Predictors of the Rate of MTCT of HIV in Ethiopia

Research Protocol

2011
 Addis Ababa, Ethiopia

January, 2011

Summary

This study will try to uncover how much the ARVs given to HIV positive pregnant women
according to the national guidelines reduces the risk of MTCT, factors that contribute for MTCT
of HIV, and the proportion of the at-risk population that accesses and correctly uses the
intervention. This research will try to test mainly the following two hypotheses: The rate of
MTCT of HIV in the study population is below 5%; and the maternal and health service related
factors that contribute to the MTCT. The study findings will have a big programmatic and policy
implication for the country. This study is planned to be carried out in 2012 in Addis Ababa and
Tigray Regions of Ethiopia- subject to the availability of fund.

2
Table of Content

Contents
Summary........................................................................................................................................................................2
Table of Content ............................................................................................................................................................3
Problem of Statement ....................................................................................................................................................4
Literature review............................................................................................................................................................6
Goal of the study............................................................................................................................................................8
Specific Objectives: ...................................................................................................................................................8
Research Question or hypothesis ...................................................................................................................................8
Methodology..................................................................................................................................................................9
Study Variables and their Definition: .......................................................................................................................... 10
Ethical Considerations ................................................................................................................................................. 11
Expected Outcomes of the Study ................................................................................................................................. 11
Plan to report and disseminate research findings ......................................................................................................... 11
Logistics plan: ............................................................................................................................................................. 12
Work Schedule: ........................................................................................................................................................... 13
Bibliography ................................................................................................................................................................ 14

3
Problem of Statement

The main means of acquiring HIV in babies is through vertical transmission of the virus. Babies
can acquire HIV from their HIV positive mother during pregnancy, delivery or breast feeding
period through breast milk of the mother. Early researches has shown that, without any
intervention, about 70 - 85% of none breast fed babies and 55 – 75% of breast feed babies who
are born to HIV positive mothers can be free from HIV (1). Currently, thanks to the advance in
medical sciences and technology, over 99% of babies are spared from HIV in the United States
and Europe(2).

The WHO four pronged strategy for PMTCT is being implemented by many countries to prevent
the vertical transmission of HIV from the mother to babies during pregnancy, labor and delivery
and breast feeding. The strategy includes, decreasing HIV incidence in women (Prong 1),
reducing unmet need for family planning (Prong 2), providing antiretroviral prophylaxis to
prevent HIV transmission during pregnancy, labor and delivery, and breastfeeding (Prong 3), and
providing care, treatment and support for mothers and their families (Prong 4) (3).

In most African countries and other resource limited settings, either single dose NVP or maternal
ZDV are practiced for PMTCT. Researches shows that in a population of breast feeding
prevalence of 40%, a short course of ZDV followed by intra-partum 3TC and single dose NVP
would decrease the MTCT rate to below 5% (4). Many African countries are in the process of
embracing the 2010 WHO Recommendation for PMTCT. This Recommendation is based on
Highly Effective Antiritroviral Therapy (HEART) for women with CD4 count of 350 or
advanced Clinical Stage, and country-level selection between tow antiretroviral options during
pregnancy and during breast feeding (“Option A” or Option B”) (5).

Nevertheless, in most African countries, many pregnant women are losing these existing
opportunities because of the drop out from service in the PMTCT service cascade after they
present for ANC through HIV testing and receiving HIV test result, to receiving ARVs and
adherence to ARVs and health workers advice during pregnancy, delivery, and breast feeding
period.

In Ethiopia, PMTCT program started in 2001 when the first National PMTCT Guidelines was
issued. The guideline was revised in 2007 and recently another revision was made following the
2010 WHO new recommendation. According to the national report, the adult HIV prevalence for
2010 was estimated to be 2.4% (6). A recent ANC Sentinel study, HIV prevalence in pregnant
women is 3.2%. This figure ranges from 5.2%? in urban areas to 1.9% in rural part of the country
(7). Based on the above evidences, in 2011 alone, ninety thousand of the pregnant women are
assumed to be HIV positive. This year about three million pregnancies are expected nationally.
To offer PMTCT service for these clients there are 1340 health facilities in the country.

4
However, only 10% of HIV positive women received ARVs for PMTCT (8). The major factor
contributing for the low PMTCT service coverage in the country is the low ANC (32%) and
institutional delivery utilizations (10%) and lost to follow up of HIV positive pregnant women to
get ARVs (39%) related to cultural, access, quality, and unfavorable health provider attitude
factors (9).

Impact studies are used to determine: whether an implemented intervention has made a
difference, factors contributing for the effective PMTCT program, the magnitude of change
observed and how the programme can be improved. This study will try to uncover how much the
ARVs given to HIV positive pregnant women according to the national guidelines reduces the
risk of MTCT, factors that contribute for MTCT of HIV, and the proportion of the at-risk
population that accesses and correctly uses the intervention.

5
Literature review

Approximately half of the 33.4 million persons living with the Human Immunodeficiency Virus
(HIV) worldwide are women of reproductive age, and among the 2.1 million HIV-infected
children, virtually all were infected during pregnancy, delivery, or breast-feeding (10). Since
2002, the number of newly infected children has declined. The probable factor can be the
increased implementation of the prevention of mother-to-child transmission of HIV in countries
and the stabilization of HIV prevalence among women in many countries (11). With these
advances, new challenges have surfaced.

UNICEF, WHO, UNAIDS and other stakeholders, had conducted a consultative meeting in
February, 2009 to evaluate the impact of PMTCT Services in Low- and Middle-Income
Countries in Averting New HIV Infections in Children and Improving Child Survival. The goal
of this Consultative meeting was just to reach consensus on the most effective way to develop
and design approaches to assess the effectiveness of national PMTCT programmes that can be
implemented widely in a number of countries. This meeting reviewed different approaches that
can be applied in different set up. These include population-based surveys, health facility-based
surveys, combined facility and population-based surveys, and cohort studies. Each of these
methods has their own merits and demerits (12).

Population based surveys are also used to estimate the impact of PMTCT programs. One good
example is DHS. This approach will come up with more representative population related
information. The major challenge of this approach is the fact that it requires large sample size to
ensure accurate estimate of outcome measures. If both the mother and the baby die, it is very
difficult to establish vertical transmission rates from population based surveys. In addition, this
method is merely practical in countries with high HIV prevalence.

The Spectrum Project Package is a tool used to estimate the impact of the AIDS epidemic at a
country level. The tool is updated to incorporate the recent survey and surveillance findings
combined with demographic data to provide the needed PMTCT outcome indicators needed for
planning purpose. The major limitation of this tool is that many of the assumptions used in the
spectrum are derived from a small number of studies which may not be representative of the
population (13).

Some researchers use programmatic data to estimate the effectiveness of PMTCT programs. For
example a study by Nicole and his colleagues in Kenya has shown that estimates from Program
data are defective to replace surveillance data. In addition, most studies in sub-Saharan African
were not strong enough to show transmission rates with various maternal ARV regimen and CD4
distribution and infant feeding status were either poorly described or overlooked. The other

6
challenge of this approach is that, it is not a direct outcome measure of programme effectiveness
(14, 15).

The other model is the CHAI PMTCT Model. This model is used to estimate the expected
outcomes of PMTCT interventions and explore strategies to minimize infant infections.
Cumulative MTCT rates segregated by drug regimen at six months and monthly their after
segregated by the infants’ feeding pattern. This model assumes that the cumulative transmission
rate at six weeks is a proxy for birth (16). However this assumption doesn’t hold true as per the
Kesho Bora observational cohort data which reports different rates at birth and six weeks (4).

Health facility based surveys which can be retrospective cohort, prospective cohort, or cross-
sectional analysis. Data collected through DBS collected from children attending immunization
clinics in their post partum period, patient files and registers, and follow up of cases can be used
for data collection in this method. This methodology is robust and relatively straight forward to
manage and implement. However, in countries where the EPI coverage is low, this method may
not be representative. The issue of logistics is also a big concern when employing this method at
national level using DBS laboratory tests. Retrospective-cohort of Health Facility Based Surveys
do not explain why mothers and their infants dropped out in the PMTCT cascade. This method
needs to track down infant outcomes or link to infant data which is expensive. Moreover, test to
follow up and the fact that it is facility based are weaknesses of this method (16).

The famous PMTCT Effectiveness in African Research and Linkage to care and Treatment
(PEARL) study. This survey employed population based survey and data was also collected from
forty- three health facilities in Africa (18).

Approaches using any two of the above methods like the PEARL study provide robust process
and out-come indicators which complimentary data on service coverage, data related to the
process in the PMTCT cascade, and also cost effectiveness of programs. This approach also
provides long term outcome indicators like HIV free child survival at two years. The demerits of
combined approaches is, their high cost and its complex evaluation that would be difficult to
replicate in multiple sites and repeatedly over time.

7
Goal of the study
The goal of this study will be to assess the rate and predictors for the MTCT infection among
babies born to HIV positive mothers in Ethiopia.

Specific Objectives:
 To determine the MTCT rate of HIV from HIV positive mothers to their infants;
 To identify the maternal predictors for MTCT of HIV;
 To identify perinatal health service predictors for MTCT of HIV;
 To pinpoint the practice of infant feeding by HIV positive mothers;
 To establish survival of HIV exposed infants at one year of age;

Research Question or hypothesis

This research will try to test mainly the following two hypotheses:

 The rate of MTCT of HIV in the study population is below 5%.

 Maternal and health service related factors do effect the MTCT.

8
Methodology

Study Area and Period: According to the FMOH report, over 1,340 government, private, and
NGO health facilities provide PMTCT services in the country. Both government hospitals and
health centers which started PMTCT program before two years and with adequate PMTCT case
load will be selected from Addis Ababa – the capital city of Ethiopia and Tigray Region. These
two regions are selected for the high HIV prevalence in these regions of the country. Tigray
region is selected to see a better representativeness.

Study design: Retrospective cohort analysis using abstraction of secondary data from ART
Clinic, ANC/PMTCT, Labor and Delivery, and EPI units.

Study population: all babies born to HIV positive mothers in the last one year irrespective of
place of delivery (home or at health facility).

Data collection and management: The source of data for this study are health facility registers,
patient fills, and interview of HIV positive mothers who gave birth in the last one year. First data
collection tool will be developed for data collection from facility registers and patient fills and
for the interview. Then training manual and trainer guide will be developed for data collectors.
Training will be given to data collectors and their supervisors on the data collection tools, ethical
issues, data quality assurance, inter-personal communication skills and how to handle challenges
that may appear in data collection process and team work. After the training the data collectors
will test the data collection tools (from registers and interview) and exercise data collection in
selected health facilities in Addis Ababa. Based on the pretest result and feedback from data
collectors, the data collection tool will be modified in such a way that it will be simple,
understandable, and easy to use for the data collectors.

The supervisors will oversee the data collection process in both the pretesting and actual data
collection time. They also give on-the-spot corrections and feedback for the data collectors to
ensure the quality of the data. They also collect the filled checklists on daily basis and check for
its completeness, consistency, and quality. Then, the data will be double entered into computer
EPI Software to check its internal validity. Descriptive statistical tests including frequencies,
percentages, and cross tabulations will be done. Then the data will be exported to SPSS software
to do analytic statistical tests including logistics and linear regressions to see the relationship of
variables by controlling for confounders. Then, results will be presented in tables and graphs to
answer the study questions.

9
Study Variables and their Definition:

Objective Indicator Definition of Indicator Study approach

To determine the MTCT rate  Proportion of HIV positive babies  HEI tested positive by either DBS Abstraction of
of HIV from HIV positive among babies born to HIV or Ab according to the guidelines; patient files and
mothers to their infants; positive mothers;  Total number babies born to HIV registers
 Number of HIV positive babies in positive mothers in the study
the study period (Nx) period; Interview with
 Number of babies born to HIV parents of HIV
positive mother in the study exposed infants
period (Dx);
To identify the maternal  Duration on ARV drug before  ARV for more than two months or Abstraction of
predictors for MTCT of HIV; delivery; less; patient files and
 Adherence on ARV drugs;  Good, Fair, or Bad adherence as registers
 Clinical condition of the mother per the WHO guidelines;
during pregnancy;  WHO Clinical staging or CD4 Interview with
 count parents of HIV
exposed infants
To identify intra perinatal  Mode of delivery;  Spontaneous Vaginal, Assisted Abstraction of
predictors for MTCT of HIV;  The labor condition of the mother; Vaginal (epithotomy, forceps, patient files and
 Place of delivery; Vacuum, ARM), CS, registers
 ARV status of the mother and the  Normal or prolonged
baby;  Material used to cut the cord; Interview with
 Post- delivery care of the baby  Feeding condition of the baby parents of HIV
(EBF, Mixed feeding, Exclusive exposed infants
Complementary Feeding)
 Medical interventions to the baby
(Cord cutting, injection, suctioning)
To establish survival of HEIs  Number of HIV exposed babies  Using the WHO definition of Abstraction of
at one year of age; alive at one year of age; follow up status/adherence the patient files and
 Number of HIV exposed babies follow up status of HEIs for care registers
died before one year of age. and treatment will be classified into
Good “G”, Fair “F”, and Poor “P”. Interview with
parents of HIV
exposed infants

10
Ethical Considerations
Ethical clearance will be first collected from the Ethiopian Federal Ministry of Health, which is
mandated to assure the ethical quality of researches carried out in the country in various fields.
Then written consent letter will be earned from the respective government offices and health
facilities where data will be collected. Then informed consent will be assured from each study
individuals after discussing the purpose of the study, the right to refuse, and assuring
confidentiality for the information they give, before any data collection undergoes.

Expected Outcomes of the Study

 Number of HIV positive babies born to HIV positive mothers;

 List of maternal factors predicting the MTCT of HIV;
 List of intra perinatal predictors for MTCT of HIV;
 Description of infant feeding practice by HIV positive mothers;
 Number of HIV exposed infants alive at one year of age;

Plan to report and disseminate research findings

The research findings will be presented in both national and international HIV/AIDS
conferences. Moreover, there is a plan to publish the findings in peer reviewed international
journal to make policy makers, program people, and researchers get access to the findings.

11
Logistics plan:
S. Activity/Item Number Unit Total
No cost cost in
USD
I Stationery
1. Pen 10 0.20 2.00
2. Pencil 10 0.20 2.00
3. Eraser 10 0.20 2.00
4. Sharpener 10 0.20 2.00
5. Plain paper 4 4.6 18.40
6. Toner 1 100 100.00
7. Photo copy 100 0.1 10.00
8. Flash disk/memory stick 5 52 260.00
9. Bag for enumerators 5 20 100.00
Total stationery cost
I Personnel cost
Perdiem for enumerators 6 20 3,000.00
(25 days including the
training days)
III Transportation cost for 4 40 160.00
enumerators
IV Miscellaneous 100.00
Total Project budget 3,756.4

12
Work Schedule:
S. Activity Time Period
No 2011 2012
January February March April May June
1. Finalize the research
proposal
2. Secure budget
3. Finalize questionnaire
and checklists
4. Earn ethical clearance
5. Training of data
collectors
6. Pilot study and revision
of the instrument
7. Data collection
8. Data entry and data
cleaning
9. Data analysis and
produce draft research
paper
10. Share the research
paper for input
11. Incorporate comments
andfinalize the research
project
12. Submit the research
paper for publication in
peer reviewed journal

13
Bibliography

1. De Cock KM et al. Prevention of Mother-to-Child HIV Transmission in Resource-Poor

Countries:
Translating Research into Policy and Practice. Journal of the American Medical Association.
2006.

2. Tubiana R, Le Chenade j, Rouzioux C, et.al. Factors associated with Mother-to-Child

Transmission of HIV-1 Despite a Maternal Viral Load < 500 Copies/ml at Delivery: A Case-
Control Study Nested in the French Perinatal Cohort (EPF-ANRS CO1). 2010. Clin Infect
Dis 50:585 – 596..
3. WHO. Antiretroviral Drugs for Treating Pregnant Women and Preventing HIV Infection in
Infants: Towards Universal Access: Recommendations for a Public Health Approach. 2006.
4. Kesho Bora Study Group. Eighteen-Month Follow-up of HIV-1 Infected Mothers and their
Children Enrolled in the Kesho Bora Study Observational Cohorts. 2010.J.Acquir Immune
Defic Syndr. 54:533-541.
5. WHO. Antiretroviral Drugs for Treating Pregnant Women and Preventing HIV Infection in
Infants: Recommendations for a Public Health Approach. 2010.
Available:http://whqlibdoc.who.int/publications/2010/9789241599818_eng.pdf.Accessed on
Jan 2012.
6. MOH. Single Point HIV Prevalence Estimate. June 2007.
7. Federal Ministry of Health. Report on the 2007 Round Antenatal Care Sentinel HIV
Surveillance in Ethiopia; June, 2010. Addis Ababa.
8. Federal Ministry of Health. 2010/2011 Annual Review Meeting Report. 2011.

9. FDRE - Central Statistic Authority. Ethiopia Demographic Health Survey. Preliminary

Report. 2011. Preliminary Report.
10. AIDS epidemic update: November 2009. Geneva: World Health Organization, United
Nations Programme on HIV/AIDS. Accessed on Dec. 2011
(http://data.unaids.org/pub/Report/2009/JC1700_Epi_Update_2009_en.pdf.)
11. WHO. Global Health Sector Strategy on HIV/AIDS 2011 – 2015. 2011.
12. UNICEF, WHO, UNAIDS, Vanderbilt School of Medicine. Consultative Meeting on
Evaluating the Impact of Prevention of Mother-to-Child Transmission of HIV (PMTCT)
Services in Low-and Middle-Income Countries in Averting New HIV Infections in Children
and Improving Child Survival. 12 -13 February, 2009. Tennessee, USA.
13. J. Stover, P. Johnson, B Zaba, et.al. The Spectrum Projection Package: Improvement in
Estimating Mortality, ART Needs, PMTCT Impact and Uncertainty Bounds.

14
14. Nicole S., Wolfgang H., Esther M. et.al. Can Data from Programs for the Prevention of
Mother-to-Child Transmission of HIV be Used for Surveillance in Kenya? Public Health
Reports / November–December 2006 / Volume 121. Pages 695 – 702.
15. Carter R. Estimates of MTCT rates from Programmatic data (Unpublished). Sept, 2010.
16. McCarthy E. Transmission Data Being Used for PMTCT Models, CHAI Model.
Unpublished. Sept, 2010.
17. Hayashi C. Clinical Transmission Rates to National Programme Transmission: Issues to
Consider and Data Gaps (Unpublished). Sept. 2010.
18. Elizabeth S, Didier E, David C, et.al. Coverage of Nevirapine-Based Services to Prevent
Mother-to-Child HIV Transmission in 4 African Countries. JAMA. 2010;304(3):293-302

15