Providing Continuity of Care in the African HIVAIDS Context

1
Providing Continuity of Care in the African
HIV/AIDS Context
Module Four Gathering, Managing, and Analyzing
Data

• M E Orientation, Atlanta
• August 12, 2004
• Mark Shields
• Strategic Information - GAP Zambia

2
The View In
You are Here
3
Zambian Perspective
4
City life 1
5
City life 2
6
Village life
7
First Things First
Intervention Cascade Inputs ? Activities
? Output ? Outcome ? Impact
Indicator 1 Indicator 2 Indicator 3
Indicator 4 Indicator 5
The What so wed like to learn 350 things,
periodically The How so HOW do we do this
what comes first ?
Information Cascade Status / Events ?
Data capture ? Monitoring ?
Evaluation Interventions
8

• Data is not the goal.
• -Meade Morgan

9
What do we really want ?

• Best Data System for ME
• OR
• Best Health Care for Nation
• Can one system serve both masters ?
• If so, HOW?

10
The Best Care Info Problem

• Zambia Standard of Care requires info
  continuity
• in RH care for Malarial prophylaxis
• in ANC care for PMTCT of HIV
• in Child care for Immunization Schedule (HIV
  vaccine?)
• in OI care for effective TB treatment
• in HIV care for effective HIV/AIDS treatment
• in STD care for effective prophylaxis, contact
  tracing
• Current health care document system fails
• to effectively link records between visits even
  in one site
• to make continuity of care possible across sites
• to integrate different patient level services

11
Why are there health record discontinuities?
-Mechanics Paper documents can be hard to
sustain, store, retrieve share. -Human
Resources Integration of care initiated during
health reform superceded local training limited
staff are too busy to document or use records.
-Local Costs Record keeping takes time and
paper, as well as system organization, oversight,
and feedback. -Political support Record keeping
requires central determination that records be
kept. -Logical Model ? KEY no standard
unique ID.
12
Medical Records
13
(No Transcript)
14
NYT HR issue
July 12, 2004
An Exodus of African Nurses Puts
Infants and the Ill in Peril By CELIA W.
DUGGER LILONGWE, Malawi Six women suddenly
went into the final, agonized minutes of
childbirth. Hlalapi Kunkeyani was the only nurse.
There were no doctors.  
15
Health ID card Orthopaedics
16
What is Possible, and Where?An African Health
Care Information Context

• - Aggregate paper information system, called
  HMIS, tallies clinic visits by type in all 1,340
  clinics in the 72 districts, quarterly.
• - HMIS forms have 77 disease categories, 2 ages,
  0 gender, are tallied by computer at district
  data moves clinic? district? province? nation
• - All 72 district clinics have electricity, phone
  at times, and 1 computer. Most 1,200 health
  care sites do not have power lines, phone, or
  doctor.
• - ART programs are in 9 provincial hospitals.
  NGOs have other sites.
• Top 10-20 of all sites will provide most ART in
  next few years.
• Separate paper systems for TB, HIV/ART, ANC,
  PMTCT, VCT, STI.
• The ANC/PMTCT system has 16 forms, TB about 10
  forms.
• -TB, ANC, Childlt5, ART ? health cards are carried
  by client.
• Some clients will carry blue books, in which
  clinician records notes.

17
(No Transcript)
18
Malawi Health Passport
19
Might new resources for and the requirements of
HIV Care ART in Africatogether represent a
opportunity to engineer One System to provide
Continuity of Carefor all health programs, and
for the Monitoring and Evaluation thereof?
temporary

20
Two essential functional elements of a Continuity
of Care Record

• A continuous personal health record requires
• 1) Data Capture useful information from each
  clinical visit is preserved and linked in an
  organized way, and
• 2) Data Access this information is available at
  any visit, whenever and where ever the client is
  next seen.
• Timely, Electronic Data is hard to get, but may
  yet be easiest way to make continuity of care
  possible.

21
Existing standards efforts CCR, .?
The CCR responded to a need to organize a basic
set of information about a mobile patients
condition. - Diagnoses - Procedures -
Allergies - Labs - Medications - Care
Provided - Care Plan - Referral / Transfer
instructions. A Zambia proposal follows this
CCR motive, but details differ, so it might be
better called ZCCR
CCR is a specification of the ASTM
International, the Massachusetts Medical Society
(MMS), and the Health Information Management and
Systems Society (HIMSS)
22
1) Electronic Data Capture
Paper data is getting more unmanageable as ME
requirements, and tracking services,
proliferate. There are two approaches Paper
first vs. digital first a) Paper first
requires separate human task of data entry, at a
later time, by person not present at interview,
who must be trusted, but who may be less
expensive. Allows clinician to use familiar
paper. Interferes less with interview. b)
Electronic first requires a clinician to change
from paper to electronic, may slow process, but
likely to improve quality of care and ME of
system itself.
23
CA Registry
Paper first The Zambian National Cancer Registry
Staff in 2001
24
Electronic First a Malawi instance
ATMs are another instance in the region
25
Dueling Palm Computers
Zambia Dueling Palm Computers
26
2) Data Access Distribution
Once data is digital, how do we get it to point
of care? There are two approaches
Client-server Peer to peer

• Centralize redistribute
• Prior to need
• On demand
• via Server and WAN
• Existing in Lusaka, 24 sites,
• single type of service - ANC

Clinic Z
Clinic A
Clinic Z
Transfer via Phone call, (e/mail) Staff, or
Client carried record Existing in rest of
Zambia, in 1,316 sites, all points
Clinic A
27
Cellular Coverage 04
Cellular Communications Infrastructure Coverage
in Zambia in 2004
28
PDA (hand held computer) with Smart card reader
29
Data Distribution Infrastructure

• WAN routers/switches/firewalls, reliable
  telecom,
• client-server technology
• E-mail computers, printers, phone lines,
  electricity
• Smart card computers, printers, electricity
• Paper roads

30
Attributes of Smart card PDA
Smart card can - hold entire patient record in
Zambia, for 2-3 - preserve health information
confidentially - be immediately received from,
and returned to client - provide low bandwidth
data link between providers - if lost, card
with prior data can be reissued, re-linked PDA
can - provide templates for data capture,
assuring minimal standards for ME and for
health records (Standards) - promote
algorithms for health care update algorithms
without retraining all health providers
(Services/Training) - hold references such
as drug guidelines, or manuals decision
support may prevent errors service gaps
(Quality) Paper can - be printed whenever
needed or desired necessary in case of absence
of power PDA Personal Digital
Assistant or hand held computer
31
Information Cycle
Structure of Learning Systems
Information-ActionFeedbackCycle
Aggregate Analyse
Perceive
Learn
Information
Observe
Data
Apply
Collect
Knowledge
Desired Outcomes SHOULD guide Processes, BUT
cause effect must first be understood to do
this.
Facts
Decisions
Monitor
Prioritize
Events
Outcome Indicators Measure Results
Actions
Circum- stances
Experience
Do
Implement
Process Indicators Measure Activity
Resource Indicators Measure Assets
Prepare
32
Ideal Training of Cliniciansclose the feedback
loop

• On Location Train at Point of Care
• In Context Train regarding health services
  provided
• Immediate Train at time of required decision
• Personal Train regarding issues this clinician
  has
• Consistent Train the same way each time error
  made
• -Correct Train to standard approved guidelines
• These standards might be expected of PDA
  feedback, but not of any other training method.

33
A Zambian Patient Level Continuity of Care
Record Design

• Electronic First, Paper Second, to permit real
  time decision support, support changing ME
  needs
• 2. Make PDA useful to clinicians, minimize
  burden.
• 3. USE existing health card carrying tradition,
  by integrating digital information on smart
  cards.
• 4. Client is the Network health record
  connectivity.
• 5. Focus first on 100 outpatient sites with
  electrical infrastructure (where all ART is
  focused, currently)
• 6. Enroll well-care clients (ANC) first.

34
Est. Cost PDA Smart card in 133 top sites

• Cost of one PDA /clinician another /facility,
  for top 133 sites Figure 300/PDA x 10
  PDAs/site x100 sites 400K /4 yrs, 100K/yr
  
• Cost of one smart card / patient, for top 133
  sites, including printing
  Figure 3/card x (50,000 TB/yr 50,000 HIV/yr
  200,000 PMCT/yr 300,000 other STD, VCT hosp.
  visits/yr) 1,800K /yr for cards
• Cost of putting 4 printers in 133 facilities 2
  replacements in 4 years Assume 300 / printer x
  800 240,000 / 4 years, or 60,000/yr
  Figure consumables at twice this cost, total
  180K / yr for printing
• Cost of 50,000 persons on ART year 1, 90,000
  persons year 2, etc. for 4 years of 50
  successful identification and treatment of AIDS
  Figure 300/ person /yr 15M in yr 1,
  moving up to 45M /yr in 4 yrs if treatment is
  80 effective, 30,000K /yr for treatment
• Cost PDAS/CsPrinters 7 1st line ARV drug
  cost, 3 USG funds

35
Patient level data is needed for TWO compelling
reasons ?
The best system to address many evolving ME
needs can be same as the best system to provide
patients Continuity of Care if it is a system
with electronic first data capture that supports
a patient-level continuity of care record.
36
The Youth
37
Thanks

• Thanks! it was a lot to digest