Presentation on theme: "Most at Risk Adolescents in Eastern Europe Building the Evidence Base Joanna Busza & Megan Douthwaite London School of Hygiene & Tropical Medicine September."— Presentation transcript:
Most at Risk Adolescents in Eastern Europe Building the Evidence Base Joanna Busza & Megan Douthwaite London School of Hygiene & Tropical Medicine September 2, 2009
Research as process … study design, data collection, and analysis remain flexible and adapt as new questions or findings emerge. Research into action … more important to produce useful information with practical implications than to create a lot of data. Research for skills building …systematically working through each step strengthens collaboration within country teams and the region. LSHTM approach to technical assistance
Technical Assistance Proposed Structure 1)Regional training & study design workshops 2)Development of standardised tools 3)Guidance on country-specific issues 4)Data analysis workshops 5)Country visits for specified activities 6)Distance based backstopping & advice
Select local sample populations & recruitment strategies Synthesise available data Collect data on risk & protective behaviours (7 countries) Explore context & dynamics (4 countries) Identify knowledge gaps on MARA Develop interventions Analyse costs and effects; Follow-up survey (?) Process evaluations (3 countries) Research Cycle
Choosing the right methodology is a logical process, with several decision-making steps … Exploring an unknown subject … … Gathering population-based data... … Comparing across the region …. … Planning interventions … … Evaluating services …
Designing Baseline Studies 1)What data on MARA already exists? 2)Who has contact with target groups? 3)What are the advocacy objectives? 4)What are plans/ goals for interventions? 5)What is the main purpose for the evidence produced?
Risk vs. Vulnerability Assess & profile % of most-at-risk populations who are adolescents? OR Determine & characterise % of specific adolescent groups who are involved in risk behaviours?
Eco-social framework for Risk Individual Community context Structural shapers Peer norms & Networks Available services Local Environment Skills Knowledge Risk perception Laws Policies Cultural attitudes & expectations Political economy Biological Susceptibility
Research Components 1)Sample selection 2)Development of instruments 3)Adaptation to country-specific contexts 4)Addressing ethical issues 5)Data analysis and interpretation 6)Qualitative studies in select countries to explore specific dimensions of MARA experience 7)Intervention research in 3 countries to evaluate & cost MARA-targeted services
Identify 2 nd dary sources of data Distill most important findings Present results in clear format Combine qualitative and quantitative data Offer rigorous interpretation Data adequate for action Results widely disseminated Findings in form compatible with other data Contributes to wider evidence base Interventions can be planned or monitored Compile & interpret existing data Identify information gaps Design study Select sample and tools Train team Use appropriate fieldworkers Monitor quality during research Collect information from multiple sources Manage ethical & logistic issues Planning Data CollectionAnalysisResults/Use Research Trajectory
Sampling MARA 1)Venue based 2)Institution based 3)Chain Referral Respondent Driven Sampling Network recruitment Snowball sampling 4)Convenience 5)Combined sampling approaches
Developing Indicators Research design workshop, Belgrade Integrating risk and vulnerability measures Ensuring ability to compile UNGASS indicators All MARA behaviours included ≈40 standardised core indicators + flexibility for country-specific topics
Data Collection Tools LSHTM drafted male & female core questionnaires Colour-coded core and recommended questions Feedback incorporated from country teams Each country adapted, translated and pilot tested Guidelines distributed for compiling indicators
Core Questionnaires Eligibility criteria Demographic profile Injecting drug use (frequency, drug choice, and sharing practices) HIV knowledge Sexual behaviour (including commercial & casual partnerships) MSM behaviour Access and use of services (including condoms & HIV testing) Experience of detention
Baseline Studies: Preliminary Results
Diversity of Study Populations & Methods Focus on Risk Young IDU in Serbia, Romania, Moldova, Albania Young sex workers in Romania & Albania Young MSM in Albania & Moldova Focus on Vulnerability Young people in Roma settlements in Montenegro Institutionalised settings in BiH & Moldova Street children in Ukraine
RecruitmentReached Albania - MSMRDS50 BiH- institutions ALL392 Moldova – institutions ALL81 Montenegro – Displaced Roma Venue-based/ Snowball 290 Ukraine – Street children Network/ Snowball 805 Sample Populations: Other
LSHTM Analysis Analysis conducted for 6 data sets Romania FSW Romania IDU Serbia IDU Moldova IDU Montenegro Roma Ukrainian street children
LSHTM Analysis Standardisation across data sets Age range limited to EXCEPT for Ukraine (10-19 yrs) Selection of indicators that maximise comparability across the region Disaggregation by country, age and sex Chi-square test for statistical significance (& Fischers exact test where numbers <5 per cell)
Data Quality Strengths - Relatively good quality re: internal consistency within data sets Weaknesses - Caution required in interpretation of some variables due to small numbers Some variation in way questions were asked Cleaning issues – Skip patterns not all followed correctly, making it difficult to choose questions for compiling indicators
Sample sizes & gender distribution
Injecting Drug Use
IDU Risk Profiles % Injecting Daily Moldova2.6 Romania IDU 76.2 Romania SW 88.3 Serbia46.3 Ukraine5.3 % Sharing injecting equipment past month Moldova88.1 Romania IDU 19.0 Romania SW 71.0 Serbia35.1 Ukraine34.8
IDU Diversity of injecting drug use patterns among the study samples Moldova has a greater % of young IDU, but injecting behaviour is sporadic In Montenegro, no IDU behaviours reported among IDP Roma Sex workers who inject drugs may have riskier behaviour and poorer service use
Sexual Risk Profile
Condom Use with different partners
Sexual Behaviour All studies found high rates of sexual experience, including among adolescents. Sexual experience increases with age Condom use follows familiar pattern, with decreasing consistency for longer term partners MSM behaviours rare, with exception of Montenegro and Ukraine sites.
Use of Services
Service Use Pharmacies appear acceptable source of both injecting equipment and condoms Knowledge of services higher than use Surprising number of respondents ever tested for HIV, and this increases with age Low use of rehabilitation services, especially among adolescents.
Knowledge by Age *** * ** *** * ** * * ***** **
Service use by Age *** * * ** * * *** ** ***
Enhanced Vulnerability Younger cohorts have poorer knowledge of HIV transmission and are less likely to seek formal services Detention & harassment by police a regular event, especially for boys Adolescent sex workers report more experience of forced sex and are less likely to use condoms consistently Association between younger age and child protection institutionalisation
Vulnerability by Sex
Girls experience unmet need for other reproductive health, especially contraception. Girls report higher rates of forced sex Sex work is NOT always higher among girls The steady partners of female IDU are more likely to also be IDU than among males.
Moving Forwards Extending programmes that already work with IDU and sex workers – addressing overlaps Considering links between harm reduction & child protection Using “entry points” identified by research – i.e. willingness of adolescents to visit pharmacies Addressing legal & institutional barriers
Next Steps: Qualitative Studies Interviews and focus group discussions conducted in Ukraine with MARA sex workers Formative interviews with MARA MSM, sex workers and providers in Moldova Focus group with IDU and interviews with sex workers in Romania Rapid assessment with IDU planned in Moldova to define intervention
Next Steps: Intervention Studies Ukraine – frontline services for street based sex workers in Mykolaev Romania – referral link network developed between child protection services and health providers Moldova – peer delivered intervention to reduce injection initiation under consideration M&E frameworks developed to guide process and output evaluations
Future Steps Write-up of baseline results (1 paper in press) Intervention and M&E framework developed for Moldova Process evaluation workshop in Ukraine; qualitative data analysis Follow up on intervention research in Romania Extend model to other countries (?)
Lessons Learned Focused, country-specific technical assistance more effective In depth research in a small number of countries better than “standardised” capacity building for many countries Regional workshops to compare study designs and results useful to national researchers Need more than 3 years to conduct baseline, qualitative and evaluation research components
Striking a balance…. Regional standardisation Country specific priorities Feasible in programme timeframe Scientific rigour Data for monitoring Data for policy advocacy Shared learning Tailored support