Community-based Distribution of DMPA in Madagascar: Pilot Introduction and Early Scale-up Theresa Hatzell Hoke Family Health International
Jan 29, 2016
Community-based Distribution of DMPA in Madagascar:
Pilot Introduction and Early Scale-up
Theresa Hatzell Hoke
Family Health International
Major Partners• Ministry of Health and Family Planning,
Madagascar
• USAID
• Family Health International
• SantéNet
• Population Services International
• Adventist Development Relief Agency (ADRA)
• Action Santé Organisation Secours (ASOS)
Pilot Implementation
Scale-up
Pilot Evaluation
Presentation Overview
Pilot Implementation
Scale-up
Pilot Evaluation
Presentation Overview
Contextual Factors Favoring CBD of DMPA in Madagascar
Population:18 M, 74% rural
Access: 65% of population lives more than 10km from a health center
Service quality: human resource shortages, heavy workloads, waiting times
Unmet need for family planning: 24%
Number of sites with CBD workers : 3422
Growing preference for DMPA
2006 update of National Reproductive Health Norms and Procedures:
Trained community-based workers permitted to provide DMPA services.
Introduced in 13 “communes ” across 4 districts
Added to pre-existing CBD programs
61 CBD agents participated, along with 26 supervisors
Pilot Intervention
Intervention Components
1. Recruitment and selection
2. Training
3. Certification and installation
4. Commodity management
5. Reporting
6. Supervision
3-Day Training
Training ResourcesCBD of DMPA Cartoon Manual
Certification and Installation
Commodity Management
• Initial stock of 15 doses with supplies
• Monthly re-supply at health center
• Single-use syringes
• Sharps container
Supervision and Reporting
DMPA intervention added to pre-existing systems and procedures:
• Monthly contact by clinician and NGO supervisors
Occasional visits by district and central-level supervisors
• Capitalized on pre-existing reporting forms
Pilot Implementation
Scale-up
Pilot Evaluation
Presentation Overview
Measuring Pilot Success1. Did CBD workers provide injection
services in accordance with quality standards?
2. Did use of contraception increase in the communities served?
3. Did the support mechanisms function as intended?
4. Is CBD of DMPA acceptable to CBD agents and their clients and supervisors?
Evaluation Methods
• Data collection 7 months after training
• Structured interviews 61 agents 25 supervisors 303 clients
• Review of service records
Essential Knowledge Score
0%
20%
40%
60%
80%
100%
ADRAMoramanga
(n=19)
ASOSMoramanga
(n=12)
ASOS Sud(n=30)
All Sites (n=61)
Pe
rce
nta
ge
of
CB
D W
ork
ers
18
16
14
CBD Agent Knowledge about DMPA Service Provision
CBD Agent Performance: Client Reports (n=303)
0 20 40 60 80 100
Knows DMPA gives no STI protection
Agent respects confidentiality
Counseled about amenorrhea
Counseled on side effects
Knows duration of pregnancy protection
No problem with injection site
Satisfied with injection technique
% of clients
Impact of Intervention on Contraceptive Use
CBD DMPA acceptors between Jan-June 2007: 1,662 women
Age range: 15 – 49 years
Number of living children: 1-11
Desire a future pregnancy?Yes 63%No 31%Unsure 6%
New FP user, 28%
New DMPA user, 25%
Used DMPA in the past,
47%
Contraceptive History of DMPA Acceptors
(n=303)
3-Month Re-injection (n=199)
No re-injection received,
4%
Other source, 2%
Received from CBD workers,
94%
Impact on Accessibility
• Mean travel time to health center : 136 minutes
• Mean travel time to CBD agent: 27 minutes
• Mean time “saved”:109 minutes
Support Mechanisms: Key Lessons
• Reduce training class size
• Capitalize on supervisors’ strengths
• Ensure consistent DMPA supplies at health centers
• Streamline reporting
• Add DMPA intervention to a well-functioning CBD family planning program
Pilot Implementation
Scale-up
Pilot Evaluation
Presentation Overview
Scale-up Intervention• Training : 3 ½ days
instead of 3 days
• Stronger link with public sector health system
• Expanded collaboration with NGOs
• Prioritized remote, populated areas with committed district health officers
Status Update
• MOHFP collaborating with 16 NGOs
• An additional 406 agents trained
• Of the 61 “pilot” agents, 58 participated in refresher training in June 2008
• By April 2009, 385 agents providing DMPA services in 111 communes across 27 districts