SmartCards in Malawi

1
SmartCards in Malawi
Matt Boxshall The Lighthouse Trust Lilongwe,
Malawi mattb_at_sdnp.org.mw
2
Malawi

• Pop 11m
• GDP / Capita (2000) US170
• 65 poor - unable to meet daily nutritional
  needs (NSO 2000)
• HIV prevalence - Urban 22.5, Rural 10.7,
  around 1million infected
• Life expectancy dropping, lt 40yrs
• Pop / nurse approximately 3,500, or about 1 per
  100 HAART eligible patients

3
HAART in Malawi

• Approximately 3,000 registered on HAART, mid 2003
• Four formal sites currently operational, total
  capacity to register perhaps 3500 new HAART
  clients annually
• Global Fund money will pay for free HAART for
  gtgt25,000 over 5 years

4
The Lighthouse

• Background - Hospital Volunteers, Complimentary
  services, Trust working as a PPP
• Strategy - Scale, Model, Build Capacity
• Services - CHBC, VCT, Clinic - HAART

5
Graph Reg
Cumulative HAART Registrations

• HAART registration graph

LCH to Lighthouse

• Government Drugs _at_ US 30 / month
• Demand vs Supply

6
Graph Clinic visits
Lighthouse Clinic Monthly Client Visits

• Cumulative Workload
• Reaching Capacity - where to next?

7
Response

• Fast Track to move review non-problematic
  reviews to more junior staff (nurses)
• Decentralize reviews to health centers
• BUT patient data management systems are also
  increasingly stretched, and decentralization (and
  ARV shopping) will only exacerbate this-

• How do we identify and follow our patients?
• How do we know who fails to pick up their drugs
  (or picks up more than one supply?)
• How do we gather information centrally for ME?
• How do we account for drugs?

8
A Technological Fix?

• We cant throw people at this problem - we dont
  have them!
• We need something
• Easy to use
• Robust
• Scalable
• Tamper-proof
• Reasonably priced

9
SmartCards

• A programmable chip on a credit card, read by a
  point of sale device (PoS)
• Successfully implemented in the region - KWS,
  petrol stations, banks, micro-finance projects
  etc
• Local providers available and interested
  (Malswitch, NET1)
• Costs approx 5 / card, rental of PoS approx 25
  / month - small vs drug costs

10
Programming the Card

• Cards issued at prescribing site
• Each card has 550 fields or wallets
• Fields can be entered at registration, updated at
  drug collection, calculated, password protected
  etc.
• Biometric information
• can be carried - in this case,
• fingerprint scans

11
Sample Fields

1. Fingerprint Biometrics x2
2. Patient ID number
3. Date of start of ART
4. Date of Registration onto SmartCard ART
5. Registering Clinic Name
6. Registering Clinician Name
7. Date of first collection of drugs with SmartCard
8. Drug regimen details ( change regimen flag?)
9. Date of last drug collection
10. Date current drug supply will finish
11. Location of last drug collection
12. Name of person dispensing drugs
13. Number of pills dispensed
14. Collection by Patient or Guardian
15. Cumulative Total Pills received
16. Patient Working
17. Drug Credit
18. Default Flag

12
Drug Collection

• Any patient should be able to pick up drugs
  anywhere a PoS device is available
• Patient (or Guardian) identified offline
• Automated checks run (eg late collection)
• Drug collection authorised, details updated to
  card
• Details downloaded to PoS
• Vendor card updated
• Printout if required

13
Data Collection

• Patient data collected electronically and largely
  automatically at PoS
• PoS downloaded either by dial-up, or by transfer
  to milking card
• Drugs credited to Cards for transfer between
  sites, and stock management (at least partly)
  automated
• Drug and Patient management, ME, closely linked

14
Unresolved Issues?

• Health worker uptake will be dependant on
  perceived value, particularly in time saved
• Patients may resist, preferring less control
  (although balanced vs flexibility of collection
  site)
• Centralized electronic data collection may raise
  confidentiality issues
• Responsibility for system management may be
  divisive - clinical services, medical stores?

15
The Way Forward

• Technical specifications have been drafted with
  suppliers
• Lighthouse will initiate with partners in MoHP
  (and others)
• Operational research should evaluate
  effectiveness
• If successful, roll-out will need to be fast to
  establish system in line with planned HAART
  scale-up - system makes a lot more sense if it is
  country wide

16
Acknowledgements

• Lighthouse - Sam Phiri, Florian Neuhann, Ralf
  Weigel
• University of North Carolina - Mina Hosseinipour
• Baobab Health Partnerships - Richard Altmann,
  Gerry Douglas
• Net1 - Brenda Stewart

17
BAOBAB