PHIA Webinar Slides FINAL

HIV Impact Assessments (PHIA):
An Introduction Focusing on
Malawi and Zimbabwe
Elizabeth Radin, PhD
Technical Specialist – Population-based Surveys
Project Director, Malawi & Zimbabwe PHIAs
November 6, 2014
Presentation Objectives
1. Explain what ICAP’s PHIAs are
2. Explain why ICAP is doing PHIAs from both a
SCIENTIFIC and POLICY perspective
3. Present how we are approaching the first two
PHIAs in Malawi and Zimbabwe
4. Share information on the future of PHIAs at
What are ICAP’s PHIAs
Population-based HIV Impact Assessment
A survey that is:
• Nationally-led (MOH, NSO)
• In collaboration with CDC
• Cross-sectional
• Household-based
• Nationally and Sub-nationally Representative
• Focused on impact-level indicators of the HIV
epidemic through biomarkers and self report
What we mean by ‘Impact Assessment’
A Description of Impacts . . .
IMPACT: long term, high-level result
(reduced transmission, reduced mortality)
OUTCOME: medium-term result
(# tested, # on ART)
OUTPUT: product of activities
(# of health staff trained)
What does a PHIA Assess?
• The currents status of the epidemic in a
• The access to and uptake of HIV care and
treatment services
Scientific Rationale for PHIAs
Health facilities, and health facility access exists in a spectrum . . .
ART Site
Summary: Scientific Rationale
Current data is largely facility-based
Facility-based data describes a subset of the
It is difficult to infer population measures– such
as prevalence or incidence – from facility based
Population-based surveys are the gold standard
for these indicators
Policy Rationale for PHIAs
Adults and children with HIV infection receiving ART
with PEPFAR support, 2004-2011
Number on ART
4 million
El-Sadr WM et al 2012
Policy Rationale for PHIAs
After more than a decade of PEPFAR what is
that status of the epidemic?
For Example:
• What is the rate of new infection following
prevention efforts?
• What is the proportion of Viral Load
Suppression following expanded ART
Sampling for the PHIAs
A Two-stage Cluster-based Sampling Strategy:
An example: What is the prevalence of coffee drinking at ICAP?
Background: ICAP has 1000 staff, 20 Offices, 50 Staff/Office
A Census: Ask all 1000 ICAP staff if they drink coffee
A Simple Random Sample: Select 10, 100 or 500 ICAP staff and ask if
they drink coffee
A Cluster-based Sample: Select 5 ICAP country offices, ask all staff in
those offices if they drink coffee
A Two-stage Cluster-based Sample: Select 10 ICAP country offices,
randomly select 25 people from each country office
Sampling for the PHIAs
Sampling Strategy:
• Using two-stage cluster-based sampling strategy
– Sample ~500 Enumeration Areas (EA),stratified by health zone
– Sample ~30 Households per EA
• Sample size includes ~15,000 HH; ~30,000 individuals
• Adults from every HH ~20,000, all children every other HH ~10,000
The DHS: A Pop Survey Celebrity
Similarities between DHS/PHIA
Population-based household survey
Cross-sectional, nationally representative
Household and individual questionnaires
National and subnational HIV prevalence
• Household and individual sample size similar
Differences between DHS & PHIA
• Biomarkers for CD4 counts, viral load, recency,
drug resistance, ARV metabolites, peds
• Point-of-Care HIV testing and CD4 testing
with return of results
• Opportunity to assess global HIV outcomes of
interest that are outside domain of DHS
– Potential for Treatment as Prevention
ICAP Experience with Pop Surveys
• Swaziland HIV Incidence Measurement Survey
• Sinazongwe Combination Prevention Evaluation
[SCOPE], in partnership with the Zambia MOH
• Bukoba without New Infections, “Bukoba Bila
Maambukizi Mapya,” [BBM2] in Tanzania
Presentation Objective
Approach to Two PHIAs:
Malawi and Zimbabwe
First Two PHIAs:
Malawi and Zimbabwe
• In collaboration with CDC
• Work with Ministries of Health to develop,
implement and disseminate findings from
PHIA Pilots in Malawi in Zimbabwe
• From April 2014-March 2016
• Currently in protocol development and preimplementation stage
• 16. 3 Million People
• Life expectancy: 54 Years
• Causes of Premature Mortality (YLL):
HIV/AIDS (23.7%), Malaria (10%), Lower
Respiratory Infection (9.7%)1
• HIV Prevalence (age 15-49)
• National: 10.3% 2
• HIV Care and Treatment:
Malawi 2010 DHS.
• 675 ART sites, 470,000 patients on ART
(83% of need)3
1 Malawi Global Burden of Disease Study 2010; Institute for Health Metrics and Evaluation (IHME);
2 UNAIDS HIV and AIDS Estimates
3 UNGASS 2013 Malawi Country Progress Report
• 13 Million People
• Life expectancy: 58 Years
• Causes of Premature Mortality
(YLL): HIV/AIDS (29.0%), Lower
respiratory infection (11.7%),
Diarrheal disease (6.0%)1
Duri Kerina et al. HIV/AIDS: The Zimbabwean Situation and Trends.
American Journal of Clinical Medicine Research, 2013, Vol. 1, No. 1.
• HIV Care and Treatment:
665,000 patients on ART
(77% of need)2
1 Zimbabwe Global Burden of Disease Study 2010; Institute for Health Metrics and Evaluation (IHME);
2 UNGASS 2013 Zimbabwe Country Progress Report
Objectives for Malawi & Zimbabwe
Primary Objectives:
• To estimate HIV incidence (i.e., prevalence of
recent HIV infection) in a household-based,
nationally representative sample of HIVinfected adults
• To estimate the sub-national prevalence of
suppressed HIV viral load (<1000 cells/ml3) in a
household-based, nationally representative
sample of HIV-infected adults
Measuring Incidence Longitudinally
Incidence: new infections in a population
January 1, 2014
January 1, 2015
+ -  
= .25 cases per person year
Measuring Incidence Cross-Sectionally
Objective 1: Incidence
Recent infection (4-6 months) is identified by:
1) Low avidity - weak bonding strength between host
antibody and virus.
2) An elevated level of HIV virus in the body
. . . And converted into an annualized rate
US HIV Incidence (06-09) = .02%
Expected Zimbabwe/Malawi Incidence= ~1%
Objective 2:
Viral Load Suppression (VLS)
HIV Treatment Cascade
% of all people with HIV
Diagnosed Linked to Care Retained in
Adapted from:
Objectives for Malawi & Zimbabwe
• To estimate HIV incidence (i.e., prevalence of
recent HIV infection) in a household-based,
nationally representative sample of HIVinfected adults
• To estimate the sub-national prevalence of
suppressed HIV viral load (<1000 cells/ml3) in a
household-based, nationally representative
sample of HIV-infected adults
Objectives for Malawi & Zimbabwe
Secondary Objectives:
HIV prevalence in adults and children
CD4 T-Cell counts
Transmitted drug resistance
ARV metabolites
Nutrition in HIV positive children
HIV-related risk behaviors
Use of HIV-related services
HIV knowledge and attitudes
Similarities in Malawi and Zimbabwe
• Incidence (national), Viral Load Suppression (zonal)
Eligibility Criteria:
• Must be a HH member
– resides or slept night before in HH
• Must give informed consent
• All adults in every household
• All children (ages 0-14) in every other household
Similarities in Malawi and Zimbabwe
Survey Procedures:
• Collect questionnaire data and blood samples
• Carry out POC HIV and CD4 testing
• Provide counseling and return results
• Refer HIV positives to care
• Transport blood samples to central lab for
additional testing
Similarities in Malawi and Zimbabwe
• Household Questionnaire
• Adult Individual Questionnaire
– Demographics including marriage
– HIV knowledge and attitudes
– Reproduction
– Sexual history
– HIV testing, care and treatment history
Similarities in Malawi and Zimbabwe
Data Management:
• Tablets  Cloud server  in-country server
Laboratory Management:
• Central level testing at a national lab (VL, EID, recency)
Country Oversight Mechanism:
• TWG chaired by MOH
• Sub-committees on Management, Protocol, Data,
Unique to Malawi PHIA
• Oversampling of high prevalence health zones
– for greater precision around cascade analysis
– interest in making programmatic
assessments/comparisons in future rounds
• Sample will include adults aged 15-64
Unique to ZIMPHIA
Secondary Objectives:
• Prevalence of Syphilis
• Describing the extent of stigma
• Sample will include all adults over 15
• Module for adolescents aged 10-14
Presentation Objective
The Future of PHIAs:
The PHIA Project
On the Horizon: ‘the next 20’
In collaboration with CDC and MOHs , ICAP will conduct
PHIAs in ~20 sub-Saharan African countries over the
next 5 years
• No country list yet
Focus on building capacity for population-based surveys
• Strengthen capacity in epidemiology, surveillance, statistics
and national reference laboratory services to collect,
analyze, and use morbidity and mortality data
On the Horizon: ‘the next 20’
• Partnerships with experienced groups:
(e.g., HPTN/MTN/VTN and SHIMS) and ASLM
• Approach for high prevalence countries may
differ for low prevalence countries
• Use results to assess impact of PEPFAR and guide
policies and future programs
Key Messages
• ICAP’s PHIAs are Population-based HIV Impact
• Rigorously measure key indicators of the
epidemic such as
• Incidence
• Viral Load Suppression
Key Messages
• They will provide information on HIV program
effectiveness that can be used to inform future
programs and policies
• The first two PHIAs will be in Malawi and
• ICAP will work on ~20 PHIAs over the next 5
• The Governments of Malawi & Zimbabwe
• United States Centers for Disease Control and
• The U.S. President’s Emergency Plan for AIDS
Relief (PEPFAR)
• Padmaja Patnaik, Suzue Saito, Jessica Justman
• The PHIA Team
Zikomo, Tatenda, Thank you!

similar documents