Transcript SmartCards in Malawi

SmartCards in Malawi
Matt Boxshall
The Lighthouse Trust
Lilongwe, Malawi
[email protected]
Malawi






Pop 11m
GDP / Capita (2000) = US$170
65% ‘poor’ - unable to meet daily
nutritional needs (NSO 2000)
HIV prevalence - Urban 22.5%, Rural
10.7%, around 1million infected
Life expectancy dropping, < 40yrs
Pop / nurse approximately 3,500, or about
1 per 100 HAART eligible patients
HAART in Malawi

Approximately 3,000 registered on HAART,
mid 2003
 Four ‘formal’ sites currently operational,
total capacity to register perhaps 3500
new HAART clients annually
 Global Fund money will pay for free
HAART for >>25,000 over 5 years
The Lighthouse

Background - Hospital Volunteers, Complimentary
services, Trust working as a PPP
Strategy - Scale, Model, Build Capacity
 Services - CHBC, VCT, Clinic - HAART

Cumulative ARV Registrations
Cumulative HAART Registrations
 HAART
Graph Reg
registration graph
• Government Drugs @ US$ 30 / month
• Demand vs Supply
Apr-03
Mar-03
Feb-03
Jan-03
Dec-02
Nov-02
Oct-02
Cumulative
Registrations
Sep-02
Jul-02
Jun-02
May-02
LCH
to Lighthouse
Aug-02
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
ClinicMonthly
Reviews Client Visits
Lighthouse Clinic
1800
Lighthouse Centre
Opens
1600
1400
Graph Clinic visits
1200
1000
Total ARV Reviews
800
Reviews (non ARV)
600
400
200
• Cumulative Workload
• Reaching Capacity - where to next?
Apr-03
Feb-03
Dec-02
Oct-02
Aug-02
Jun-02
Apr-02
Feb-02
Dec-01
Oct-01
Aug-01
Jun-01
Apr-01
0
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
M+E?
How do we account for drugs?
A Technological Fix?
We can’t throw people at this
problem - we don’t have them!
 We need something;

–
–
–
–
–
Easy to use
Robust
Scalable
Tamper-proof
Reasonably priced
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
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
Sample Fields
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Fingerprint Biometrics x2
Patient ID number
Date of start of ART
Date of Registration onto SmartCard ART
Registering Clinic Name
Registering Clinician Name
Date of first collection of drugs with SmartCard
Drug regimen details (+ change regimen flag?)
Date of last drug collection
Date current drug supply will finish
Location of last drug collection
Name of person dispensing drugs
Number of pills dispensed
Collection by Patient or Guardian
Cumulative Total Pills received
Patient Working
Drug Credit
Default Flag
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
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, M&E,
closely linked
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?
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
Acknowledgements

Lighthouse - Sam Phiri, Florian Neuhann, Ralf
Weigel

University of North Carolina - Mina Hosseinipour

Baobab Health Partnerships - Richard Altmann,
Gerry Douglas

Net1 - Brenda Stewart
BAOBAB