safe 
motherhood (Change Champions)
■■ Provide basic newborn/birth supplies to incentivize 
facility delivery (Mama Kits, Mama Packs)
■■ Run mass media campaigns on safe motherhood, 
engage community drama groups
Access
■■ Build new operating theatres with CEmONC capacity
■■ Expand/refurbish maternity wards, labs, pharmacies
■■ Renovate mothers’ shelters near hospitals for 
high risk women 
■■ Provide service delivery vouchers and vouchers for 
transport to facilities and referral to higher-level 
private facilities
■■ Buy emergency obstetric and newborn care 
(BEmONC) equipment
■■ Buy ambulances, motorcycles, motorbikes, 
and communication equipment for transportation 
and referrals 
■■ Form district-level transport committees to 
improve referral
■■ Contract with private providers for EmONC
Quality
■■ Hire new doctors, nurses, and midwives in 
collaboration with the Ministry of Health 
■■ Train health workers in EmONC
■■ Train health workers in newborn resuscitation 
(“Helping Babies Breathe”)
■■ Train doctors in surgical obstetric care and 
anesthetic officers in anesthesia
■■ Provide regular supportive supervision to frontline 
health workers to maintain and improve skills in 
obstetrics, newborn care and anesthesia
■■ Provide training and oversight for maternal 
death reviews 
■■ Provide essential medicines to health facilities
System strengthening
■■ Build provincial and district health team 
capacity with SMGL staff
■■ Strengthen supply chains through training on 
procurement and stock management
■■ Build capacity of facility staff to supervise 
community health workers
■■ Provide computer-based medical records 
(SmartCare)
■■ Strengthen pharmacy, lab, and blood supply
■■ Support roll-out of electronic district health 
information system
■■ Train health workers in data collection and 
health information systems
■■ Both countries ■ UGANDA ■ ZAMBIA
EXHIBIT 5 
SMGL logic model
24
Saving Mothers, Giving Life External Evaluation / Interim Report 25
3 SMGL PROGRAM THEORY
population), and delay in receiving appropriate care (through training, mentoring, 
equipment, drugs, referral).24
To create the SMGL logic model, we have categorized the interventions into 
Demand, Access, Quality, and Health Systems Strengthening (DAQS) activities. The 
logic model (Exhibit 5) is based on our review of ongoing and planned activities in 
each country. The logic model summarizes SMGL efforts to increase awareness of 
the need for care, facilitate transportation to reach care, provide equipment, medi￾cines, and training to improve the quality of care, and ensure that the health system 
develops and sustains approaches to maintain accessible high quality services.
While the DAQS domains, the programmatic approach, and the majority of inter￾ventions are consistent across the two countries, several specific activities and 
means of implementation differ. One major difference between the two countries is 
the role of the private sector. Uganda has involved physician and nurse professional 
associations in training and mentorship and has contracted with private transport 
and health care providers. In Zambia, there is little involvement of the private sector, 
reflecting the minor role that private providers play in the overall health system.
While the logic model describes SMGL activities, we sought to understand which 
activities were most visible at the district level. To this end, we measured how 
frequently activities from the logic model were mentioned in all of our district￾level interviews. Exhibit 6 below summarizes this graphically. The word size gives 
an indication of the salience of these activities to implementers in the district. 
In Uganda, vouchers were most frequently mentioned by interview participants. 
Trainings were emphasized in both countries, although more commonly noted 
in Zambia. The prominence of trainings is likely a function of the frequency and 
intensity of trainings as well as the logistic complexity in implementing them. 
Finally, community outreach efforts—Volunteer Health Teams (in Uganda) and 
Safe Motherhood Action Groups (in Zambia)—were also highlighted.
3.3 SMGL theory of change
The theory of change underlying SMGL holds that it is the combination of increased 
information and transport at the community level coupled with availability of better 
services at health facilities that will dramatically expand health system utilization 
and maternal and newborn survival. Central to this thesis is that neither demand 
nor supply is sufficient to transform the possibility of survival—hence the whole￾of-health systems approach. This is an innovative approach in the field of maternal 
and newborn health where many global initiatives tackle either demand-side25-28 or 
supply-side29, 30 interventions. SMGL’s approach is supported by the experiences of 
countries that accomplished large reductions in maternal deaths historically—from 
Sri Lanka to Egypt to China.31-33 The whole-of-health systems approach also echoes 
and in key ways seeks to replicate the successful experience of PEPFAR programs, 
which simultaneously worked at the community level to overcome stigma around 
HIV and built up the health system to provide high quality clinical interventions.
A secondary feature of SMGL is its role as a catalyst for health system strength￾ening. There are two mechanisms by which SMGL can strengthen health systems. 
First, in both countries SMGL provides concrete support for core health system 
functions: e.g. improving management skills, health information functions, phar￾macy and other supply chains, and blood banks. These activities can benefit not 
only mothers and newborns, but other users as well. Second, by rapidly improving 
the quality of services and with it, the number of facility deliveries, resuscitations 
for asphyxiated newborns, and averted deaths, the initiative demonstrates what is 
possible to achieve in the existing health system, raising the enthusiasm of provid￾ers and managers and the confidence of women and communities. Ideally, these 
efforts can infuse other health services.
“The Ministry of 
Health sees SMGL
as a learning 
opportunity for the 
rest of the system—
performance￾based financing, 
ambulance 
management, and 
better health data.” 
— Uganda MoH Official
26 MARCH 2013
EXHIBIT 6 
Frequency of SMGL activities 
mentioned at the district level
ZAMBIA
UGANDA
Saving Mothers, Giving Life External Evaluation / Interim Report 27
4.1 Partners
The implementation of SMGL is a collaborative effort led by USAID, CDC, and 
the Ministries of Health in Uganda and Zambia. The newly formed Ministry of 
Community Development, Mother and Child Health (MCDMCH) in Zambia is 
another key SMGL stakeholder. As described below, CDC and USAID worked 
with reproductive and child health leaders at the Ministries of Health to design 
SMGL at the country level and select intervention districts. Ministry of Health 
officials were invited to coordinate SMGL planning and implementation meetings, 
although the frequency of the meetings and numerous competing priorities 
limited senior officials’ ability to participate.
In both Uganda and Zambia, CDC and USAID played a central role in contracting 
implementing partners for SMGL activities, coordinating the work of the partners, 
and hosting frequent partner meetings to share progress. CDC, together with the 
countries’ statistical offices and in-country research institutions, is also charged 
with leading the monitoring and evaluation activities including baseline facility 
surveys and maternal mortality surveillance. In the first year, other global SMGL 
partners were in the planning and early implementation stages of their activities. 
Following the PEPFAR model, the majority of SMGL funding—which came from 
the US Government—was awarded to non-governmental organizations who 
had existing relationships with USG agencies and most of whom were already 
working in the eight SMGL districts. These “implementing partners”§
 were largely 
responsible for day-to-day program activities in Phase 1, including: 1) Designing 
interventions, recruitment, training, supervision, and mentoring of health workers 
and community advocacy groups; 2) Outreach, advocacy, and communications; 
3) Procuring medicines, equipment, and supplies; 4) Refurbishing and renovating 
health facilities; and 5) Collecting data. As additional global partners joined the 
initiative, new implementing partners have been added. For example, in Uganda, 
Merck for Mothers began work with local NGOs to engage private sector partners, 
and Every Mother Counts supported transportation vouchers.
In Uganda, a single organization—Baylor-Uganda—is the lead implementing 
partner in three of the four SMGL districts where they are responsible for the 
majority of program activities, while IDI is the main implementing partner in the 
fourth district. Some organizations work across multiple districts. For example, 
Marie Stopes Uganda manages distribution of Healthy Baby vouchers, the Health 
Care Improvement Project (URC) supports quality improvement activities, and 
SURE provides technical oversight and training for supply chain management. 
In Zambia, CIDRZ is the lead implementing partner in two of the four SMGL 
districts, while ZCAHRD and MCHIP are lead partners in the remaining two 
§ NB In this report we use the term “implementing partners” to refer to non-governmental 
implementers. Ministry of Health implementers will be separately identified.
Ambulance in Kamwenge District, Uganda
4
Implementation partners and process
Ministry of Health
■■ Kabarole District Health Office
MAIN IMPLEMENTING PARTNER
■■ Baylor-Uganda
OTHER IMPLEMENTING PARTNERS
■■ STRIDES and SURE (Management 
Sciences for Health) (MSH)
■■ Association of Obstetricians and 
Gynaecologists of Uganda (AOGU)
■■ Uganda Pediatric Association 
(UPA)
■■ Uganda Society of Anaesthesia 
(USoA)
■■ Marie Stopes Uganda (MSU)
■■ Peace Corps
■■ Uganda Health Marketing 
Group (UHMG)
■■ Health Care Improvement 
Project (HCI) (University 
Research Co., LLC)
KABAROLE
Ministry of Health
■■ Kamwenge District Health Office
MAIN IMPLEMENTING PARTNER
■■ Baylor-Uganda
OTHER IMPLEMENTING PARTNERS
■■ STRIDES and SURE (Management 
Sciences for Health) (MSH)
■■ Association of Obstetricians and 
Gynaecologists of Uganda (AOGU)
■■ Uganda Pediatric Association 
(UPA)
■■ Uganda Society of Anaesthesia 
(USoA)
■■ Marie Stopes Uganda (MSU)
■■ Peace Corps
■■ Uganda Health Marketing 
Group (UHMG)
■■ Health Care Improvement 
Project (HCI) (University 
Research Co., LLC)
KAMWENGE
Ministry of Health
■■ Kyenjojo District Health Office
MAIN IMPLEMENTING PARTNER
■■ Baylor-Uganda
OTHER IMPLEMENTING PARTNERS
■■ STRIDES and SURE (Management 
Sciences for Health) (MSH)
■■ Association of Obstetricians and 
Gynaecologists of Uganda (AOGU)
■■ Uganda Pediatric Association 
(UPA)
■■ Uganda Society of Anaesthesia 
(USoA)
■■ Marie Stopes Uganda (MSU)
■■ Peace Corps
■■ Uganda Health Marketing 
Group (UHMG)
■■ Health Care Improvement 
Project (HCI) (University 
Research Co., LLC)
KYENJOJO
Ministry of Health
■■ Kibaale District Health Office
MAIN IMPLEMENTING PARTNER
■■ Infectious Disease Institute (IDI)
OTHER IMPLEMENTING PARTNERS
■■ Association of Obstetricians and 
Gynaecologists of Uganda (AOGU)
■■ Uganda Pediatric Association 
(UPA)
■■ Uganda Society of Anaesthesia 
(USoA)
■■ Marie Stopes Uganda (MSU)
■■ Peace Corps
■■ Uganda Health Marketing 
Group (UHMG)
■■ Health Care Improvement 
Project (HCI) (University 
Research Co., LLC)
KIBAALE
■■ Ugandan Ministry of Health (MoH)
■■ National Medical Stores (NMS)
■■ National Blood Bank
■■ Association of Obstetricians and 
Gynaecologists of Uganda (AOGU)
■■ Uganda Pediatric Association 
(UPA)
■■ Uganda Society of Anaesthesia 
(USoA)
■■ Uganda Bureau of Statistics 
(UBOS)
Uganda
■■ United States Government
■■ U.S. Agency for International Development (USAID)
■■ CENTERS FOR DISEASE CONTROL AND PREVENTION (CDC)
■■ U.S. Peace Corps
■■ U.S. Department of State
■■ U.S. Department of Defense
■■ U.S. Department of Health and Human Services
■■ OGAC (OFFICE OF THE GLOBAL AIDS COORDINATOR)
■■ National Institutes of health (NIH)
■■ Government of Norway
■■ Merck for Mothers
■■ Every Mother Counts
■■ American College of Obstetricians 
and Gynecologists (ACOG)
GLOBAL
■■ Zambian Ministry of Health (MoH)
■■ Zambian National Blood 
Transfusion Services (ZNBTS)
■■ Ministry of Community 
Development, Mother and Child 
Health (MCDMCH)
■■ Central Statistics Office (CSO)
■■ Zambian Defence Force (ZDF)
■■ University of Zambia (UNZA)
Zambia
Ministry of Health
■■ Kalomo District Health Office
MAIN IMPLEMENTING PARTNER
■■ Zambia Center for Applied Health 
Research and Development 
(ZCAHRD) (Boston University)
OTHER IMPLEMENTING PARTNERS
■■ Zambia Integrated Systems 
Strengthening Program (ZISSP)
(Abt Associates)
■■ John Snow, Inc. (JSI)
■■ Elizabeth Glaser Pediatric AIDS 
Foundation (EGPAF)
■■ Communications Support for 
Health (CSH) (Chemonics)
■■ Maternal and Child Health 
Integrated Program (MCHIP)
(Jhpiego)
■■ Society for Family Health 
(SFH) (Population Services 
International)
KALOMO
Ministry of Health
■■ Lundazi District Health Office
MAIN IMPLEMENTING PARTNER
■■ Centre for Infectious Disease 
Research in Zambia (CIDRZ)
OTHER IMPLEMENTING PARTNERS
■■ Zambia Integrated Systems 
Strengthening Program (ZISSP)
(Abt Associates)
■■ John Snow, Inc. (JSI)
■■ Elizabeth Glaser Pediatric AIDS 
Foundation (EGPAF)
■■ Communications Support for 
Health (CSH) (Chemonics)
■■ Maternal and Child Health 
Integrated Program (MCHIP)
(Jhpiego)
■■ Society for Family Health 
(SFH) (Population Services 
International)
SUPPORTING ORGANIZATION
■■ Thandizani
LUNDAZI
Ministry of Health
■■ Mansa District Health Office
MAIN IMPLEMENTING PARTNER
■■ Maternal and Child Health 
Integrated Program (MCHIP)
(Jhpiego)
OTHER IMPLEMENTING PARTNERS
■■ Zambia Integrated Systems 
Strengthening Program (ZISSP)
(Abt Associates)
■■ John Snow, Inc. (JSI)
■■ Elizabeth Glaser Pediatric AIDS 
Foundation (EGPAF)
■■ Communications Support for 
Health (CSH) (Chemonics)
■■ Society for Family Health 
(SFH) (Population Services 
International)
■■ Zambia Prevention, Care & Treat￾ment Program (ZPCT II) (FHI360)
SUPPORTING ORGANIZATION
■■ Groups Focused Consultants
MANSA
Ministry of Health
■■ Nyimba District Health Office
MAIN IMPLEMENTING PARTNER
■■ Centre for Infectious Disease 
Research in Zambia (CIDRZ)
OTHER IMPLEMENTING PARTNERS
■■ Zambia Integrated Systems 
Strengthening Program (ZISSP)
(Abt Associates)
■■ John Snow, Inc. (JSI)
■■ Elizabeth Glaser Pediatric 
AIDS Foundation (EGPAF)
■■ Communications Support for 
Health (CSH) (Chemonics)
■■ Maternal and Child Health 
Integrated Program (MCHIP)
(Jhpiego)
■■ Society for Family Health 
(SFH) (Population Services 
International)
NYIMBA
EXHIBIT 7 
SMGL
Partner 
Network
Saving Mothers, Giving Life External Evaluation / Interim Report 29
districts. In addition to the lead implementing partners, the following partners 
provide SMGL services in all four districts: ZISSP (SMGL coordinators, EmONC 
and SMAG trainings), MCHIP (Helping Babies Breathe trainings), EGPAF (comput￾ers, supporting solar panels, SmartCare and rapid syphilis test trainings), CSH 
(birth plans and other community outreach materials), and JSI (essential medi￾cines and logistics improvement trainings).
Since SMGL activities are focused at the district level, district medical or 
health officers and their teams, who oversee all district health activities, are the 
pillars of SMGL program implementation. The eight district health and district 
medical officers (DHOs/DMOs) oversee and approve implementation activities. 
To assist them in coordinating such a complex program, USG agencies assigned 
an SMGL coordinator funded by the initiative, to manage the day-to-day imple￾mentation of activities as well as to provide a link between DHOs/DMOs and 
implementing partners.
At the facility level, SMGL services are primarily provided by public sector health 
workers. In Uganda, nearly 200 additional health workers and health system 
managers were hired to augment the workforce. In addition to the public sector 
workforce, private service providers in Uganda have been contracted to address 
transport and delivery care. In contrast, the private health sector has not played a 
role in Zambia, where private providers have a smaller presence in rural areas.
4.2 Planning and implementation
SMGL was conceptualized during a visit to Zambia by former Secretary of 
State, Hillary Rodham Clinton, in June 2011. Initial planning began at the US 
Department of State, in partnership with USAID and CDC. Planning in Zambia 
began shortly thereafter with the creation of an SMGL technical task force and 
steering committee. Frequent meetings between CDC, USAID, and the MoH 
shaped the Zambian vision for SMGL, drawing from Zambia’s National Maternal 
and Child Health Road Map, as well as the Zambian adaptation of the Campaign to 
Accelerate the Reduction of Maternal Mortality in Africa (CARMMA). During these 
initial planning meetings, the details of the SMGL logic model were expanded and 
neonatal health was included as a program focus. By October of 2011, US imple￾menting partners had been identified and contracts were awarded. Implementing 
partners conducted planning activities and facility assessments for the remain￾der of 2011; implementation of SMGL activities began in January 2012.
In Uganda, planning began in October 2011 as CDC, USAID, and the Ugandan 
MoH met to shape the program. Uganda’s national Road Map provided a key start￾ing point; SMGL was envisioned as the operationalization of the main components 
of the strategy. In December 2011, SMGL leadership convened a large meeting of 
national and district MoH officials and implementing partners, including private 
and faith-based facility managers, in the regional capital of Fort Portal. 
Contracts and initial work plans for SMGL Uganda were finalized in December 
2011. Preparatory work began in January 2012, including staff recruitment and 
the development of baseline assessment tools. Baseline facility assessments 
and the training of existing staff were completed by February 2012. In April 2012, 
SMGL Uganda implementation began in earnest with new staff reporting for duty, 
ambulances being deployed to the districts, and facility renovations underway. 
Exhibit 9 is a planning and implementation calendar, noting SMGL meetings 
and delegation visits in both countries. This calendar highlights the great inten￾sity of the initiative for funders and implementers, who are simultaneously 
managing their regular portfolio of activities.
For district health teams, SMGL was added to numerous existing health 
programs in the districts, managed by a variety of local and global, public and 
4 IMPLEMENTATION PARTNERS AND PROCESS
A typical mothers’ shelter in 
Mansa District, Zambia
A renovated mothers’ shelter in 
Mansa District, Zambia
PHOTOS: ZPCT II
30 MARCH 2013
private actors. To illustrate this point, Exhibit 8 shows the crowded health partner 
landscape that SMGL joined in each district. The Exhibit shows the organizations 
engaged in health activities in the SMGL districts, in addition to core Ministry of 
Health programs. As of December 2012, each of the eight SMGL districts had 
between 13 and 28 organizations.
4.3 Use of PEPFAR platform
The intensity displayed in the implementation calendar (Exhibit 9) alludes to 
the highly competent teams involved in the planning and implementation of the 
program. Another facilitator of SMGL’s quick launch was the ability to leverage 
PEPFAR partnerships. As PEPFAR focus countries, Uganda and Zambia have 
longstanding and substantially funded programs supporting HIV prevention, care, 
and treatment. In order to expedite program start-up, the USG largely built on 
existing partnerships, grants, and contracts rather than initiating new funding 
mechanisms or competitive bidding for SMGL activities.
CDC, for example, added supplemental funding to existing cooperative agree￾ments, expanding the remit of selected PEPFAR implementing partners to include 
SMGL activities. USAID, in contrast, shifted funding and adjusted scopes of work 
Kabarole
Kibaale
Kyenjojo
Kamwenge
AOGU
NESH
UCP
USOA
AOGU
USOA
AOGU
USOA
HIPS
EMESCO
SDS
AOGU
USOA
SPEAR
EXHIBIT 8 
Organizations active in health 
initiatives in SMGL districts
Uganda
4 IMPLEMENTATION PARTNERS AND PROCESS
Saving Mothers, Giving Life External Evaluation / Interim Report 31
Kalomo
Mansa
Lundazi
Nyimba
MUMUNI PROJECT
LUFAID
GROUPS FOCUSED
CONSULTANTS
ANGLICAN DIOCESE OF 
MANSA
CATHOLIC 
DIOCESE OF 
MANSA
ACTION FOR SOCIAL 
DEVELOPMENT
MANSA BUSINESS 
ASSOCIATION
to permit current implementing partners to focus on SMGL activities and districts. 
This approach, described by one USG officer as “bureaucratically nimble,” enabled 
very rapid start-up of program activities by implementing partners, most of whom 
were already at work in the focus districts. In this way, SMGL leveraged the lessons, 
resources, partners, and programs developed by PEPFAR.
4.4 Funding
Although a costing analysis of SMGL is outside the purview of this evaluation 
(a separate costing study is being conducted by USAID), we will present an estimate 
of the funds expended thus far on SMGL to highlight the magnitude of health system 
investments and, in turn, the scope and ambition of the initiative.
The funding for Phase 1 of SMGL came almost exclusively from the US 
Government. As of December 2012, the USG had allocated approximately $37 million 
USD—$31.5 million for implementation and an additional $5.5 million for monitoring 
and evaluation activities. USAID attributed $6 million in Uganda and $12 million in 
Zambia. These were not new funds; rather they were monies that had been budgeted 
for activities related to PEPFAR, family planning, maternal and child health, nutrition, 
and malaria that were now concentrated on SMGL districts.
Zambia
4 IMPLEMENTATION PARTNERS AND PROCESS
32 MARCH 2013
CDC contributed approximately $9.5 million USD in new money to each country: 
$6.5 to 7 million for implementation activities and $2.5 to 3 million for monitoring 
and evaluation. The US Department of Defense committed a smaller amount for 
facility renovations in Zambia and the Peace Corps contributed the services of 
Peace Corps Response Volunteers in all four focus districts in Zambia and two 
districts in Uganda.
In terms of other SMGL global partners, Merck for Mothers is in the early stages 
of funding activities in Zambia and Uganda. In Zambia, Merck for Mothers has 
issued a request for proposals for two to three grants of $200,000 to $250,000 
each for planning projects to determine the feasibility and design of a maternity 
waiting home model that can ensure high quality services in a financially self-sus￾taining manner. Merck for Mothers plans to award an additional $3 million for one 
to two projects in Phase 2 to conduct implementation research and test the model. 
In Uganda, Merck for Mothers has commited $9 million in the initial four SMGL 
districts and several other districts in year 1 and approximately 30 districts over 
the three-year course of the project to strengthen the private health sector. Merck 
for Mothers is working with a consortium led by PACE, a local NGO, to address 
several aspects of the Demand-Access-Quality-System Strengthening framework: 
bolster the private provider network (e.g. providing training and business loans to 
private clinics), improve women’s access to supplies and information (e.g. work 
with drug shops, Mama Ambassadors, and care providers to market and distribute 
Mama Kits), make care more affordable (e.g. community based health insurance 
and savings schemes), and improve emergency transport (e.g. training taxi drivers 
4 IMPLEMENTATION PARTNERS AND PROCESS
COUNTRY LEVEL ACTIVITY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
UGANDA
Central
CDC, USAID, 
& implementing 
partners
Partners meetings • • • • • • • • • • • • • • • • • • • • • • • • • •
Hosting visiting delegations • • • • •
District
Health offices 
& implementing 
partners
Partners meetings • • • • • • • • • • • • • • • • • • • • •
ZAMBIA
Central
CDC, USAID, 
& implementing 
partners
Partners meetings • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •
Hosting visiting delegations • • • • • • •
District
Health offices 
& implementing 
partners
Partners meetings • • • • • • • • • • • • • • • • • • • • • • •
2011 2012
Planning month Implementation month Indicates a single event 
(i.e. meeting or visiting 
delegation)
EXHIBIT 9 
Implementation calendar
Saving Mothers, Giving Life External Evaluation / Interim Report 33
EXHIBIT 10 
SMGL budget 
in the context of 
national health 
expenditures
who transport women in labor). Recruitment of private clinics into the private 
franchise network, and clinical training activities, began in early 2013.
Every Mother Counts provided $41,000 USD to support transportation vouchers 
in Uganda. The Government of Norway will provide funding in Phase 2, as origi￾nally planned, and the American Congress of Obstetricians and Gynecologists is 
in the process of finalizing its support for activities in the two countries.
Both the Government of Uganda and the Government of Zambia made addi￾tional investments that directly or indirectly support the goals of SMGL. For 
example, the Government of Uganda plans to increase physician salaries and the 
Government of Zambia hired more than 2,000 new health workers nationwide in 
2012, of which 75% were nurses.
The US Government funding for SMGL translates into an approximate 
investment of $7.4 per person in the Ugandan focus districts and $20.8 per 
person in the Zambian focus districts.¶ To put these figures into context, the 
total health spending in Uganda is $46.7 per person, of which $10.0 comes 
from the Ugandan government. In Zambia, the total per capita health spending 
is $72.9, of which $43.9 is government-funded.10 As Exhibit 10 shows, SMGL 
funding represents 16% and 29% of all health spending—and 74% and 47% of 
government spending—in Uganda and Zambia, respectively. This is a substantial 
investment in maternal and newborn survival compared to countries’ current 
resource envelopes.
¶ Figures obtained by dividing funds for implementation (approximately $13 million in Uganda and 
$18.5 million in Zambia) by the combined population of the four SMGL districts per country. Population 
figures were taken from the SMGL operational plans.
Uganda Zambia
$0
$10
$20
$30
$40
$50
$60
$70
$46.7
$10.0
$7.4
$72.9
$43.9
$20.8
Total health spending per capita
Government health spending per capita
SMGL spending per capita
“Quite a lot of 
investment has 
gone into the 
focus districts 
in terms of extra 
dollars reaching 
the ground... and 
the partners who 
are implementing 
the program have 
received extra 
funds to increase 
their capacity.”
— Zambia MoH Official
34 MARCH 2013
5.1 Methods
Our analysis of the “dose delivered” assessed the extent to which SMGL activities 
had been implemented in the intervention districts as of December 1, 2012—the 
mid-point of Phase 1 of the program. The data in Exhibit 11 thus provide a district￾level perspective on the extent of implementation of SMGL at six months. Our 
analysis excludes any planned or future activities. In interpreting the figures, 
it is important to note that some activities were sequenced for later implemen￾tation, and that in some cases, the lack of activities may reflect lower need for 
those inputs in a particular district rather than slow implementation. An additional 
caveat is that there may be some peripheral or pilot activities that were not 
included, as we sought to assess only the major activities reflected in the country 
SMGL operational plans.
Our dose delivered instrument consisted of 202 indicator variables representing 
the 28 core SMGL activities from the logic model. To ensure that we captured 
all relevant activities, we reviewed the SMGL operational plans for each country, 
and data from interviews with USG, MoH, and implementing partners in Zambia 
and Uganda.
Once all data were collected and vetted with in-country partners, we then 
selected 33 (Uganda) and 32 (Zambia) tracer indicators that captured implemen￾tation of the core SMGL activities. Tracer indicators provide a useful summary of 
overall SMGL implementation from the district perspective. These indicators were 
selected based on three criteria: 1) The extent to which the indicator variable was 
representative of the larger activity it was nested under; 2) The availability of data 
for the indicator across all SMGL districts; and 3) The reliability of the available 
data for the indicator. Specific details on the methods for collecting, vetting, and 
analyzing the dose delivered data can be found in Appendix B.
5.2 Results
Uganda
There was little variability in the extent to which activities had been implemented 
across districts in Uganda in the first six months of the program. Some variation 
in activities implemented was noted in Kibaale, which has a different implement￾ing partner (IDI) than the other three districts (Baylor-Uganda).
Upwards of 600 Village Health Team (VHT) members were trained in each 
district, with Kibaale, the most densely populated district, training over 1,900. 
This was the major demand-building activity in the four Uganda districts and 
it was widely reported to have facilitated community-level data collection and 
demand generation for maternal health services.
Activities to increase access to maternal and newborn care included facility 
upgrades to CEmONC capacity in each district (11 facilities), with all but Kibaale 
5
Mid-point results
“Kit boxes” for obstetric emergencies in 
Kalomo District, Zambia
PHOTO: Dr. Davidson Hamer, ZCAHRD
Saving Mothers, Giving Life External Evaluation / Interim Report 35
receiving facility upgrades to BEmONC capacity (11 facilities). The four districts 
had a total of 17 refurbished maternity wards (with refurbishments defined as 
any upgrades that fell short of the construction of a new building). Kibaale was 
the only district that reported completed renovations of mothers’ shelters (4). 
All three of the Baylor districts implemented a voucher system and private 
provider contracts for EmONC services. Additionally, Marie Stopes Uganda 
distributed transport and delivery vouchers in all four districts. Transport was 
another major activity aimed at increasing access. A total of 30 vehicles were 
distributed across the four districts with each district receiving at least one 
vehicle ambulance and at least two motorcycle ambulances. Furthermore, each 
district established a transport committee responsible for coordinating vehicle 
use and maintenance.
Many activities aimed at increasing the quality of maternal and newborn health 
care were implemented, including the hiring and training of health providers. 
A total of 20 additional doctors were hired across SMGL districts, while 20 addi￾tional nurses were hired in Kyenjojo. The largest stimulus to human resources, 
however, was the addition of 94 midwives across the four districts. One of the 
most intensive activities was the training of health workers in EmONC and 
newborn resuscitation. Doctors and nurses also received trainings in surgical 
obstetric care and anesthesia, respectively, and substantial numbers of frontline 
health workers in all districts received regular supportive supervision. A total of 
250 health workers received SMGL-sponsored clinical trainings. Maternal and 
perinatal death review (MPDR) trainings were also conducted in all four districts.
Several SMGL activities were aimed at strengthening health systems within the 
focus districts. Health workers were trained in data collection and health infor￾mation systems. Kibaale had the highest number of health workers trained in 
supply chain strengthening (30). Baylor and IDI each hired an SMGL coordinator 
to manage project logistics in the districts.
Zambia
In Zambia, there was considerable variation in implementation status between 
different activities. Demand creation activities, especially Safe Motherhood Action 
Group (SMAG) trainings and birth plan dissemination, were implemented in all 
four districts. Lundazi, the district with the largest population, trained six SMAGs 
(120 individuals), whereas Nyimba, with approximately one-third the population, 
trained twelve (210 individuals). Exhibit 12 is a video clip showing a training of 
a SMAG in Zambia. District-level respondents reported that neither the planned 
documentary about safe motherhood nor a mass media campaign had been 
completed, although a radio campaign was noted to be in production. All districts 
engaged community influencers as “Change Champions” for safe motherhood, 
but only Lundazi and Kalomo reported the use of Mama Packs (i.e. gift packages 
containing a combination of a traditional fabric, diapers, and soap) to create 
demand. Several of these activities, such as SMAGs and Mama Packs, predate 
SMGL and were mostly scaled-up and intensified under SMGL.
Activities to improve access to maternal and newborn health services included 
facility upgrades to BEmONC capacity in all districts but Mansa. It is possible 
that most facilities in Mansa, a provincial capital, already had sufficient BEmONC 
capacity. No facilities were upgraded to CEmONC capacity in any of the districts. 
Mansa and Kalomo were the only districts to see maternity wards refurbished, 
while Nyimba was the only district to report a refurbished pharmacy. Renovations 
were in progress for five mothers’ shelters in Kalomo and twelve in Mansa. 
Districts reported substantial hurdles in procuring emergency and utility vehi￾cles. Nyimba still had no SMGL vehicle as of December 1, 2012, while Lundazi 
5 MID-POINT RESULTS
Operating theatre before renovation, 
Kyenjojo District, Uganda
Operating theatre after renovation, 
Kyenjojo District, Uganda
Train community groups to promote delivery/birth preparedness (VHTs/SMAGs) Individuals trained as VHT members 692 626 1,912 750
Provide basic newborn/birth supplies to pregnant women (Mama Kits) Mama Kits distributed to pregnant women ** ** ** **
Run mass media campaigns on radio, engage community drama groups Radio spots broadcast 27,391
Drama skits conducted 500
Increase facility EmONC capacity Health facilities upgraded to BEmONC capacity 6 2 3
Health facilities upgraded to CEmONC capacity 3 2 5 1
Expand/refurbish maternity wards, labs, pharmacies Maternity wards with infrastructure improvements 10 2 4 1
Expanded/refurbished pharmacies 0 0 0 0
Renovate mothers’ shelters near hospitals for high risk women Renovated mothers' shelters 0 0 4 0
Buy ambulances, motorcycles, and motorbikes for transportation 
and referrals
Vehicle ambulances provided to district 2 2 2 1
Motorcycle ambulances provided to district 3 3 8 2
Provide vouchers for delivery, EmONC, transport, postnatal care 
in private facilities
Vouchers distributed by Baylor to women 17,192 5,340 7,360
Baylor vouchers redeemed by women 8,649 2,207 4,128
Healthy baby vouchers distributed to women 16,631
Private facilities enrolled in voucher program 42
Form district-level transport committees to improve referral Transport committees established in district 1 1 2 1
Buy emergency obstetric and newborn care (EmONC) equipment EmONC equipment procured ** ** ** **
Contract with private providers for EmONC Private providers contracted to provide EmONC 7 6 6
Hire new doctors, nurses, and midwives in collaboration with the 
Ministry of Health
New doctors hired in district 5 3 9 3
New nurses hired in district 0 0 0 20
New midwives hired in district 19 12 48 15
Train health workers in EmONC Health workers trained in EmONC 90 44 29 60
Train health workers in newborn resuscitation Health workers trained in newborn resuscitation 90 44 29 60
Supervision of frontline workers to improve obstetrics and newborn care skills Health workers who received supportive supervision 120 110 41
Provide training and oversight for maternal death reviews Health workers trained in maternal death reviews 12 8 29 10
Provide essential medicines Provide essential medicines to health facilities ** ** ** **
Train doctors in surgical obstetric care and nurses in anesthesia Doctors trained in surgical obstetric care 7 3 ** 5
Nurses trained in anesthesia 7 2 3
Strengthen supply chains through training on procurement and stock mgmt Workers trained in procurement and stock mgmt 7 0 30 14
Build provincial and district health team capacity with SMGL staff SMGL coordinators in district* 1 1 1 1
Strengthen pharmacy, lab, and blood supply Blood banks 1 0 5 2
Equipment provided to strengthen blood supply ** **
Train health workers in data collection and health information systems (HIS) Workers trained in data collection and HIS 213 206 11*** 214
ACTIVITY TRACER INDICATORS FOR ACTIVITY
Uganda
DEMAND
ACCESS
QUALITY
STRENGTHENING
409,400 324,400 646,500 369,700
KABAROLE KAMWENGE KIBAALE KYENJOJO
POPULATION
DISTRICTS
EXHIBIT 11 “Dose delivered:” SMGL activities 
implemented in districts as of December 1, 2012
Data not available
* A single SMGL coordinator oversees the Kabarole, Kamwenge, and Kyenjojo districts
** Activity under way, district-level data not available
*** This figure only refers to the record clerks and assistants trained in data collection—
it does not include the number of nurses and midwives trained in data collection
ACTIVITY TRACER INDICATORS FOR ACTIVITY
Zambia
Train community groups to promote facility delivery and birth 
preparedness (VHTs/SMAGs)
Safe Motherhood Action Groups trained 13 6 24 12
Individuals trained as SMAG members 220 120 240 210
Improve birth preparedness at home through VHTs/SMAGs Birth plans distributed ** ** 56,000 16,000
Produce documentary about safe motherhood using traditional leaders Documentary airings 0 0 0 0
Run mass media campaigns on radio, engage community drama groups Campaigns conducted 0 0 0 0
Identify and engage community influencers in safe motherhood Change Champions trained 11 35 13 21
Provide basic newborn/birth supplies to pregnant women (Mama Packs) Mama Packs distributed to pregnant women 6,000 350 0 0
Increase facility EmONC capacity Health facilities upgraded to BEmONC capacity 28 34 0 7
Health facilities upgraded to CEmONC capacity 0 0 0 0
Expand/refurbish maternity wards, labs, pharmacies Expanded/refurbished maternity wards 5 0 12 0
Expanded/reburbished pharmacies 0 0 0 1
Renovate mothers’ shelters near hospitals for high risk women Mothers' shelters under renovation 5 0 12 0
Buy emergency obstetric and newborn care (EmONC) equipment EmONC equipment procured ** ** ** **
Buy ambulances, motorcycles, and motorbikes for 
transportation and referrals
Vehicle ambulances provided to district 1 1 3 0
Motorcycle ambulances provided to district 0 0 12 0
Hire new doctors, nurses, and midwives in collaboration with the 
Ministry of Health
New doctors hired 0 2 0 0
New nurses hired 0 7 0 0
New midwives hired 12 0 0 0
Train health workers in EmONC Health workers trained in EmONC 38 40 80 21
Train health workers in newborn resuscitation Health workers trained in newborn resuscitation 22 24 80 20
Train health workers in rapid syphilis testing Health workers trained in rapid syphilis testing 36 39 43 22
Supervision provided to maintain/improve skills in obstetrics/newborn care Health workers who received supportive supervision 24 46 20
Provide essential medicines Provide essential medicines to health facilities ** ** ** **
Provide training and oversight for maternal death reviews Health workers trained in maternal death reviews 0 24 9
Strengthen supply chains through training on procurement and stock mgmt Workers trained in procurement and stock mgmt 35 56 25 30
Build provincial and district health team capacity with SMGL staff SMGL coordinators in district 1 1 1 1
Build capacity of facility staff to supervise community health workers Facility staff trained in supervising health workers 13 0 3 26
Provide computer-based medical records (SmartCare) Computers provided to health facilities 35 39 30 17
Staff trained in computer literacy and SmartCare 73 38 78 40
Strengthen pharmacy, lab, and blood supply Blood banks 1 0 1 1
Equipment provided to strengthen blood supply ** ** ** **
Train health workers in data collection and health information systems (HIS) Workers trained in data collection and HIS 95 48 80 30
DEMAND
ACCESS
QUALITY
STRENGTHENING
254,211 314,281 217,603 101,616
KALOMO LUNDAZI MANSA NYIMBA
POPULATION
DISTRICTS
Data not available
** Activity under way, district-level data not available
38 MARCH 2013
had one vehicle ambulance of the three they were expecting. Kalomo received one 
vehicle ambulance and Mansa received three in addition to twelve motorcycle ambu￾lances. The slow pace of vehicle procurement was a major source of frustration 
among district health officials and implementing partners in all districts.
In contrast to facility refurbishment and transport, implementing partners in 
Zambia made rapid and substantial progress on activities related to quality improve￾ment. A total of 179 health workers in the four districts were trained in EmONC and 
newborn resuscitation (“Helping Babies Breathe”), with the goal to have at least one 
newly trained skilled provider per facility. Following the trainings, implementing 
partners provided onsite supportive supervision. With the exception of Lundazi, 
all districts reported trainings in maternal death reviews (MDRs). Compared to 
Uganda, few facility-level providers were hired by SMGL partners, an issue that 
was mentioned frequently by district respondents. Kalomo, however, received 
12 retired midwives who were affectionately referred to as “the twelve disciples” 
in the district. Lundazi was the only other district to report additional human 
resources through SMGL, consisting of two doctors and seven nurses.
Implementing partners carried out a number of health systems strengthening
activities in the four districts. Each district has a dedicated SMGL district coordinator 
to oversee activity coordination. Between 25 and 56 health workers in each district 
attended an extensive five-day training on logistics and supply chain management 
through the Essential Medicines and Logistics Improvement Program (EMLIP) 
provided by John Snow, Inc. (JSI). Providers in each district were also trained in 
computer literacy, SmartCare, and rapid syphilis testing by the Elizabeth Glaser 
Pediatric AIDS Foundation (EGPAF), which also provided computers and supporting 
solar panels to every SMGL facility. Additionally, some providers received trainings 
in health worker supervision and health information systems.
Synthesis
Trainings and mentorship were the two most strongly implemented activities across 
the eight SMGL districts. VHT and SMAG trainings, EmONC and newborn resusci￾tation trainings, and onsite mentoring were also implemented extensively in both 
countries. This was likely due to the ability of implementing partners to rapidly 
recruit qualified trainers and mentors and may have been facilitated by partners’ 
existing relationships with district health teams. Although it is not reflected in the 
table due to a lack of specific data, Uganda, like Zambia, provided facility delivery 
incentives—Mama Kits containing some basic newborn supplies (e.g. a baby blan￾ket)—in some districts.
Some mothers’ shelters were in the process of being upgraded in both countries. 
Although shelters and housing for health workers topped the wish lists of many of 
the district medical officers interviewed, new construction is not permitted under 
USG foreign assistance regulations. Upgrades to pharmacies and laboratories were 
also lacking in the first six months of the project.
Notable differences between activities implemented in Uganda and Zambia were 
the lack of CEmONC upgrades, additional health care workers, and emergency 
transport vehicles in Zambia. Some of these activities may be pursued in the second 
half of Phase 1. The substantial increase in human resources in Uganda helped to 
allay the absenteeism experienced during off-site trainings, a common complaint in 
Zambia, which is discussed in Section 6.4 of this report. Another point of difference 
was that only Uganda used a voucher system and only Zambia reported the use of 
community influencers as Change Champions. While both countries provided train￾ings on supply chain management and logistics, this was implemented with more 
intensity in Zambia, perhaps because it was streamlined across the four districts 
by JSI, one of Zambia’s main procurers of medical supplies.
5 MID-POINT RESULTS
EXHIBIT 12 
Video of SMAG 
training in 
Nyimba District, 
Zambia
SMGL health worker training, 
Nyimba District, Zambia
Saving Mothers, Giving Life External Evaluation / Interim Report 39
6.1 Successes
Perhaps the most notable findings of this evaluation is that although the effect of 
the SMGL on maternal mortality is not yet known there was a unanimous percep￾tion among the over 150 respondents we interviewed that SMGL was already a 
success on many fronts. While we heard a range of thoughtful criticisms of the 
program and many ideas for improvement, no interviewees considered SMGL 
a failure. The respondents interviewed for this interim report were policymak￾ers, funders, and implementers involved in SMGL in Zambia and Uganda. The 
perspectives of communities and health providers will be captured in the next 
wave of our data collection and included in the final report.
In listing key SMGL achievements, participants noted that early data from 
routine health information systems showed increases in institutional deliveries in 
intervention districts. They also reported improved awareness of and demand for 
maternity services from women in communities, and the perception that the level 
and quality of obstetric and newborn services at facilities had improved markedly 
since the start of the program. This perspective was shared by interviewees at 
central and district levels in both countries and by respondents with different 
roles in SMGL from policy to clinical care.
The activities deemed most successful spanned the full range of the Demand￾Access-Quality-System Strengthening framework. These included Village Health 
Teams and Safe Motherhood Action Groups in Uganda and Zambia, respectively, 
and Mama Kits/Packs in both countries (Demand); transportation vouchers and 
hiring of new health workers in Uganda and building new maternity wards and 
operating theaters (Access); new equipment, emergency obstetric and newborn 
resuscitation training and mentoring (Quality); and community data collection 
and improvement in health information systems (System Strengthening). In short, 
those interviewed agreed that in the first six months of operations SMGL had 
achieved notable successes in both the community and the health system. 
Respondents were also asked to comment on the drivers of SMGL’s success 
to date. One of the most frequently cited success factors in both Zambia and 
Uganda was the strong national partnerships—between Ministry of Health, US 
Government, and a range of implementing partner organizations—that formed 
around SMGL in each country. Respondents pointed to the support for the initiative 
from the Ministry of Health, particularly at the outset in selecting interventions 
and SMGL districts. They also noted that despite SMGL funds being “off-budget” 
(i.e. not channeled through national governments) SMGL worked largely through 
the existing health system, rather than pursuing stand-alone solutions.
Many participants noted that although there were points of tension, the 
“unusually collegial” relationship between CDC and USAID in both countries was 
an essential ingredient to the early success of such a large program. An imple￾menting partner respondent in Lusaka commented on the “fabulous interagency 
“SMGL did not 
create parallel 
structures like 
other projects 
in the past 
have: SMGL was 
designed to work 
within the existing 
[health system] 
structure and 
communities.”
— Uganda District Level
Implementing Partner
6
Successes, challenges, lessons learned
40 MARCH 2013
collaboration” on SMGL as a key driver of success. This sentiment was echoed by 
a senior Ministry of Health official in Uganda, who observed that the collaboration 
among USG agencies in SMGL was “the best I’ve ever seen” of any US-funded 
health program. USG respondents in country also pointed to the important role of 
CDC and USAID headquarters—and “USG big-wigs”—in supporting programming. 
Respondents agreed that implementation within two months of the original 
SMGL plan would have been impossible without the presence in both countries 
of organizations with existing relationships and, crucially, contractual arrange￾ments with USG agencies. Most of these relationships were originally developed 
for PEPFAR programs. These implementing partners’ experience delivering 
complex HIV and other programs at district level and their strong relationships 
with Ministry of Health counterparts at central and district levels greatly facili￾tated their ability to deliver interventions quickly. In this way the large PEPFAR 
programs in the two countries served as a launching pad for SMGL. Frequent 
meetings among partners, weekly at the outset, enabled input and participation 
and promoted a sense of common purpose. 
While some noted that district medical officers were sometimes overwhelmed 
with the speed and intensity of SMGL implementation amidst other health prior￾ities, all agreed that the buy-in, commitment, and in many cases the personal 
commitment of district medical officers in intervention districts were essential 
preconditions for success. 
Respondents noted that the large financial commitment by the US Government 
in SMGL captured stakeholders’ attention and elevated maternal health efforts—
at least temporarily—on the agendas of policymakers and implementers. While 
many pointed to the tremendous challenges presented by the compressed time￾line, US government respondents in particular noted that the “sprint” nature of 
Phase 1 of SMGL generated excitement among national actors and their develop￾ment partners that one rarely sees in maternal health or health system initiatives. 
At the same time, respondents pointed out that this level of energy cannot be 
maintained over the long run and that the program must adapt to permit integra￾tion into the health system. Interviewees from the USG and Ministries of Health 
acknowledged, however, that the need to produce results within 12 months 
hindered country ownership. This is discussed in more detail below.
Finally, the emphasis on community involvement from the outset of SMGL was 
cited as a critical driver of the early success of SMGL. Safe Motherhood Action 
Groups (Zambia) and Volunteer Health Teams (Uganda) were credited not only 
with promoting facility births but also with changing community views about the 
inevitability of maternal deaths.
Exhibit 13 below summarizes the SMGL program activities that were the 
most frequently cited successes by showing the proportion of interviewees who 
mentioned each item in response to a question on key successes in the two coun￾tries. Below we briefly discuss these.
Community engagement through VHTs and SMAGs
While Safe Motherhood Action Groups (SMAGs) and Volunteer Health Teams 
(VHTs) existed in Zambia and Uganda before SMGL their role was greatly 
intensified for SMGL. This was done through both additional recruitment and 
training of existing participants. Both groups were also charged with new tasks, 
notably community-level data collection on pregnancies and maternal and 
newborn deaths.
In Uganda, respondents reported that VHTs helped engender “community 
commitment” and “increased confidence in providers.” SMGL “brought hope to a 
lot of people in the community” who were said to have received the project it with 
“[It is] very 
special in my 
experience to see 
organizations 
work this closely 
and this well.”
— USG Zambia Official
“[Before SMGL] 
there were many 
mothers dying 
in silence. At 
least now when 
mothers die, 
people notice and 
they try to learn 
from it. It’s a big 
issue. Now when 
a mother dies 
we know before 
lunch.”
— Uganda District Level
Implementing Partner
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
Saving Mothers, Giving Life External Evaluation / Interim Report 41
0% 10% 20% 30% 40%
Refurbishing operating theaters
Equipment & supplies
Improved data management
Mentoring
Transport
Obstetric & newborn care trainings
Community engagement (VHTs/SMAGs) 32%
15%
25%
6%
15%
17%
43%
30%
11%
25%
20%
13%
8%
“open arms.” Community dialogue meetings facilitated by VHTs were mentioned 
as a particularly powerful means to mobilize community members.
In Zambia, SMAGs and community leaders were credited with improving the 
communities’ acceptance of SMGL and building trust in the health system. Chiefs 
and other community leaders participated in brainstorming sessions for SMGL, 
for example, identifying poor provider attitudes as a key barrier to women deliv￾ering in facilities. SMAG members were praised as the true demand creators, 
with a number of respondents reporting facility deliveries were higher at facili￾ties with trained SMAGs.
Health provider training and mentoring
in obstetric and newborn care
The extensive trainings in emergency obstetric care, neonatal resuscitation 
(“Helping Babies Breathe”), and specialized training in obstetric surgery in 
Uganda were highly praised in both countries. District medical officers and 
providers said that the trainings improved skills and the confidence and morale 
of doctors, nurses, midwives, and clinical officers. In health systems where 
providers are frequently overworked and do not have the skills or tools to do 
their job, we heard that SMGL trainings and mentoring made providers feel 
“invested in instead of policed.” In Zambia, district respondents noted that staff 
could now manage abortion complications, postpartum hemorrhage, and sutur￾ing on the ward rather than in the operating theater. Participants also noted the 
high levels of enthusiasm among clinical staff for the initiative.
Improved availability of transport
The success of transport investments was mixed, with a much more positive 
assessment in Uganda than Zambia. The investment in transport was seen as 
“Health workers 
have been 
empowered [by 
the trainings] 
and have risen to 
the occasion in 
a way that they 
haven’t with many 
projects.”
— USG ZAMBIA OFFICIAL
EXHIBIT 13 
Frequency 
of reported 
successful 
activities by 
country
Uganda
Zambia
42 MARCH 2013
a key contribution of SMGL in Uganda. Specifically, respondents in Uganda felt 
that the provision of emergency vehicles and motorcycles (boda bodas) was 
one of the most important successes of SMGL. A district MoH respondent noted 
that, “boda bodas for mothers is a huge success because it integrates services 
and norms that already exist in our community to encourage maternal health. 
Vehicle ambulances are also a hugely important success and have increased our 
capacity.” An implementing partner in the district observed, “Transport was a big 
problem solved and the mothers have been enjoying coming to the health units 
much more than before.”
In Zambia district health officials expressed frustration that few ambulances 
had been delivered six months into the program.
Improved data management
In both countries MoH respondents and implementing partners cited improve￾ments in health information systems that resulted in better and more timely 
data on mothers and newborns. This was particularly the case in Zambia where 
there was a substantial investment in electronic health cards that linked with 
facility patient databases. Also in Zambia, community key informants trained 
by the Central Statistics Office to conduct verbal autopsies were cited as key to 
improved understanding of deaths in the community.
In Uganda, an implementing partner commented that as a result of SMGL 
efforts data utilization in SMGL districts was higher than he had witnessed in 
other districts. Another benefit observed in Uganda was that data from private 
sector facilities is now better integrated into district information systems.
Provision of equipment and supplies
and facility refurbishment
Facility improvements were another important SMGL success. In Uganda, 
respondents enthusiastically reported that refurbished operating theaters were 
functional for the “first time in years.” The expansion of maternity wards was 
another important success given that lack of privacy was a frequently mentioned 
barrier to women wanting to deliver in a facility.
Although facility improvements in Zambia were at an earlier stage than 
in Uganda, respondents noted that new supplies (e.g. delivery kits, IV tubing, 
mattresses, and blankets), medicines, and equipment (e.g. manual vacuum aspi￾rators, ultrasound machines) helped to improve provider morale, the quality of 
obstetric care, and encouraged women to use facilities.
6.2 Innovations
Reducing maternal mortality has proved exceedingly difficult for low- and 
middle-income countries around the globe. Most regions, with the exception of 
East and Central Asia, will not meet Millennium Development Goal 5—reduction 
of the maternal mortality ratio by 75% between 1990 and 2015.34 It is clear 
therefore that business as usual will not be sufficient to bring about meaningful 
improvements in maternal survival and that there is a need for new ideas on how 
to implement what we know works. In our implementation evaluation of SMGL 
we were particularly interested in identifying interventions, service delivery 
approaches, and organizational strategies that provided new approaches for 
tackling maternal and newborn mortality. To this end, we asked SMGL stakehold￾ers in country what they felt were the most innovative aspects of SMGL.
We identified three themes on innovation: new technologies, improved ways 
of working, and new ideas for community mobilization. It was notable that the 
innovations mentioned were not related to provision of clinical services per se. 
Boda boda in Kyenjojo District, Uganda
Newly refurbished Maternity Ward, 
Kamwenge District, Uganda
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
Saving Mothers, Giving Life External Evaluation / Interim Report 43
Instead, respondents spoke about new ways to link women to services, change 
community perceptions, and organize health services. The results are summa￾rized in Exhibit 14.
Respondents in both countries described the intense interagency collabora￾tion between CDC and USAID as a new way of working on a shared agenda for 
US agencies in global health. The use of SMGL coordinators in the districts to 
support district Ministry of Health structures and maintain communication was 
also cited, as was the use of VHTs and SMAGs to collect community-level data. 
Engaging non-standard actors in the fight for maternal survival was another 
innovation. This included collaboration with private facilities in Uganda to expand 
safe delivery options for women. Churches in Zambia were encouraged to deliver 
safe motherhood messages to congregants on Sundays. Traditional community 
leaders were engaged as Change Champions in several Zambian districts, acting 
to motivate women to use services and to communicate barriers to access to 
medical staff.
Another new way of working was the explicit effort to deliver services to 
women who could not get to the health system. In Kibaale, mobile clinics 
attended to pregnant women living in remote areas who may otherwise have no 
access to services. Finally, mentorship by experienced clinicians onsite was cited 
as an extremely effective innovation that permitted nurses to stay in their facili￾ties and hone their skills in “real-world” settings.
Ways of 
working ■■ Onsite mentoring
■■ SMGL district 
coordinators 
■■ Improved referral 
networks
■■ High-level partner 
coordination
■■ Data collection by 
VHTs & SMAGs
■■ Mobile clinics for 
maternal health 
services
■■ Involvement of 
private facilities
■■ Churches deliver safe 
motherhood messages
■■ Local leaders as Change 
Champions
■■ Phased implementation 
approach
■■ Emergency drills 
for mentoring
■■ Referral & 
feedback pilot
■■ Mothers shelter logs
Community 
mobilization ■■ Safe motherhood 
day/week
■■ Mama Kits/Packs
■■ Transport committee
■■ ANC4 incentive system
■■ Transport & service 
vouchers
■■ Income generating 
activities for VHTs
■■ Male involvement 
promoted at soccer 
■■ Partnership with village 
savings association
■■ Incentives for 
postnatal care
■■ Culturally appropriate 
birth plans
■■ Incentives for traditional 
birth attendants 
■■ Drumming group 
promotes safe 
motherhood
■■ Recognition meetings 
for exceptional facilities
Technology ■■ Local radio programs
■■ Solar lighting
■■ Phones for VHTs 
■■ Bodas for mothers
■■ BABIES matrix for 
facility monitoring of 
newborn indicators
■■ Zambulances-terrain 
appropriate vehicles
■■ Kit (tackle) boxes for 
obstetric emergencies
■■ SMS pilot for following 
pregnant women in the 
community
UGANDA BOTH ZAMBIA EXHIBIT 14 
SMGL
innovations 
by country
“[The husband] 
has the bicycle or 
the money. Bring 
men on board from 
the onset—this is 
critical—or you do 
all this [awareness 
raising] and he will 
say ‘stay, don’t go.’”
— Zambia District Level
Implementing Partner
44 MARCH 2013
Prominent community mobilization innovations were Mama Kits and Mama 
Packs in Uganda and Zambia, respectively, to create demand for facility deliv￾eries. Districts also held a Safe Motherhood Day in Uganda and were planning 
a Safe Motherhood Week in Zambia, to galvanize women to come for services 
and increase community awareness. In Uganda, one district organized soccer 
matches using the half time break to encourage men to become informed about 
and involved in their partners’ pregnancies. One district in Zambia spoke about 
traditional drumming and drama groups that used performances to inform 
community members about maternal and newborn health.
Solar lighting installed in facilities was an innovative technology used in Uganda 
and Zambia. Providers were relieved that they no longer had to deliver babies at 
night by the light of their mobile phones. In Zambia, one implementing partner 
created “kit boxes” from fishing tackle boxes after they observed that providers 
frequently left a woman’s bedside during an obstetric emergency to obtain 
supplies. Kit boxes for postpartum hemorrhage, eclampsia, and other obstetric 
emergencies are stocked with the necessary supplies on maternity wards in 
this district. In Uganda, mobile phones were provided to VHTs to improve their 
communication with implementing partners, pregnant women, and drivers. 
An implementing partner in Zambia is conducting a pilot using mobile health 
technology to send text messages between facilities and community volunteers 
to assist in tracking pregnant women within their zones and promote ANC atten￾dance and facility delivery.
6.3 Challenges
Although the overall perception of SMGL was positive, respondents were frank 
in noting a number of challenges and weaknesses in program design and imple￾mentation. In analyzing the interviews we repeatedly found that some successes 
were also mentioned as challenges—sometimes by the same respondent. This is 
unsurprising given that large, complex programs, however helpful, create heavy 
burdens on policymakers and implementers, creating new problems and exacer￾bating existing bottlenecks.
Respondents listed a number of challenges that they felt impeded SMGL imple￾mentation and its potential for achieving and sustaining reductions in deaths 
of women and newborns. These ranged from organizational issues related to 
SMGL itself to deep-rooted problems endemic to development efforts. They are 
discussed below.
Short timeframe
While in many ways a motivator, the one-year timeframe of Phase 1 was a major 
challenge for implementation and an even larger one for sustainability. The pres￾sure to deliver was tremendous, making it difficult to meaningfully involve the 
Ministry of Health in implementation. As one respondent noted, “The short time￾frame … incentivizes ‘short cuts’ and approaches that do not promote capacity 
building or sustainability. For example, it is very tempting to just hire short-term 
staff to go to the districts and do things directly, instead of doing the right thing 
and working through the MoH.”
SMGL’s rapid start-up required nimble use of existing USG funding and 
procurement mechanisms. New funds were made available to CDC-funded imple￾menting partners within a very short time and existing funds were leveraged 
to support USAID-funded implementing partners, enabling the swift launch of 
SMGL in both Zambia and Uganda. However, many of the awards were limited to 
one year. Multiple implementing partners concurred that “running a long term 
program on a short term budget” posed important implementation problems, 
“We are moving at 
such a fast pace that 
one actually needs 
to sit down and 
think, ‘out of what 
we have done, what 
really is practical 
and sustainable and 
what can continue 
to be done?’”
— Zambia Implementing
Partner
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
Saving Mothers, Giving Life External Evaluation / Interim Report 45
most notably when it came to hiring staff and sequencing interventions: “We 
were trying to cobble things together on the fly, sometimes not as efficiently as 
you would want things to be.”
Some activities were explicitly modified due to concerns that funding may not 
be there after the end of Phase 1. For example, recently in Uganda voucher distri￾bution was limited to women in their third trimester of pregnancy to ensure that 
funds would be available when women actually deliver. 
The rapid pace and short timeframe of the SMGL program also occasioned 
frustration among USG agency staff, who were attempting to accelerate tradition￾ally slow processes such as procurement and approval of no-cost extensions in 
Atlanta and Washington.
Weaknesses in coordination and 
duplication of activities
Despite the predominantly positive view of the SMGL partnership, respondents 
in both countries pointed out that coordinating the large portfolio of partners and 
interventions presented major challenges. Respondents were concerned about 
a lack of clarity about the roles of CDC and USAID in the leadership of the SMGL 
program. The involvement of both US agencies and their respective grantees 
(i.e. implementing partners) created a large group of stakeholders, each with 
their respective agendas and programmatic approaches. For example, many 
implementing organizations already had maternal care interventions “on the 
shelf” and were understandably reluctant to change these, particularly for a 
one-year program—despite acknowledging the value of a coherent intervention 
package across districts for implementation and assessment of effects.
Although respondents agreed that frequent partners’ meetings were helpful 
for sharing action plans, many wished for more explicit guidance from the US 
government leads and/or the Ministry of Health on the content of SMGL activi￾ties—an arbiter of the SMGL approach. Respondents noted that only a mandate 
“from above” would convince implementing partners to change their existing 
approaches to activities like training, mentorship, supervision and support for 
procurement and infrastructure.
Respondents also spoke about the lack of clear roles for different implement￾ing partners. Implementing organizations and district MoH leaders in Uganda 
pointed to duplication and competition among partners, most notably with the 
voucher program in which several partners offered different versions of vouch￾ers at different prices in SMGL districts. Private health providers involved in 
Uganda mentioned incurring higher operational costs due to a lack of harmoni￾zation and coordination, again in the voucher system. Zambian respondents also 
noted the duplication of some activities, such as training for Safe Motherhood 
Action Groups, which was done by several groups, and complained about confu￾sion about who is doing what.
Communication between capital and district was also not optimal. District offi￾cials in Zambia reported that partners did not always inform the district health 
office about SMGL activities they were conducting in the communities, and that 
decisions were sometimes made with the MoH centrally, without discussion with 
the implementing districts.
Funding and logistic gaps
While the use of existing contractual mechanisms between USAID and CDC and 
their respective implementing partners enabled rapid provision of start-up 
funds and swift initiation of SMGL services, the one-year funding stream created 
difficulties for several implementing partners, whose grants ended at the end of 
“It was surprising 
to be so 
uncoordinated 
even though we 
were all working 
on the same 
thing in the same 
district.”
— Uganda District Level
Implementing Partner
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
46 MARCH 2013
September 2012. The slow pace of acquiring no-cost extensions meant that at 
least one Zambian partner had to suspend activities and pull staff out of the field 
for several months. USG respondents in Zambia mentioned that budgetary issues 
led to a scaling back of their initial transportation and refurbishment goals.
USG procurement regulations also markedly limited the speed with which 
partners could purchase vehicles for use as ambulances; this took more than a 
year in some cases. As of December 1, 2012 some districts were still waiting for 
renovated mother’s shelters, additional facility-level midwives, updated radio 
systems, and essential medical equipment. District-level MoH respondents in 
Uganda noted that inadequate blood supply was a remaining barrier to safe 
motherhood that had yet to be addressed by SMGL.
Finally, respondents cited ambiguity around roles and responsibilities for 
equipment training, delivery, and maintenance, which contributed to underutiliza￾tion of equipment that was available.
Difficulties with transport
The same difficult geography and weak transport infrastructure that contributed 
to the low rates of facility delivery in SMGL districts proved formidable barriers 
to SMGL implementation. The most common issue noted by district-level respon￾dents in both countries was the ongoing transportation barriers for both SMGL 
implementers and women seeking care, due to rough terrain, distance, and inac￾cessibility of health facilities during the rainy season. Although the issues had 
been anticipated, and SMGL included interventions designed to address them— 
in particular different types of vehicles—they continued to undermine successful 
implementation and facility delivery rates. Indeed, SMGL transport interventions 
were sometimes not suited to the terrain. For example E-rangers (three-wheeled 
motorcycle ambulances) were not able to traverse waterlogged dirt roads. In 
Zambia, frustration with transport was high due to procurement delays for badly 
needed ambulances and motorbikes.
Human resources
Staffing shortages were overwhelmingly cited as a persistent challenge in 
SMGL implementation at all levels in both countries. The baseline health work￾force levels were very low—even compared to other countries in Africa—with an 
estimated 0.48 skilled health workers (e.g. doctors, nurses, midwives, clinical 
officers) per 1,000 population in Zambian SMGL districts. There were 0.25 health 
workers per 1,000 in Uganda’s SMGL districts although these figures do not 
include all private sector providers. These numbers are one-fifth and one-tenth 
of the recommended minimum health worker numbers to provide essential 
maternal and child health services, respectively.35
Respondents in Uganda noted that the 134 additional doctors, nurses, and 
midwives hired by SMGL implementing partners—representing a 25% increase 
in overall health workforce—were essential to scaling up obstetric services. 
However, they noted that even this number was quickly proving insufficient as 
more and more women came to health clinics for antenatal care and delivery. 
Because the new staff were offered higher salaries to boost recruiting, Ministry 
of Health respondents at central and district levels were concerned that they 
would be unable to retain the staff once the project was over.
Zambia reportedly received 21 providers across two districts—far lower than 
was initially planned—and concerns over insufficient health workforce were 
voiced in all interviews in the districts. As one district-level informant noted, 
the scarcity of human resources “will undermine everything we do if it is not 
addressed in some way.”
“Low staffing 
cannot keep 
up with the 
increasing 
demand brought 
on by SMGL; we 
are swamped and 
overworked.”
— Zambia District Level
MoH Official
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
Saving Mothers, Giving Life External Evaluation / Interim Report 47
“SMGL has helped 
with our ability to 
gather local data, 
which has helped 
us with planning in 
other areas of our 
health system.”
— Uganda Implementing
Partner
Finally, as described in greater length in the following section (6.4), several 
respondents in Zambia in particular felt that the proliferation of off-site training 
in Zambia had exacerbated facility-level staffing shortages by taking workers 
away from their clinics.
6.4 Synergies and unintended consequences
Because health systems are complex adaptive systems, the introduction of 
new programs is often accompanied by a host of secondary effects—including 
synergies and unintended consequences. In some cases, these may be a surprise 
to program implementers, while in others, they are predictable enough to antici￾pate and plan around. In both Uganda and Zambia, the launch of SMGL activities 
resulted in outcomes that were not primary program objectives. These “secondary” 
effects—both positive and negative—are discussed below.
Positive synergies and positive
unintended consequences
At the national level, respondents in both countries described a range of positive 
changes due to SMGL. SMGL clearly increased the enthusiasm for action on 
maternal and newborn health among national policymakers and their devel￾opment partners. However, in Uganda, the program also sparked action to 
strengthen the health system more generally. The high level of parliamentary 
interest in SMGL, most notably in Uganda, was cited as a key factor for prompting 
lawmakers to conduct a review of the national health budget and increase salaries 
for physicians nationwide.
At the district level, Ministry of Health respondents in Zambia observed a 
“spill over” of skills and knowledge from SMGL to non-SMGL staff. There were 
positive effects on staff morale, as health care workers saw and appreciated the 
improvement in the quality of care provided to patients. Some respondents felt 
that the increased attention to record-keeping and data management would also 
be likely to spill over to non-SMGL programs, to beneficial effect. There were also 
synergies with HIV and syphilis programs. Increased demand for facility-based 
delivery led to reports of increased uptake of HIV testing, PMTCT services, and 
testing for syphilis and other sexually transmitted infections.
At the community level, the inclusion of traditional birth attendants (TBAs) in 
SMAGs and VHTs was described as an important and positive secondary effect of 
SMGL in both Zambia and Uganda, shifting their role from an impediment to safe 
motherhood to an ally in promoting facility deliveries.
Negative synergies and negative
unintended consequences
Concerns were expressed about maternal health overwhelming other health 
sector priorities. One implementing partner characterized this risk by observing 
that “HIV has become stale. This [SMGL] seemed tangible and new.” Similarly, 
dozens of respondents expressed concern that the exclusion of family plan￾ning from SMGL had not only created a missed opportunity to reduce maternal 
mortality, but had de-emphasized the importance of family planning as a health 
strategy in the country more broadly.
Respondents also noted that the SMGL districts had, to some extent, benefited 
at the expense of other districts, with implementing partners in both countries 
noting that some activities and programs were redirected from non-SMGL to 
SMGL districts. Similarly district level MoH officials noted the diversion of time 
and district-level staff away from non-SMGL activities: one District Medical 
Officer noted that 60% of his time was spent managing SMGL, which necessitated 
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
48 MARCH 2013
O F F S I T E
O N S I T E
ZAMBIA
NYIMBA DISTRICT
O N S I T E
O F F S I T E
UGANDA
KIBAALE DISTRICT
“Alarm” trainings are a series of trainings on emergency obstetric 
and newborn care (including CEmONC, BEmONC and “Helping 
Babies Breathe”), conducted by the Association of Obstetricians 
and Gynaecologists of Uganda.
Graphs are based on a 250 day work year. One 
district was selected from each country based 
on data availability. Shaded areas represent SMGL￾related trainings and mentorings experienced by 
a representative SMGL health worker in a Ugandan 
HC III and a Zambian Health Facility.
EXHIBIT 15 
A “year in the life” 
of a typical SMGL 
health worker
5 Days First alarm training
5 Days Fourth alarm training
4 Days Second alarm training
231 Days Work days with no SMGL training
5 Days Third alarm training
20 Days Emergency obstetric and newborn care
5 Days Essential medicines logistics improvement program
191 Days Work days with no SMGL training or mentoring
4 Days Helping Babies Breathe
5 Days Maternal death review
5 Days Medical records
20 Days Mentoring
delegating other health concerns to subordinates.
Although the SMGL logic model addresses the full range of Demand, Access, 
Quality and Systems strengthening interventions, the sequencing of activities 
was not optimal and created unintended negative effects. For example, trainings 
of health workers and community groups (SMAGs and VHTs) were launched 
months and in some cases more than a year before improvements in infrastruc￾ture, supply chains, and transportation. 
The community level activities generated demand for services before the 
expansion and improvement of supply. This led to surging numbers of pregnant 
women coming to facilities before facilities were expanded, upgraded, or staffed 
with new workers. Respondents were concerned that poor experiences in the 
facility would lead to disappointed mothers and communities and a decline in 
trust in the health system. There were also complaints of burnout and overload 
for “overwhelmed” staff at some SMGL facilities as well as stock outs of medi￾cines and supplies, power shortages, and water shortages that were attributed 
to the marked increase in demand for services.
SMGL provided extensive training to facility-level clinicians. Because the bulk 
of this training was provided off-site and not at the health facility, the training 
removed health workers from clinics and thus itself reduced availability of 
services. Off-site training exacerbated health worker shortages particularly in 
Zambia where only a handful of new health workers were hired for SMGL. Exhibit 
15 illustrates a “year in the life” of a typical SMGL health worker. Using a base 
of 250 work days, we estimated how many days per year a typical midwife at a 
health center spent in SMGL-related trainings (off-site) and mentoring (on-site), 
based on actual trainings delivered in the district. In Nyimba, a midwife could be 
expected to spend 39 work days at off-site trainings.
District officials in Uganda observed that the higher salaries of workers hired 
as part of SMGL created resentment among “non-SMGL staff caring for other 
patients.” A few respondents were also concerned about the “poaching” of staff 
from other health programs.
Two potential unintended consequences were noted at the community level. 
In Uganda, we repeatedly heard concerns that efforts to improve access to and 
quality of facility-based deliveries were actually incentivizing women to have 
more babies, although no concrete data were provided to support this assertion. 
In Zambia, district health officials and implementing partners reported that some 
village leaders had begun imposing penalties (e.g. a fine of a chicken or a goat) 
on the husbands of women who decided to deliver at home instead of a health 
facility. These efforts, which were not sanctioned by SMGL and maybe perceived 
as coercive, are concerning as they may create a punitive instead of a supportive 
environment for women and are likely to affect the most vulnerable women —the 
poor and women living farthest from facilities—disproportionately. Their effects 
on women and communities should be carefully studied as SMGL goes forward.
Exhibit 16 synthesizes the primary and secondary effects—both positive and 
negative—that were most apparent to SMGL stakeholders by the six-month mark 
and that we discussed above.
6.5 Country ownership and sustainability
Respondents in both countries reported that SMGL was highly congruent with 
national maternal health strategies both in terms of policy focus and specific 
activities, with the notable exception of family planning. All agreed that this 
congruence was a key prerequisite for country ownership. From the outset, 
USG leads engaged their Ministry of Health counterparts in planning for SMGL 
implementation and made a concerted effort to include them in implementation 
“I think that we 
must ensure that 
our communities 
should own their 
health. No one 
should feel that 
their health is in 
the hands of the 
US government.”
— Uganda District Level
MoH Official
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
50
EXHIBIT 16 
Matrix of reported primary 
and secondary effects
PRIMARY EFFECTS SECONDARY EFFECTS
POSITIVE NEGATIVE
■
■
■
■
■
■
■ ■
■
■
Increased 
community 
commitment Increased 
community 
confidence 
in providers
■
■
■
■
■
Increased male 
engagement
■
■
■
■
■
■
■
■
■
Increased 
demand for 
facility-based 
delivery
Increased 
advocacy for 
maternal health
Enhanced 
infrastructure
Improved quality of BeMONC, 
EmONC, HBB services
Community maternal 
deaths captured due 
to community key 
informants in Zambia
Enhanced 
supply chain
Increased uptake of HIV 
and syphilis testing
Increased interest 
from new/other donors
Improved overall 
coordination between 
CDC, USAID, DoD, PC
Spillover 
of skills 
from SMGL￾trained staff 
to other staff
Increased 
advocacy for 
health in general
Increased national health 
budget, higher salaries for 
doctors (in Uganda)
Spillover data management 
skills to outside programs
Shift of TBA role from safe 
motherhood impediment to 
facility delivery advocate
■
Improved 
health 
worker 
morale
Improved and 
more timely data 
on mothers and 
newborns
Major time commitment 
for DMOs; of SMGL
Heavy workload 
for staff, stock 
outs, and 
shortages due 
to increased 
demand
Health programs 
diverted from 
non-SMGL districts
Penalties imposed by 
chiefs for non-facility 
deliveries in Zambia
Tension amongst 
SMGL and non-SMGL 
health workers
Staff shortages due to 
off-site trainings
Crowd-out of 
non-MCH services and 
programs in districts
■
Saving Mothers, Giving Life External Evaluation / Interim Report 51
meetings and decisions. In both countries, the ultimate selection of districts 
was a Ministry of Health decision. The Ministries’ visible role during the planning 
process, including in the development of district work plans, in turn generated 
support for the program among district health officials and community lead￾ers. Respondents in both countries also pointed to the support of First Ladies, 
high-ranking MoH staff, and parliamentarians as examples of high-level buy-in.
Informants highlighted the key distinction between government ownership versus 
government support of the program. The fact that SMGL was almost entirely donor￾funded and followed an externally-set timeline raised questions about the extent of 
true government ownership at the national level as well as the sustainability of the 
initiative. Many respondents pointed out that for true ownership national govern￾ments should have made a financial commitment supporting the aims of SMGL at 
the outset, particularly in areas such as poor roads, human resource shortages, and 
crumbling infrastructure, that cannot be addressed by donors.
Hands-on engagement with the program was felt to be highest at the district 
level. Respondents in both countries highlighted the important role that local politi￾cians and traditional leaders had played in generating district- and community-level 
support and facility-based deliveries. “Ownership is at the district level because 
this is where implementation occurs,” said one informant. Below, we report some 
specific findings for each country.
Uganda
Although the majority of respondents at both district and central levels rated 
MoH ownership of the SMG program relatively high, with most giving it an average 
rating of 7 out of 10 on a scale where 10 was excellent, respondents were quick 
to point out that this only referred to the MoH’s “enthusiasm” for the program 
rather than a guiding role in implementation. Some noted that the program was 
initially viewed as “made in the USA” and that “you don’t see the Ministry involved 
in planning meetings, at the Kampala and district level” to the extent they should be 
involved. A general sentiment voiced by central and district level partners was that 
government ownership was lower than it could have been due to the short timeline 
for the project: “The MoH…are willing to participate [but] one year is too short. It’s 
not how government works.”
However, respondents noted despite their limited role in design and oversight of 
SMGL, the MoH demonstrated its commitment to the goals of SMGL in numerous 
ways, including supporting construction and refurbishment projects, and, above all, 
investing in health workers: “No one knew that government could be pushed as far 
as they were on human resources. And now the government is trying to sustain the 
program.” Respondents pointed to the doubling of doctors’ salaries at the Health 
Center IV level (health facilities with inpatient care) that was approved by Uganda’s 
parliament in fall of 2012. The Honorary Victoria Rusoke, Member of Parliament in 
the SMGL district, Kabarole, was mentioned numerous times for her skill in advo￾cating for the salary increase as a means to bring health workers to SMGL and other 
rural districts. Her role as a champion for women’s health tremendously energized 
SMGL district implementers.
Most respondents agreed that, “at the districts, country ownership has been high. 
Honestly, they are doing everything they should do and it is working.” District health 
officials were especially commended for being aware of all the SMGL partners and 
activities taking place in their districts, being responsive to the needs of SMGL part￾ners, and for making time to attend and coordinate meetings among the different 
stakeholders in their districts.
“[A maternal 
death] is a big big 
deal, even the DHO
and other local 
politicians are 
talking about it 
and they are now 
vigilant.”
— Uganda District Level
Implementing Partner
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
52 MARCH 2013
A number of respondents highlighted the role that the Director General of the 
MoH had played in chairing SMGL meetings in the districts and explained that this 
showed that the program was important to the government. An implementing 
partner had this to say about district ownership: “I would say whenever we have 
needed the district to do something for us they have done their best to get it done. 
The political will is there and the technical people have gone out of their way 
to make sure services flow seamlessly.” District health officials, in turn, stated 
that their interests had been adequately taken into account. One district health 
official had this to say: “We are really involved in these activities, they have taken 
us into account as they have planned, designed, and implemented”. A commonly 
expressed sentiment across all respondents was that the level of district 
ownership was being greatly hampered by the fact that they have not committed 
funding: “The districts can’t be 10 out of 10 because they do not have financial 
resources. The next phase should be the government commitment.”
Zambia
Many respondents commended the MoH for being committed to reducing mater￾nal mortality prior to SMGL, citing as an example that the President’s wife was 
a physician who promoted facility deliveries during public appearances. The 
general view among respondents in Zambia was that ownership of SMGL at MoH 
level was “good,” but that actual participation in planning and implementation 
lagged the expressed interest. Multiple respondents attributed this largely to the 
fact that the MoH was “stretched thin,” both financially and in terms of human 
resources, and was thus not always able to participate in the many meetings 
and planning activities required to launch SMGL. Some respondents noted that 
although the MoH had been actively involved in the planning phase, including the 
selection of SMGL districts, the organization of various trainings, and visits to the 
districts, this level of involvement had waned during actual implementation.
The rapid pace of SMGL did not fit well into existing Ministry decision-making 
processes. One respondent noted that, “this is a fast moving freight train and it 
was difficult for them to jump on board.” Several respondents pointed out that 
the elections and subsequent creation of the Ministry of Community Development 
Mother and Child Health further hindered effective engagement of national 
officials: “Between December 2011 and December 2012 we had a great lack of 
clarity as to which Ministry was going to be responsible for what. In December 
2012 they finally laid out what each ministry would do.”
Positive examples of MoH support for maternal survival included introduction 
of programs with direct entry into midwifery training, introduction of a quality 
improvement curriculum for nursing schools, and the fact that many SMGL train￾ings had been initiated and organized at national level. 
The majority of respondents in Zambia rated district ownership of SMGL very 
highly and described the districts as being “in the driver’s seat.” Implementing 
partners reported that district health officials were highly invested in the initia￾tive and regularly participated in SMGL activities taking place in their districts. 
The District Medical Officer of Mansa even became a mentor in the SMGL mentor￾ship program, supporting nurses and midwives in providing quality obstetric care 
in the clinics. An implementing partner noted: “I can just call the district office 
and they will give me a vehicle, because they know what I’m doing is not for [this 
implementing partner], but it is for Mansa District”. Respondents in Zambia noted 
the involvement of key local leaders, such as the chiefs and district commission￾ers, as positive, and important, examples of SMGL ownership at district level.
“For us the district 
has been very 
instrumental to 
coordinate and 
provide everything 
that we need to 
do. They’re excited 
like everybody 
else... They’ve 
been involved in 
everything that is 
being done.”
— Zambia District Level
Implementing Partner
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
Saving Mothers, Giving Life External Evaluation / Interim Report 53
“The private 
sector brings in 
innovative ways 
of dealing with 
[equipment] 
like ultrasound 
machines, for 
example. The 
public sector 
is very poor in 
maintenance and 
the private sector 
has the ability to 
play an important 
role.”
— Uganda District Level
Implementing Partner
Sustainability
Two major opinions emerged among respondents in both countries regarding the 
sustainability of SMGL moving forward. A large group of respondents expressed 
concern that all the gains and progress made in addressing maternal mortality 
in the districts would be threatened by the end SMGL funding without a clear 
transition plan to the national government. These respondents cautioned that, 
“if we don’t think about sustainability “we are just ‘fixing pipes’ but not seeing 
where there are leaks.” The lack of a long-term plan for SMGL created “a wave of 
uncertainty [for managers and health workers] because they have seen so many 
programs come and go. It creates anxiety because they are worried the tap will 
be turned off.”
There were clear tensions between the short period of time that partners had 
to implement activities and planning for sustainability. One respondent noted “It 
would take five years of sustained inputs for this program to be anywhere near 
sustainable.”
In Uganda, Ministry of Health respondents worried about the country’s ability 
to retain SMGL facility staff whose higher salaries would be difficult to “absorb by 
a government with limited resources.” District health officials and district-level 
implementing partners felt strongly that putting the community in the driver’s 
seat was important for sustainability. For example, one partner noted that a year 
into the project villages should start running the community dialogue meetings 
themselves and contribute to them without continued support from implementing 
partners. Another noted that the community should be empowered to look after 
certain costs themselves, such as purchasing a community motorcycle to bring 
women to facilities, to prevent the attitude of “letting others do things for them.” 
There was recognition that this would require income-generating activities to 
create capital in the community. Most implementing partners felt that volunteer 
health team members should receive meaningful monthly stipends if they were 
to continue their work into the future.
In Zambia, an implementing partner respondent in one district worried that 
the level of help their organization was providing the DHO (e.g. bringing in equip￾ment, providing support by phone during obstetric emergencies) may be creating 
a dependence that would prevent the district from running the program once 
external funders withdraw from SMGL activities. Certain activities were consid￾ered more sustainable included maternal death reviews and trainings, which had 
occurred to varying degrees before SMGL.
A smaller group of respondents in both countries was more positive about 
SMGL sustainability moving forward. These respondents noted that the SMGL 
program had made substantial investments in strengthening district health 
systems and that these investments would ensure continuity of SMGL. These 
included infrastructure investments such as operating theater renovations and 
health information systems. They also pointed to the existing maternal health 
expertise at the Ministry of Health and the involvement of key leaders and 
stakeholders as an important investment that would ensure SMGL sustainability 
moving forward.
6.6 Role of the private sector
Private sector involvement was substantially greater in Uganda than in Zambia, 
reflecting the larger role of the private sector in health care delivery. The private 
sector in Uganda was involved in transport and delivery care. Pregnant women 
were given vouchers that could be redeemed for motorcycle and taxi transport to 
facilities for delivery and antenatal care as well as for ambulance services from 
lower- to higher-level facilities if an obstetric emergency arose. Further, SMGL 
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
54 MARCH 2013
implementers contracted with private clinics and hospitals to provide antenatal 
and delivery care to quickly expand the range of facilities that women could 
deliver in using delivery vouchers. One of the implementing partners in Uganda 
also provided performance-based contracts in which investments were made in 
equipment and private providers were paid bonuses for meeting and exceeding 
targets for four antenatal care visits and deliveries. Staff from private facilities 
also provided mentorship and supportive supervision at lower level facilities.
The private sector was also involved in other components of SMGL in Uganda. 
Corporate contributions to SMGL were noted, including discounted phones and 
airtime (from Warid and MTN), fuel (from Total Gas) and discounted radio airtime. 
Ugandan professional associations of obstetricians, pediatricians, and anesthe￾siologists were contracted to conduct training and mentoring of health staff to 
increase technical competence in life saving maternal, surgical, and newborn 
resuscitation.
There was broad consensus among Ugandan implementing partners and 
district government officials that the public/private partnerships in the SMGL 
focus districts enhanced service delivery and quality of care for women in the 
districts. However, some expressed concern that SMGL’s encouragement of 
private facilities to expand their capacity might draw patients away from the 
public system. Subsequently, direct support to the private sector should be 
balanced with overall goals of strengthening the public health system—the main 
means for sustaining SMGL services for the foreseeable future. One specific flaw 
that some pointed to in the model was that private-sector vouchers were not 
usable in public sector facilities, even in the case of a medical emergency.
Overall, private sector respondents reported a positive experience with the 
SMGL program and noted several benefits to participation in the project—partic￾ularly the ability to upgrade their equipment and expand their client base. Private 
facility managers suggested that standardizing vouchers would help reduce their 
transaction costs and that the flat fee reimbursement method for Caesarean 
sections needed to be adjusted for complicated cases. All respondents were 
concerned about sustainability: the private sector worried that they would lose 
patient volume if vouchers went away and public sector workers expressed fears 
that their private colleagues would cease to collaborate in the SMGL program 
once direct subsidies were reduced.
Zambian respondents noted that there had yet to be engagement either at the 
corporate level or at the private provider level in Zambia, where there are few or 
no private providers in the SMGL districts. Partners were enthusiastic about the 
potential for private sector partners, identifying key gaps and challenges that 
might be best met by the private sector, including the construction of mothers’ 
waiting shelters and purchase of vehicles—both activities that are challenging for 
USG partners under their procurement regulations. Merck for Mothers has issued 
a request for proposals to test maternity waiting shelter models.
Beyond gap-filling, respondents were eager for new thinking about how to 
tackle maternal mortality that they felt private sector partners could bring to 
the table. One suggestion was that Merck for Mothers, an SMGL lead partner 
with extensive experience in corporate social responsibility, could help to enlist 
the domestic private sector in taking a greater role in fundraising or promoting 
visibility of maternal and newborn health: “We need help with the private sector. 
We are public health people, so this isn’t our expertise…[but] if the CEO of Merck 
spoke to the CEO of Airtel, they could make some real progress.”
“Vouchers need 
to be restructured 
so as not to 
kill the current 
system. Otherwise, 
they will have a 
negative impact on 
the public sectors.”
— Uganda Implementing
Partner
6 SUCCESSES, CHALLENGES, LESSONS LEARNED
Saving Mothers, Giving Life External Evaluation / Interim Report 55
7.1 Assessment of SMGL’s first six months
Our interim assessment found a tremendous amount of enthusiasm for SMGL 
in both countries. It was striking to hear positive sentiments about SMGL from 
respondents ranging from senior national government and USG officials in the 
capitals to maternity ward nurses in the districts. The support for continuing and 
expanding SMGL was strong and undiminished by the assortment of challenges 
encountered in the start-up phase.
In synthesizing the lessons learned from the countries at mid-point of Phase 1,
 we found that the “big push” model—high global visibility, large investment, 
demand for rapid results—succeeded politically and operationally by raising the 
stakes for national governments and US development partners in country.
SMGL catapulted maternal and newborn health to prominence on national 
health agendas. It gave Ugandan and Zambian health ministries a way to oper￾ationalize national commitments to reducing maternal mortality by bringing 
visible improvements in services for mothers and newborns in intervention 
districts within a short time. This large investment, combined with Ministry of 
Health involvement, motivated district health managers to tackle the challenge 
of managing a large influx of activities. In sum, the magnitude of investments and 
high-level attention to SMGL meant that failure was not an option—for USG or 
national governments.
However, the imperative for rapid results combined with the nearly 100% US 
government funding which—as per US law, was spent by US agencies working 
with NGOs rather than through government budgets—diminished countries’ 
ability to own and direct the program. Ministry of Health respondents at central 
and district levels pointed out the important distinction between support and 
leadership: while they strongly back the goals of SMGL and, for the most part, the 
means to achieve them, they do not feel to be in the driver’s seat. They want this 
to change as SMGL moves into Phase 2. As one senior official put it, “We are very 
grateful [for SMGL investments] but not to be figureheads.”
Another downside of the rapid nature of the initiative was the lack of a plan on 
how to sustain gains made. When SMGL was launched, there was no clear plan 
for how the training would be repeated for new entrants, how mentoring can 
be continued, how equipment will be serviced, maintained, and replaced, how 
community groups would be supported to continue their work beyond the end 
of the funding commitment. Some of the challenges brought on by the abbre￾viated planning period and the short timeline included problems in sequencing 
inputs and activities, a large training load for health workers, and budgetary and 
procurement delays due to off-cycle budgeting—all of which frustrated imple￾menting partners and their district counterparts.
“I started working 
in maternal and 
child health 35 
years ago. It’s been 
a dream come 
true to participate 
in a program this 
serious and of this 
magnitude.”
— USG Zambia Official
7
Preliminary assessment and 
recommendations for Phase 2
56 MARCH 2013
7.2 Recommendations for Phase 2
As it expands to other districts and countries, SMGL will adapt to local needs, 
goals, and capacities. However, useful lessons have emerged from our evaluation 
of the experience of Uganda and Zambia that can inform program design and 
implementation. One important lesson for Phase 2 is that successful imple￾mentation of SMGL requires a national commitment to maternal and newborn 
survival at the highest level and a highly functional team of development part￾ners with a strong relationship with the Ministry of Health. Below we propose 
several recommendations for the content of the intervention package and its 
implementation in Phase 2.
Content of SMGL package
1. Continue to engage communities as change agents 
SMGL’s work with communities to raise demand for maternal care was univer￾sally viewed as a game-changer in the effort to save maternal and newborn 
lives. Both community action models—the Volunteer Health Teams from Uganda 
staffed by community health workers and the Safe Motherhood Action Groups in 
Zambia, which include community members at large—were felt to be successful 
in bridging the gap between women and the health system. We also heard early 
indications—which we will confirm in our final evaluation—that the community 
initiatives were playing an important role in reducing the inevitability of maternal 
and newborn deaths in the community. A community action group model adapted 
to community structures of the country should be part of the intervention pack￾age in Phase 2.
Community health activities require people’s time and are unlikely to be 
sustained on a volunteer basis indefinitely. In Phase 2, implementers, including 
the Ministry of Health, should consult communities on ways to compensate 
groups for their efforts. Ideas range from modest payments from the health 
system to providing seed funds and training for income-generating activities as 
an incentive to stay involved in the effort. In Zambia, implementers mentioned 
funding small mills for grinding maize or providing bicycles for personal, as well 
as program, use to facilitate access to markets.
2. Invest in the health system triad: train staff, improve, 
and equip facilities
The combination of trained, confident staff and functioning facilities with reliable 
equipment is highly synergistic. While training on how to treat hemorrhage or to 
resuscitate an unresponsive newborn is crucial, health workers cannot practice 
these skills effectively without running water, lights, or steady supply of oxytocin 
or IV tubing. In addition, research shows that lack of functional equipment is 
a major disincentive for recent medical graduates to accepting rural postings 
in Africa.36, 37 The combined investment in human capacity and infrastructure 
in the first phase of SMGL is an excellent model for future efforts. There is no 
question that external donors cannot do this alone. Facility upgrades—and criti￾cally—repair and maintenance of facilities need to be part of core health system 
investments by national governments.
One caveat is timing. It is important to invest in improving quality at the outset, 
before women begin arriving in large numbers. This sequencing will help to 
ensure women and newborns receive appropriate care and do not leave the facil￾ity disappointed with the quality of care, affecting their future health choices and 
potentially those of their peers.
“Before we could 
only watch babies 
die. ‘Helping 
Babies Breathe’ 
[SMGL-supported 
training] changed 
everything; it 
showed me I can 
do something.”
— Zambian nurse
7 Preliminary assessment and recommendations for Phase 2
Saving Mothers, Giving Life External Evaluation / Interim Report 57
3. Expand the focus beyond 24 hours
While intrapartum and immediate postpartum care is an excellent starting point 
for reducing maternal and newborn mortality, solitary focus on the 24 hours 
around delivery in SMGL misses crucial opportunities to save lives and will 
limit its overall effectiveness in the future. Family planning, newborn care, and 
HIV services—in particular PMTCT— figured more prominently among national 
maternal and reproductive health plans than they did within SMGL. These 
services are also provided using shared platforms (e.g. maternity wards and 
outpatient clinics) often by the same health workers. The focus on 24 hours intro￾duced unnecessary cleavages between this set of essential services.
Family planning, which reduces women’s risk of maternal death, should be 
strengthened as part of Phase 2. Three in ten women in Uganda and Zambia have 
an unmet need for family planning, which results in unwanted or mistimed preg￾nancies and, in turn, maternal deaths.12, 16 Care of post-abortion complications, 
already being implemented in some SMGL districts, is another important inter￾vention to reduce maternal deaths.
Other services that require greater focus in Phase 2 include prevention of 
mother-to-child transmission of HIV; a necessity for decisive reductions in mater￾nal and newborn mortality in countries with high HIV prevalence. Finally, half of 
newborn deaths occur after the first 24 hours of life—these infants require close 
high quality health care after they leave the hospital with their mothers.21, 22
4. Develop and test solutions for two bottlenecks: human resources 
and distance to care
Although SMGL made investments in human resources (particularly in Uganda) 
and transport to bring women to facilities, it has not produced effective or lasting 
solutions for these key constraints to maternal survival. This is because the 
size and distribution of the health workforce and extent of the road and health 
infrastructure are issues that lie largely outside the scope of a globally led effort, 
requiring national policies and investments.
It is unlikely that SMGL can sustainably provide high quality obstetric care 
to all women in the district with the number of health workers currently on the 
ground—not without staff burnout and attrition. Concerns about work overload 
were already expressed six months into the program and will only grow with 
the volume of facility deliveries. Phase 2 should expand on approaches already 
attempted in Phase 1: contracting with private providers where they are available 
and hiring recently retired nurses and midwives. Other potentially effective solu￾tions include rural incentives such as salary increases, more rural rotations for 
senior medical students and residents, task shifting to non-physicians for obstet￾ric surgery, and remote clinical phone or radio support to frontline providers in 
smaller clinics. Ministries of Health should spearhead these efforts.
Even with expanding health budgets, health systems in resource-constrained 
countries will not be able to provide high-quality clinics close to all women, 
particularly in large areas with low population densities. Indeed, it is not clear 
that all women want to deliver at primary care clinics nearby, as studies show 
substantial proportions favor hospitals.38 Bringing obstetric services to women 
will thus require scaling up of transport options—ambulances, private taxis—but 
also other innovative solutions, such as mothers’ shelters near hospitals. The 
human resource and transport approaches need to be carefully tested in Phase 
2 to ensure they meet the needs of women and contribute to maternal and 
newborn survival.
“[Lack of] family 
planning is a huge 
barrier and makes 
our job harder 
when women 
are coming in 
for their eighth 
child or they get 
pregnant too soon 
after another 
pregnancy.”
— Ugandan Midwife
7 Preliminary assessment and recommendations for Phase 2
58 MARCH 2013
Implementation
5. Coordinate around one plan
Initiatives to promote universal coverage with quality obstetric and newborn care 
are growing; SMGL is one among many initiatives to focus on this issue in low-in￾come countries. This proliferation of good intentions is an exciting new stage in 
the global fight for maternal survival but also presents a growing challenge for 
coordination. Despite the presence of SMGL district coordinators, health system 
managers have told us they struggle to stay abreast of activities of implementing 
partners and manage their regular portfolios. Similarly, at the national level, 
implementing partners were concerned about duplication of activities, which led 
to inefficiencies and confusion among stakeholders in districts. Finally, the lack of 
a single harmonized set of activities across districts was cited as a weakness in 
the design of SMGL that hampered comparison of effectiveness and development 
of an ultimate model for the country.
For Phase 2, SMGL should strive for greater coherence in program activities, 
while allowing for some local adaptation. One useful precedent may be the 
“Three Ones” effort, which sought to harmonize the multitude of AIDS activities 
and in-country programs in the era of rapidly expanding antiretroviral treat￾ment.39 The “Three Ones” are: one action framework, one coordinating authority, 
and one monitoring and evaluation system. It would be ideal for global and 
national SMGL partners to agree to a single set of interventions, with implemen￾tation clearly allocated in country to avoid duplication. The global partners should 
then work to support key aspects of the action framework, whether through 
funding, implementation, advocacy, or testing of innovations. To minimize the 
burden of donor management for overstretched Ministry of Health senior staff, 
USG agencies should continue to play a key liaison role with host governments. 
Finally, a single standard approach to monitoring and evaluation will be critical to 
tracking progress and making corrections.
6. Commit to five years—with a transition plan
One clear observation emerging from our evaluation of Phase 1 is that interven￾tions attempting to change health systems and the customs and behaviors of 
populations cannot be done on a one-year timeline and budget. While there was 
value to having a first year “big push” to raise attention to issue and build a cadre 
of champions in country, SMGL should approach the next phase with a health 
system legacy plan that clearly spells out how SMGL will leave health systems 
better off for mothers, babies, and health workers.
In Phase 2, SMGL global partners should make minimum commitments 
of three to five years to enable appropriate planning, engagement of local 
ministries, sequencing of interventions, and crucially, permit the laying of the 
groundwork for sustainability. Future SMGL countries should also start with a 
three to five year plan to ensure lasting gains.
Going forward, the role of national governments and district authorities should 
be clearly outlined in SMGL plans. From the outset this should include govern￾ment investments in core areas such as infrastructure and human resources. 
Beyond this, there needs to be a transition plan that lays out how countries will 
assume responsibility for the program beyond five years—perhaps linked to 
greater direct budget support from donors. This approach leverages the lessons 
of the PEPFAR model, which started out as emergency response managed by the 
US and is now transitioning to being a source of funds for national governments 
and their partners to use to sustain services.
7 Preliminary assessment and recommendations for Phase 2
“We can’t solve 
the country’s 
problems in only 
4 districts, but at 
least we have a 
model that works.
Let’s use these 
lessons. Let’s 
project how long it 
will take and then 
leave it totally to 
local people.”
— USG UGANDA OFFICIAL
Saving Mothers, Giving Life External Evaluation / Interim Report 59
7. Collect (a few) standard measures to optimize implementation
SMGL is nothing less than an overhaul of maternal and newborn healthcare with 
consequences across the entire health system. Reliable data, collected and analyzed in 
a timely fashion, will be essential to inform health system managers and other stake￾holders about progress and alert them to any negative consequences and interventions 
going off track.
In Phase 2, a small set of standard tracer indicators could be collected quarterly from 
facilities in SMGL districts to give a picture of utilization, mortality, referrals within the 
system, as well as attrition or hiring of health workers, trainings, and overall facility 
functioning. The indicators, some of which may already be routinely collected, will be 
valuable for management of the health system overall and not just SMGL. These data 
will only be useful to the extent they are accurate; enhancing quality of data collection 
should thus remain a core SMGL activity. Ideally, Phase 2 should institute similar moni￾toring systems in non-SMGL districts to understand how SMGL affects care delivery 
over and above the underlying secular trends affecting the entire health system.
These tracer indicators could be augmented with special studies, as required. 
Examples of special studies include assessments of quality of care, provider motivation 
and intent to leave, women’s satisfaction, and causes of maternal and newborn deaths. 
As noted above, SMGL partners should also commission implementation science experi￾ments on what interventions are most effective and feasible to add to the SMGL portfolio.
8. Harness the private sector 
Although there were positive experiences in engaging the private sector in Uganda, 
there was a feeling in both countries that private sector expertise and capacities were 
underutilized. Respondents in our evaluation noted that they needed “all hands on 
deck” to reduce maternal mortality: public sector, private companies, and churches 
and faith based organizations. In particular, while the bulk of responsibility for health 
care in both countries lies with the government—as it does in most industrialized 
countries—there is a fundamental role for the private sector to play in supporting 
maternal survival.
Three arenas for private sector action emerged from our study. First, as discussed 
in Section 6.6, the global and national private sector can help in areas where USG and 
other partners can offer limited assistance and do not have core expertise, in particular 
in infrastructure, transport, and communications. Second, thriving domestic industries, 
such as mining and tourism should be encouraged to get involved in maternal and 
newborn survival, by raising money and conducting advertising campaigns, for exam￾ple. Large mobile phone companies could expand networks or donate airtime for health 
workers and community volunteers. Private global SMGL partners could help make this 
case most effectively.
Finally, the private sector has technical, managerial, and communications expertise 
often lacking in the public sector—expertise that will be essential to sustain high quality 
of care and community demand beyond the initial investment. From leadership training 
for health sector managers to assisting with supply chain, logistics, and equipment 
maintenance to helping community health volunteers launch small businesses, to 
promoting a customer service mentality in clinics—these are just a few ideas we heard 
about how private sector partners can strengthen the implementation of SMGL.
7.3 Next steps
In the next stage of the external evaluation we will focus on three areas: 1) Response 
of the community and providers; 2) Extent of implementation at endline; and 3) 
Assessment of the SGML partnership. This work will be conducted in collaboration with 
local research partners in Uganda and Zambia and will take place between May and 
July, 2013. We will produce a final report in fall of 2013.
“[Maternal 
and newborn 
mortality] 
is a fight for 
everybody. 
Anybody’s wife 
or daughter 
can die. Many 
people are 
looking for the 
government to 
do everything 
and they can’t.”
— Zambia Implementing
Partner
7 Preliminary assessment and recommendations for Phase 2

61
8
Appendix
62 MARCH 2013
Uganda Year
SOURCE
Zambia Year
SOURCE
Population & Economy
Population 33,425,000 2010 1 13,089,000 2010 1
Number of women of 
reproductive age (15-49) 7,298,000 2010 2 2,893,000 2010 2
Total fertility rate * 6.1 2010 1 6.3 2010 1
Crude birth rate (births per 
1,000 people per year) 45.0 2010 1 46.0 2010 1
Percent of the population living 
on less than $1.25 (PPP) a day 
(at 2005 international prices)
38.0 2009 3 69.0 2006 3
Human development 
index (0-1) ** 0.45 2011 4 0.43 2011 4
Gross national income 
per capita (PPP current 
international $)
$1,310 2011 5 $1,490 2011 5
Gross national income per 
capita (current USD) $510 2011 5 $1,160 2011 5
GDP growth (annual %) 6.7 2011 5 6.5 2011 5
Gender inequality 
index (0-1) *** 0.58 2011 4 0.63 2011 4
Transparency International 
corruption perceptions 
index (0-100) ****
29.0 2012 6 37.0 2012 6
Health
HIV prevalence rate 
among adults 15-49 7.2% 2011 1 12.5% 2011 1
Malaria deaths (annual 
number of reported 
deaths due to malaria)
8,431 2010 1 4,834 2010 1
Incidence of tuberculosis 
(estimated number of new 
pulmonary, smear positive, and 
extra-pulmonary tuberculosis 
cases per 100,000 people)
193 2011 1 444 2011 1
Annual number of maternal 
deaths (per 100,000 live births) 4,700 2010 7 2,600 2010 7
Maternal mortality ratio 
(MMR) (maternal deaths 
per 100,000 live births)
310 2010 1 440 2010 1
Annual number of 
neonatal deaths (deaths 
under 28 days of life)
43,000 2011 1 17,000 2011 1
Neonatal mortality rate 
(deaths under 28 days of 
life per 1,000 live births)
28 2011 1 27 2011 1
Infant mortality rate (deaths 
of infants under 1 year old 
per 1,000 live births)
58 2011 1 53 2011 1
Under five mortality 
rate (U5MR) (deaths of 
children under 5 years old 
per 1,000 live births)
90 2011 1 83 2011 1
Progress on MDG 4 – U5MR Insufficient 
progress
2012 8 Insufficient 
progress
2012 8
Progress on MDG 5 – MMR Making 
progress
2012 8 Insufficient 
progress
2012 8
APPENDIX A
National 
health and 
health system 
indicators
Saving Mothers, Giving Life External Evaluation / Interim Report 63
Sources
1 WHO Global Health Observatory
Data Repository
2 United Nations, Department of
Economic and Social Affairs – World
Population Prospects, the 2010 Revision
3 World Bank PovcalNet
4 UNDP International Human Development
Indicators
5 World Bank Databank , 2011
6 Transparency International
Coruption Perceptions Index 2012
7 Trends in Maternal Mortality:
1990 to 2010. UNFPA, WHO, UNICEF and
World Bank Estimates
8 Countdown to 2015
9 Global Health Workforce Alliance, 2010
10 DHS
Uganda Year
SOURCE
Zambia Year
SOURCE
Health System
Number of physicians 3,361 2005 1 649 2006 1
Number of nurses & midwives 37,625 2005 1 8,369 2006 1
Health workers per population 
(density of midwives/
nurses/doctors per 1,000)
1.4 2005 1 0.8 2006 1
Hospital beds per 
10,000 population 5.0 2010 1 20.0 2010 1
Health worker vacancy rate (%) 48.0 2010 9 61.0 2010 9
Per capita total expenditure 
on health (USD) $47 2010 1 $73 2010 1
Per capita total expenditure on 
health (PPP international $) $124 2010 1 $90 2010 1
Per capita government 
expenditure on health (USD) $10 2010 1 $44 2010 1
Per capita government 
expenditure on health 
(PPP international $)
$27 2010 1 $54 2010 1
General government 
expenditure on health as % of 
total expenditure on health
22% 2010 1 60% 2010 1
Out of pocket expenditure 
as a % of total 
expenditure on health
50% 2010 1 27% 2010 1
Private prepaid insurance 
plans as a % of total 
expenditure on health
<1% 2010 1 1% 2010 1
External resources for 
health as a % of total 
expenditures for health
26% 2010 1 39% 2010 1
General government 
expenditure on health as % of 
total government expenditures
12% 2010 1 16% 2010 1
Total expenditure on 
health as a % of GDP 9% 2010 1 6% 2011 4
Health Care Utilization
Percentage of live births 
in health facilities 57% 2011 11 48% 2007 11
Percentage of births attendant 
by skilled birth attendant 57% 2011 11 47% 2007 11
Women receiving 4+ 
antenatal care visits 48% 2011 11 60% 2007 11
Percentage of deliveries 
by Caesarean section 5% 2011 11 3% 2007 11
ART coverage among people 
with advanced HIV infection 54% 2011 1 82% 2011 1
PMTCT coverage (maternal 
ARV coverage) 50% 2011 1 86% 2011 1
Percentage of children 
under 5 sleeping under 
bednet previous night
43% 2011 11 33% 2007 11
DTP3 coverage 82% 2010 1 81% 2010 1
Contraceptive prevalence 
rate (any method) 24% 2011 11 30% 2007 11
Contraceptive prevalence 
rate (modern) 21% 2011 11 25% 2007 11
* The total fertility rate is the number of children 
that would be born to a woman if she were to live to 
the end of her reproductive life, and if she were to 
experience the current age-specific fertility rates for 
the duration of her life.
** The human development index (HDI) is a composite 
of life expectancy, educational attainment, and income. 
It ranges from a minimum of 0 (equal) to a maximum 
of 1 (unequal).
*** The gender inequality index (GII) combines 
indicators for reproductive health, female 
empowerment, and female labor force participation 
into a single metric that ranges from a minimum of 0 
to a maximum of 1.
**** The corruption perceptions index scores countries 
on a scale from 0 (highly corrupt) to 100 (very clean).
64 MARCH 2013
This external evaluation is a strategic evaluation of 
implementation of Phase 1 of the Saving Mothers, Giving 
Life initiative in Uganda and Zambia.
Conceptual framework
We use a multi-disciplinary approach to evaluate the imple￾mentation of Saving Mothers, Giving Life. Our conceptual 
framework combines elements from (a) the process eval￾uation framework of Linnan and Steckler and (b) complex 
adaptive systems analysis.40, 41 The key components assessed 
in the evaluation are: context, program theory, implementa￾tion process, dose delivered, community response, fidelity, 
and emergent properties. These are described below:
■■ Context: baseline context and recent changes in social, 
political, and economic environment that may influence 
implementation of SMGL. 
■■ Program theory: the “active ingredients” of the program 
and how these will deliver desired change. 
■■ Implementation process: actors involved in program 
implementation; planning and implementation timelines; 
funding; obstacles/facilitators of implementation.
■■ Dose delivered: measure of intervention activities deliv￾ered to the field (e.g. supplies received, trainings given, 
etc.) in a given timeframe (i.e. 6, 12 months).
■■ Community response: perceptions of SMGL interventions 
by women in the community (i.e. users and non-users of 
the health system), community leaders, health workers; 
women’s satisfaction with SMGL-supported delivery care.
■■ Fidelity: this aspect focuses on the quality and integrity 
of the partnership model (extent of partner engagement; 
participation of stakeholders, e.g. MoH, health workers, 
and the private sector) and the content of the intervention 
(e.g. is training leading to improved obstetric knowledge 
and improved perceived quality by women?).
■■ Emergent properties: dynamic evolution of the inter￾vention, real-life complications, and unexpected results: 
unintended consequences (positive and negative); syner￾gies and extent of integration with local health systems; 
innovations and mechanisms of action.
This interim report focuses on a subset of these constructs 
indicated by the previous bullets in bold text: context, 
program theory, implementation process, dose delivered, and 
some emergent properties. The findings will inform recom￾mendations for Phase 2. The final report, to be completed in 
fall of 2013, will address: 1) response of the community and 
providers 2) extent of implementation at endline and 3) func￾tioning of the SGML partnership.
Setting 
As the focus of the interim report was on country activities, 
data collection occurred in Uganda and Zambia. Our teams 
conducted data collection in Kampala and Lusaka, primar￾ily with officials from the CDC, USAID, Ministry of Health, 
SMGL implementing partners and key stakeholders, such 
as the WHO, Central Statistical Offices, and UNICEF, among 
others. The teams also visited all eight SMGL districts: 
Lundazi, Nyimba, Kalomo and Mansa in Zambia and Kibaale, 
Kyenjonjo, Kamwenge and Kabarole in Uganda.
Instruments
We used a mixed methods design, which included both 
qualitative and quantitative data. We conducted in-depth 
interviews with key SMGL partners at central (i.e. CDC, USAID 
and the Ministry of Health) and district levels (i.e. district 
health teams, SMGL Coordinators and implementing part￾ners). The content of the interview guides is described below:
a. Descriptive interview guide: descriptive information 
on the main roles of the different SMGL partners, 
and the implementation process and timeline. 
Interviews typically lasted 30 to 45 minutes.
b. Lessons learned interview guide: this examined the 
successes, challenges, unintended consequences, 
evolution and innovation of SMGL; emerging lessons 
and recommendations for scaling up in Phase 2, 
among other issues. Interviews lasted 45 to120 
minutes.
We assessed the SMGL “dose delivered” by December 1, 
2012 using a survey of 166 questions for Zambia and 178 
questions for Uganda, enumerating all inputs and activities 
falling under the 28 key SMGL activities of the logic model. 
“Dose delivered” refers to the activities and inputs actually 
implemented in the districts by the mid-point of the first year, 
as reported by the district medical officers and implementing 
partners in districts. Data was entered into Excel and verified 
by in-country partners.
APPENDIX B
Methods
Saving Mothers, Giving Life External Evaluation / Interim Report 65
Data collection
The majority of data collection occurred between October 
1 and December 20, 2012. Data collection commenced with 
a preliminary visit to Zambia and Uganda by the principal 
investigators, Drs. Kruk and Galea, who met with key SMGL 
partners and visited one SMGL district in each country. 
The main goals of the preliminary visits were to gain famil￾iarity with the SMGL program in the two SMGL countries, 
share the scope of work of the external evaluation and 
obtain feedback on it. 
Following the preliminary visits, three-person teams 
comprising a faculty member and two dedicated research￾ers visited each country. Data collection in Zambia occurred 
from November 5 to December 20 and in Uganda from 
November 25 to December 15. Two faculty advisors, Dr. 
Rabkin and Dr. Grépin, led data collection efforts in Zambia 
and Uganda, respectively. Evaluation team members were: 
Dana Greeson (MPH) and Daniel Vail (BA) in Zambia and 
Emma Sacks (PhD candidate) and Katherine Austin-Evelyn 
(MSc, MPH) in Uganda. The teams were assisted from New 
York by Program Manager, Tsitsi Masvawure (DPhil).
Our team conducted a total of 143 interviews (67 Uganda, 
76 Zambia) with 152 individuals. Of these, 89 were in-depth, 
lessons-learned interviews (47 Uganda, 42 Zambia). These 
are broken down as follows: 52 interviews conducted in 
the districts (23 Uganda, 29 Zambia) and 37 interviews 
conducted at central level (24 Uganda, 13 Zambia). We 
interviewed all key US Government SMGL leads. We also 
interviewed five senior Ministry of Health officials who were 
centrally involved in SMGL in both countries (2 in Zambia, 
3 in Uganda). We interviewed representatives of 20 imple￾menting partner organizations (12 in Zambia, 8 in Uganda), 
at capital and district level. We met with and interviewed 
all eight SMGL district medical officers in the SMGL focus 
districts. We also interviewed all seven SMGL coordinators 
in both countries (2 in Uganda, 5 in Zambia, including the 
provincial SMGL coordinator). Lastly, the team attended a 
total of 6 SMGL partner meetings in both countries (5 in 
Zambia, 1 in Uganda). 
To obtain the quantitative dose delivered data, we inter￾viewed district health and district medical officers and 
SMGL coordinators in each country and collected data on 
a total of 202 indicators we developed for the 28 SMGL 
activities contained in the logic model. Informants were 
asked about inputs and activities already implemented as 
of December 1, 2012, not those that were planned for future 
implementation. We collected information on 166 SMGL 
activity indicators in Zambia and on 178 activity indicators 
in Uganda. The data for this came from two main sources: 
key informant interviews with all SMGL district coordina￾tors in each country (4 in Zambia, 2 in Uganda) and other 
key implementers in the districts, including district-level 
members of the Ugandan and Zambian Ministries of Health 
(specifically, district medical officers, chief medical officers 
at the district hospitals, district planners, district infor￾mation officers, district human resource officers, district 
equipment officers, and district nursing officers), and 
district-level representatives from the SMGL implementing 
partners. Multiple respondents were interviewed in each 
district, and, wherever possible, interviewers confirmed 
evidence of implementation firsthand. 
After initial data collection was completed, the evaluation 
team examined all documents collected from each country 
to verify information contained in the dose delivered dataset. 
A reliability score was given to each piece of data, depending 
on the number of people who corroborated the information, 
and if we were able to verify the information with docu￾mentation. All dose delivered data were then shared with 
in-country partners at district level for their review and 
correction in January, 2013. We received clarifications and 
corrections from all eight SMGL districts and incorporated 
all suggestions and edits. In addition to the dose delivered 
data, we also obtained geocoded maps on all health facilities 
in the eight SMGL districts and collected quantitative infor￾mation about the planning and implementation activities and 
timeline from the descriptive interviews.
Our teams spent a total of 134 person-days in country 
(86 in Zambia, 48 in Uganda), the majority of which was 
time spent on the ground in the eight SMGL districts. The 
teams visited a total of 50 sites in the districts (26 in Zambia, 
24 in Uganda). Sites visited included district health offices, 
provincial health offices, district hospitals, mothers’ shelters 
and clinics. 
We also collected 536 program, health system and 
national strategy documents from USG and SMGL imple￾menting partners and the MoH at central and district levels. 
Documents collected included policies, program plans, task 
force meeting minutes and monthly updates on SMGL activ￾ities prepared by the SMGL coordinators in each district and 
reports prepared by the various SMGL implementing part￾ners. District utilization and other health system statistics 
were collected to inform the context section of the report.
Data collection in country was complemented by a review 
of the current literature on maternal and newborn health in 
Zambia and Uganda, and a review of the the Demographic 
Health Surveys (DHS) data for each country. 
The data collected for the interim report were exempted 
from IRB by Columbia University as they concerned program 
implementation activities.
66 MARCH 2013
Data management and analysis
Qualitative data were entered, cleaned, and analyzed using 
the NVivo Software Package (Version 10). Transcripts of 
all interviews and notes were independently coded by two 
analysts using pre-existing codes and a process of constant 
comparison as described by Glaser.2
 Discrepancies were 
reconciled through a practice of consensus-building between 
the two coders. A second level of open, iterative coding was 
then conducted, creating new codes around major themes, 
which were then reviewed by multiple team members to 
synthesize and summarize. These summaries were the 
basis of the content in this report. The summaries also 
provided data for the logic model, partner map, and imple￾mentation calendar.
Dose delivered quantitative data were entered and verified 
against field notes and documents provided and assigned 
a reliability score that indicated our level of confidence in 
the results, with the most reliable being data corroborated 
by two people (including one at the MoH) and backed up by 
documentation. We sought additional input for indicators with 
low reliability scores. A subset of indicators, “tracer indi￾cators,” were identified, which are presented in the report. 
These tracer indicators reflect SMGL core activities across 
the two countries. All dose delivered data were then shared 
with in-country partners at district level for their review 
and correction in January, 2013. All districts responded with 
useful clarifications and corrections. 
The descriptive interviews with lead USG and implement￾ing partners in each country were reviewed individually to 
extract information about SMGL partners, and the timeline 
and key activities in the implementation process. These were 
shared with countries for their review ahead of publication.
APPENDIX B METHODS
Saving Mothers, Giving Life External Evaluation / Interim Report 67
This project was funded by a grant from Merck for Mothers 
to Drs. Margaret E. Kruk and Sandro Galea at Columbia 
University.
We would like to acknowledge the important contributions 
made by various individuals and institutions to the produc￾tion of this report. 
Global SMGL Partners
We would like to thank the SMGL Global Secretariat 
and members of the SMGL Monitoring and Evaluation 
Committee, including Ms. Celina Schocken (Merck for 
Mothers), Ms. Angeli Achrekar, Ms. Mary Ellen Stanton, 
Dr. Claudia Conlon, Ms. Annie Schwartz, Ms. Karen Fogg, 
Ms. Sandhya Rao, and Ms. Marta Levitt (USAID), Dr. Howard 
Goldberg, Dr. Mary Goodwin, and Mr. Paul Stupp (CDC) for 
their assistance in the evaluation. We would like to thank 
Ms. Julie Becker and Ms. Sarah Hank (Rabin Martin) for 
their assistance in the project’s formative phase.
We would like to acknowledge the support of the SMGL 
Leadership Council, including Ms. Lois Quam (US Global 
Health Initiative), Mr. Robert Clay (USAID), Dr. Naveen Rao 
and Dr. Priya Agrawal (Merck for Mothers), Dr. Herbert 
Peterson (ACOG), Ms. Christy Turlington Burns and Ms. Erin 
Thornton (Every Mother Counts), and Dr. Tore Godal and 
Ms. Helga Fogstad (Government of Norway) 
National Ministries of Health
The participation of the Ministries of Health of Uganda and 
Zambia was critically important to our project. In Uganda, 
we extend our gratitude to Dr. Miriam Sentongo, Senior 
Medical Officer, Reproductive Health Office, Dr. Anthony 
Mugasa, Reproductive Health Advisor, MoH, and Dr. Jennifer 
Wanyana, former Assistant Commissioner of Reproductive 
Health, MoH. In Zambia, we extend our thanks to Dr. Max 
Bweupe, Deputy Director of Public Health and Research, 
MoH, and Dr. Caroline Phiri, Acting Director of Maternal and 
Child Health at the Ministry of Community Development, 
Mother and Child Health (MCDMCH). Our conversations with 
them yielded rich insights into the SMGL planning and imple￾mentation process. 
District Health Officials
We extend very special thanks to the eight SMGL district 
health teams for their central assistance in obtaining dose 
delivered data and their reflections on program imple￾mentation. We are indebted to the four District Medical/
Health Officers in Uganda, Dr. William Mucunguzi (Kyenjojo), 
Dr. Richard Mugahi (Kabarole), Ms. Winnie Rurangaranga 
(Kamwenge), and Dr. Dan Kyamanywa (Kibaale). In Zambia, 
we received assistance from Dr. Christopher Nyambe 
Sinyange (Nyimba), Dr. Mutinta Mudenda (Mansa), Dr. Allan 
Chisenga (Lundazi), and Dr. Kenneth Chibwe (Kalomo). We 
also thank Dr. Kennedy Malama, the Provincial Medical 
Officer of Eastern Province and Dr. Elicho Bwalya, the 
Provincial Medical Officer of Luapula Province. 
The DMOs and DHOs in both countries also facilitated 
our access to local leaders, including Mr. Nathan Musingusi, 
the Secretary for Health in Kyenjojo and Sister Harriet 
Kembabazi (representative for the Secretary for Health 
in Kamwenge). 
We would also like to recognize the assistance of other 
health officials in the districts, such as the district planners, 
chief medical officers, district nursing officers and medical 
superintendents, among others. In Uganda, they include 
Dr. Fred Omio (Ntara Health Centre), Sister Regina Mary 
Nyaga (Regional Referral Hospital in Fort Portal), Ms. Joanne 
Chesamg (Mentor Midwife in Kyenjojo), Ms. Zidah Najjuma 
(Mentor Midwife in Kibaale), Ms. Grace Akot (Mentor Midwife), 
and Dr. Charles Olaro (Medical Superintendent of Fort Portal 
Regional Referral Hospital). 
In Zambia, we extend our thanks to Ms. Christina Banda 
(Lundazi Director of MCDMCH), Mr. Joseph Sakala (Kalomo 
District Planner), Mr. Haswell Kapomba (Nyimba Chief 
Clinical Officer and former Acting Planner), Dr. Jonathan 
Chama (Chief Medical Officer, Nyimba District Hospital), 
Ms. Ireen Chiama, (Kalomo MNCH Coordinator), Ms. Agness 
Phiri (Lundazi MNCH Coordinator), Ms. Beatrice Ling’umbwa 
(Lundazi District Hospital Enrolled Midwife), Ms. Bether 
Muzyamba (Acting Chief Nursing Officer of Zimba Mission 
Hospital), Mr. Hastings Chileshe (Acting Nursing Officer at 
Nyimba Hospital), Ms. Numbi (Nyimba HR Officer) and Mr. 
William Mangimela (Chipata Information Officer), Ms. Betty 
Kunda (Mansa District Nursing Officer), Mr. Muleya Siakakole 
(Lundazi HR Officer). We would also like to acknowledge 
Lieutenant Colonel Joyce Mulilo (Zambia Defence Force).
We are especially grateful to the following individuals 
in all four districts in Uganda for swiftly responding to our 
requests for information. Mr. Abel Katuramu generously 
obtained the geocodes of all thirty-seven health facilities 
in his district for us. Mr. Tumuhairwe Collins, Mr. Justine 
Mugisa, Ms. Hope Katushabe, and Mr. Robert Mugabe also 
provided key data.
US Government Agencies
We would like to extend our sincere gratitude to the U.S. 
Centers for Disease Control and Prevention (CDC) and the 
United States Agency for International Development (USAID) 
in the two countries for their essential help in organizing 
data collection and introductions to key stakeholders. We 
appreciate their willingness to share their insights and 
APPENDIX C
Acknowledgements
68 MARCH 2013
reflections about the SMGL program and for inviting us to 
observe a range of SMGL meetings. 
In Uganda, we thank Dr. Frank Kaharuza (Chief, 
Epidemiology Branch, CDC), Dr. Jayne Tusiime-Byakikia (CDC 
Team Leader, Operations Research,), Mr. Prosper Behumbiize 
(Data Manager, CDC) and Mr. George Acire (GIS Analyst, CDC). 
We also thank Ms. Karen Klimowski (USAID Director, Health, 
HIV and Education) and Dr. Janex Kabarangira (Deputy 
Health Team Leader, USAID), and Dr. James Tanu Duworko 
(Reproductive Health Manager, USAID). 
In Zambia, we thank Dr. Lawrence Marum (CDC Country 
Director), Dr. James McAuley (Medical Officer, CDC, 
SMGL Focal Point), Ms. Fatma Soud (CDC, Monitoring and 
Evaluation Lead), and Dr. Jorge Velasco (USAID Deputy 
Health Team Leader). 
We extend our gratitude to the Peace Corps. In Uganda, we 
thank Ms. Loucine Hayes (Country Director), Ms. Bernadine 
Nganda (Coordinator), Ms. Cotious Tukashaba (Country 
Manager), Ms. Shiphrah Mutungi (Program Manager, Health), 
and Mr. Gordon Parola (Volunteer, Kamwenge). In Zambia, 
we thank Ms. Lauren Mamane (Director of Programing and 
Training) and Ms. Sally-Rose Mwachilenga (Peace Corps 
Response and Extension Coordinator). We also thank Peace 
Corps Response volunteers: Ms. Carrie Nichols (Mansa), Ms. 
Hazel Domanque (Kalomo), and Ms. Karen Janssen (Kalomo), 
with a special thanks to Ms. Carrie Vogelsang (Lundazi) and 
Ms. Ashley Arnold (Nyimba), who were invaluable in assisting 
us during and after district data collection in Zambia.
We thank Ms. Patricia Ulaya (Department of Defense 
PEPFAR Program Assistant), Ms. Yvonne Mulenga (Project 
Concern International Zambia Department of Defense HIV/
AIDS Program Manager), and Ms. Mary Mfula (Department 
of Defense Finance Department). 
SMGL Coordinators
We extend sincere thanks to the SMGL coordinators for 
helping facilitate our district visits and for graciously 
responding to email requests for further information 
after our visits. In Uganda, we thank Dr. Joseph Lubwama 
(Program Coordinator for Baylor-Uganda Regional Office) 
and Mr. Edward Kyagulanyi (IDI-Uganda). In Zambia, we 
extend special thanks to Ms. Rebecca Lusumpa (Eastern 
Province SMGL Coordinator), and district coordinators: Mr. 
Jonathan Kunyi (Nyimba), Ms. Esnart Juunza (Kalomo), Ms. 
Cathrine Konie (Lundazi), and Ms. Mutinta Maingaila (Mansa). 
SMGL Implementing Partners
We thank the SMGL implementing partners for their 
enthusiasm and cordiality. On numerous occasions, 
partners devoted their entire mornings or afternoons to 
talk to us and gave us access to key staff members and 
planning documents.
In Uganda we are grateful to the following individuals at 
Baylor-Uganda: Dr. Adeodata Kekintiiwa (Executive Director), 
Ms. Patricia Mwebaze (SMGL Program Manager), Dr. George 
Akol (SMGL Program Manager), Ms. Barbara Sentiba 
(Human Resources Director), Dr. Robert Iriso (Medical 
and Psychosocial Programs Director), Dr. Dan Murokora 
(Technical Advisor), Mr. Paul Mayende (Demand Creation), 
Mr. Rogers Sekabira (Logistics Management), Mr. Joseph 
Mukasa (Vouchers Grants Manager), Ms. Alice Asiimwe 
(Researcher, Community Monitoring and Evaluation), Mr. 
Sam Wasike (Monitoring and Evaluation Officer), Mr. Nelson 
Businge (Community Linkage Officer), and Mr. Thomas 
Kulumba (Community Linkage Office). We would also like 
to extend our gratitude to the following individuals at 
IDI-Uganda: Dr. Alex Coutinho (Executive Director), Ms. 
Brenda Picho (SMGL Director), Ms. Grace Oling (Community 
Linkage Officer), and Ms. Zida Najjuma (mentor midwife). 
At STRIDES, we are grateful to the following: Ms. Cecilia 
Kakande (Chief of Party), Dr. Henry Kakande (Deputy Chief 
of Party), Ms. Rita Lulua (Regional Coordinator), Ms. Manjit 
Kaur (Regional Operations Coordinator), Dr. Miriam Mutabazi 
(Technical Director), and Mr. Richard Ssewajje (Director of 
Grants/Contracts). We would also like to thank the following 
individuals from AOGU: Dr. Romano Byaruhanga (President), 
Dr. Jolly Beyeza, Dr. Eve Nakabembe, Prof. Florence Mirembe, 
Ms. Juliet Oinomugisho and Dr. Imelda Namagembe; Ms. 
Harriet Nambanya from UPA and the following individuals 
from IPA: Ms. Jessica Kiessel (Director of Country Offices), 
and Mr. Dan Pakerson (Country Director). We extend our 
gratitude to the private sector partners we met with, namely 
Dr. David Mwinganiza (Director, Midas Touch) and Dr. Priscilla 
Busingye (Director, Virika Hospital). At Marie Stopes Uganda
we are grateful to Mr. Jon Cooper (Country Director), Ms. 
Christine Namayanja (Director, Private Sector Partnerships), 
Ms. Julia Mayersohn (Public Relations), and Mr. William 
Nyombi (Senior Manager of Franchising). 
In Zambia, we sincerely thank the following individuals 
at CIDRZ: Dr. Carla Chibwesa (Gynaecologist), Ms. Mari 
Katundu (Program Manager), Mr. Jameson Mubonde (Nyimba 
AMAI Coordinator), and Mr. Simon Mutibo (Lundazi AMAI 
Coordinator). At ZCAHRD, we sincerely thank Dr. Davidson 
Hamer (Director of Research and Evaluation), Ms. Irene 
Singogo (SMGL Project Director), Ms. Martha Musweu (Senior 
Clinical Mentor), Ms. Shari Davis (Logistics Coordinator), and 
Dr. Don Thea (Principal Investigator). At MCHIP, we thank 
Mr. Kwame Asiedu (Country Representative), Ms. Michelle 
Wallon (Project Director), Ms. Martha Ndhlovu (Technical 
Advisor), and Ms. Brenda Mutinta (MNH Technical Officer). 
At ZISSP, we thank Ms. Kathleen Poer (Chief of Party), 
Ms. Mutinta Nalubamba (MNCH Team Leader), Mr. Mukata 
Mulonda (Eastern Province Community Health Officer), and 
Mr. Timothy Silweya (Luapula Community Health Officer), in 
addition to the SMGL coordinators previously mentioned. 
APPENDIX C ACKNOWLEDGEMENTS
Saving Mothers, Giving Life External Evaluation / Interim Report 69
At EGPAF, we thank Dr. Jack Menke (Technical Director), Ms. 
Ruth Nswana (Strategic Information Officer), and Mr. Bright 
Kulukulu (Strategic Information Officer). At JSI, we thank Mr. 
Walter Proper (Country Director), Mr. Sydney Chanda (Senior 
Public Health Logistics Advisor), and Mr. Derrick Nyimbili 
(Public Health Logistics Advisor). At ZPCT II, we thank 
Dr. Michael Welsh (Country Director), Dr. Thierry Malebe 
(Senior Advisor, PMTCT/CT), Ms. Gail Bryan,Mofya (Senior 
Advisor, Pharmacy and Labs), Ms. Prisca Kasonda (Director 
of Technical Services), and Mr. Suzgo Kapanda (Provincial 
Program Manager). 
Other Stakeholders
In Uganda, we would like to thank Dr. Olive Sentumbwe 
(Reproductive Health Director) at the World Health 
Organization, Mr. Sean Blaschke (Health Systems 
Strengthening Coordinator, UNICEF) and the following indi￾viduals at PACE: Ms. Hanna Baldwin (Technical Advisor) and 
Ms. Dorothy Balaba (Country Manager, Reproductive Health). 
In Zambia, we also thank Mr. Nicholas Shiliya at SFH/PSI, 
Mr. Alexander Mkosi (SMGL M&E) at CSH, Mr. Paul Pascal 
Chimedza (Country Director) and Mr. Elizabeth Simwawa—
both at Africare, and Kaiko Mkilolo, Results Based Financing 
Technical Specialist at the World Bank.
APPENDIX C ACKNOWLEDGEMENTS
70 MARCH 2013
1. SMGL. Saving Mothers, Giving Life. 2012; http://www.
savingmothersgivinglife.org. Accessed January 28, 2013.
2. Glaser B. The Constant Comparative Method of Qualitative 
Analysis. Social Problems. 1965;12(4):436-445.
3. Lozano R, Wang H, Foreman KJ, et al. Progress towards 
Millennium Development Goals 4 and 5 on maternal and 
child mortality: an updated systematic analysis. Lancet. 
2011;378(9797):1139-1165.
4. World Health Organization (WHO). Trends in Maternal Mortality 
1990-2010: WHO, UNICEF, UNFPA and The World Bank Estimates.
Geneva: World Health Organization;2012.
5. 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(9832):2151-2161.
6. Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO 
analysis of causes of maternal death: a systematic review. 
Lancet. 2006;367(9516):1066-1074.
7. Ronsmans C, Graham WJ, Lancet Maternal Survival Series 
steering group. Maternal mortality: who, when, where, and why. 
Lancet. Sep 30 2006;368(9542):1189-1200.
8. World Development Indicators database. 2013. http://databank.
worldbank.org/ddp. Accessed February 19, 2013.
9. Global Health Initiative (GHI). Saving Mothers, Giving Life Uganda: 
Operational Plan. January 2012 – February 2013. Washington DC: 
U.S. Global Health Initiative;2011.
10. World Health Organization (WHO). World Health Organization 
Global Health Observatory Data Repository. http://apps.who.
int/gho/data/. Accessed January 29, 2013.
11. World Health Organization (WHO). World Health Report 
2006: Working together for health. Geneva: World Health 
Organization;2006.
12. Uganda Bureau of Statistics (UBOS), ICF International Inc. 
Uganda Demographic and Health Survey 2011. Kampala, Uganda: 
UBOS and Calverton, Maryland: ICF International Inc.: Uganda 
Bureau of Statistics;2012.
13. Green E, Nantulya V, Stoneburner RL, Stover J. What Happened 
in Uganda? Declining HIV Prevalence, Behavior Change, and the 
National Response Washington DC: United States Agency for 
International Development;2002.
14. Stoneburner RL, Low-Beer D. Population-level HIV 
declines and behavioral risk avoidance in Uganda. Science. 
2004;304(5671):714-718.
15. Government of Uganda (GOU). Roadmap for Accelerating the 
Reduction of Maternal and Neonatal Mortality and Morbidity in 
Uganda 2007-2015. Kampala: Government of Uganda, Ministry 
of Health;2007.
16. Central Statistics Office (CSO), Ministry of Health (MoH), Tropical 
Diseases Research Centre (TDRC), Univeristy of Zambia, Macro 
International Inc. Zambia Demographic and Health Survey 2007.
Calverton, Maryland, USA: CSO and Macro International Inc.: 
Zambia Central Statistics Office;2009.
17. Joint United Nations Programme on HIV/AIDS (UNAIDS). 
Zambia: HIV and AIDS Estimates 2011. 2011; http://www.
unaids.org/en/regionscountries/countries/zambia/. Accessed 
January 29, 2013.
18. Fylkesnes K, Musonda RM, Sichone M, Ndhlovu Z, Tembo F, 
Monze M. Declining HIV prevalence and risk behaviours in 
Zambia: evidence from surveillance and population-based 
surveys. Aids. 2001;15(7):907-916.
19. Government of the Republic of Zambia (GRZ). Road Map for 
Accelerating the Attainment of the Millennium Development Goals 
Related to Maternal, Newborn and Child health in Zambia. Lusaka: 
Government of Zambia;2007.
20. Global Health Initiative (GHI). Saving Mothers’ Lives Zambia: 
Operational Plan. October 2011-December 2012. Washington DC: 
U.S. Global Health Initiative;2011.
21. World Health Organization (WHO). Maternal, newborn, child 
and adolescent health; Newborn health epidemiology. 2013; 
http://www.who.int/maternal_child_adolescent/epidemiology/
newborn/en/index.html. Accessed February 1, 2013.
22. Lawn JE, Cousens S, Zupan J. 4 million neonatal deaths: When? 
Where? Why? Lancet. 2005;365(9462):891-900.
APPENDIX D
References
Saving Mothers, Giving Life External Evaluation / Interim Report 71
23. Ahmed S, Li Q, Liu L, Tsui AO. Maternal deaths averted by 
contraceptive use: an analysis of 172 countries. Lancet. 
2012;380(9837):111-125.
24. Freedman L, Waldman R, de Pinho H, et al. Who’s got the power? 
Transforming health systems for women and children. Task Force 
on Child Health and Maternal Health. New York: UN Millennium 
Project;2005.
25. Manandhar D, Osrin D, Shreshtha B, et al. Effect of a 
participatory intervention with women’s groups on birth 
outcomes in Nepal: cluster-randomized controlled trial. Lancet. 
2004;364(9438):970-979.
26. Mullany BC, Becker S, Hindin MJ. The impact of including 
husbands in antenatal health education services on maternal 
health practices in urban Nepal: results from a randomized 
controlled trial. Health Education Research. 2007;22(2):166-176.
27. India Ministry of Health and Family Welfare. Janani Suraksha 
Yojana: Guidelines for Implementation. New Delhi: Government of 
India;2009.
28. Lim SS, Dandona L, Hoisington JA, James SL, Hogan MC, 
Gakidou E. India’s Janani Suraksha Yojana, a conditional cash 
transfer programme to increase births in health facilities: an 
impact evaluation. Lancet. Jun 5 2010;375(9730):2009-2023.
29. Uxa F, Bacci A, Mangiaterra V, Chiaffoni GP. Essential newborn 
care training activities: 8 years of experience in Eastern 
European, Caucasian and Central Asian countries. Seminars in 
Fetal and Neonatal Medicine. 2006;11(1):58-64.
30. Santos C, Diante D, Jr., Baptista A, Matediane E, Bique C, Bailey 
P. Improving emergency obstetric care in Mozambique: the 
story of Sofala. International Journal of Gynecology & Obstetrics. 
2006;94(2):190-201.
31. Koblinsky M, ed Reducing maternal mortality: learning from 
Bolivia, China, Egypt, Honduras, Indonesia, Jamaica and 
Zimbabwe. Washington, DC: The World Bank; 2003.
32. Campbell O, Gipson R, Issa AH, et al. National maternal 
mortality ratio in Egypt halved between 1992–93 and 2002. 
Bulletin of the World Health Organization. 2005;83(6):462-471.
33. Hogan MC, Foreman KJ, Naghavi M, et al. Maternal mortality 
for 181 countries, 1980-2008: a systematic analysis of 
progress towards Millennium Development Goal 5. Lancet. 
2010;375:1609-1623.
34. United Nations. Millennium Development Goals: 2012 Progress 
Chart. 2012; http://www.un.org/millenniumgoals/pdf/2012_
Progress_E.pdf. Accessed February 22, 2013.
35. Chen L, Evans T, Anand S, et al. Human resources for 
health: overcoming the crisis. Lancet. Nov 27-Dec 3 
2004;364(9449):1984-1990.
36. Kruk ME, Johnson JC, Gyakobo M, et al. Rural practice 
preferences among medical students in Ghana: a 
discrete choice experiment. Bull World Health Organ. May 
2010;88(5):333-341.
37. Lehmann U, Dieleman M, Martineau T. Staffing remote rural 
areas in middle- and low-income countries: a literature review 
of attraction and retention. BMC Health Serv Res. 2008;8(19):19.
38. Kruk ME, Mbaruku G, McCord CW, Moran M, Rockers PC, 
Galea S. Bypassing primary care facilities for childbirth: a 
population-based study in rural Tanzania. Health Policy Plan. Jul 
2009;24(4):279-288.
39. Joint United Nations Programme on HIV/AIDS (UNAIDS). The 
“Three Ones” in action: where we are and where we go from here.
Geneva2005.
40. Linnan L, Steckler A. Process Evaluation for Public Health: 
Interventions and Research. San Francisco: Jossey-Bass; 2002.
41. de Savigny D, Adam T, eds. Systems thinking for health systems 
strengthening. Geneva: Alliance for Health Policy and Systems 
Research, WHO; 2009.