A Collaborative Health Research and Service Program in northern Tanzania

1
A Collaborative Health Research and Service
Program in northern Tanzania

• John A. Crump MB ChB DTMH
• Assistant Professor of Medicine
• Division of Infectious Diseases and International
  Health
  
• Senior Lecturer Kilimanjaro Christian Medical
  College
  

2
Overview

• Tanzania
• Medicine collaboration in Tanzania
• Philosophy and core values
• Training
• Research
• Past research accomplishments
• Current research portfolio
• Focus on febrile illness studies
• Future directions

3
(No Transcript)
4
Tanzania

• Population 33 million
• Area 886000 km2
• Human development index rank 162nd
• Per capita GDP USD 251
• HIV seroprevalence 7.0

5
(No Transcript)
6
History of collaboration in Tanzania

• 1986-1994 Coast
• Muhimbili National Hospital Dar es Salaam
• Mid-1990s northern Tanzania
• Kilimanjaro Christian Medical Centre Moshi
• Medical resident rotations
• 2002 scale-up of activities
• Place faculty develop research program
• KIWAKUKKI Moshi
• Kibongoto National Tuberculosis Hospital Sanya
  Juu
  
• Mawenzi Regional Hospital Moshi
• HIV prevention treatment and care

7
Survival of KIWAKKUKI home-based care clients
2003-2005 n226
Tillekeratne LG et al. World AIDS Conference
2006 Abstract MoPe0303
8
Philosophy and core values

• Research with service
• Training
• Patient care and public health
• True collaboration
• Long-term commitment
• Decades not years months or days
• Progress may be slow
• Fair weather and foul
• International health as a discipline
• 100 year history
• International health best practiced from the
  field
  
• Minimize public health tourism

9
Department of Medicine Trainees
10
Tanzanian trainees 2002-present
11
Research with Service
12
Cost-effectiveness of free versus co-pay HIV
voluntary counseling and testing

• Would provision of free HIV voluntary counseling
  and testing (VCT) in Tanzania be cost-effective
  

13
Methods

• 813 VCT clients
• KIWAKKUKI
• May Nov 2003
• Before during after free VCT campaign
• Cost-effectiveness model
• Number of tests per day
• Costs of testing
• Benefits of knowing HIV serostatus infections
  averted access to treatment
  

14
Persons receiving VCT per day KIWAKKUKI VCT
2003 n813
Mean daily volume prior to free testing
Mean 15.0 pMean daily volume during free testing
Mean daily volume after free testing
Mean 7.1 pNumber of clients
Mean 4.1
June
August
July
September
October
November
2003
Thielman NM et al. Am J Public Health 2006 96
114-9
15
Cost-Effectiveness of VCT
Thielman NM et al. Am J Public Health 2006 96
114-9
16
Simple low cost approaches to identifying
patients with CD4 count

How can we identify HIV-infected adults with CD4
count capacity is limited

17
Methods

• 202 subjects recently diagnosed with HIV
  referred
  
• VCT centers
• Aug 2004 Jun 2005
• WHO staging history and examination
  anthropometry and simple lab tests
  
• ESR
• CBC
• CD4 count by manual Beckman Coulter method
• Bivariable analysis to predict CD4
• Partition tree analysis of significant variables

18
Distribution of CD4 count by WHO stage Moshi
Tanzania 2004-5
CD4 count mm3
WHO stage
Interquartile range range
Morpeth SC et al. CROI 2005 Boston Ma.
Abstract 638
19
Number of recently diagnosed HIV-infected persons
who would be triaged to treatment using 4
different strategies by CD4-count stratum
Receiver-operator characteristics
area-under-the-curve
Morpeth SC et al. CROI 2005 abstract 638a
20
Trimethoprim-sulfamethoxazole study

• What effect is use of TMP-SXT prophylaxis in
  persons with symptomatic HIV disease in Africa
  having on emergence of antimicrobial resistance

21
Methods

• 184 subjects recently diagnosed with HIV
  referred
  
• VCT centers
• Aug 2004 Dec 2005
• Prospective observational cohort study
• Arm A HIV-uninfected
• Arm B HIV-infected asymptomatic (no TMP-SXT)
• Arm C HIV-infected symptomatic (TMP-SXT)
• Follow-up
• Baseline
• Weeks 1 2 4 24
• Stool
• Screened for Escherichia coli not susceptible to
  TMP-SXT

22
Results

• Baseline non-susceptibility was high
• Arm A 26 (57) of 46
• Arm B 28 (70) of 40
• Arm C 41 (67) of 61
• Introduction on TMP-SXT rapidly led to
  non-susceptibility
  
• 95 of Arm C patients had non-susceptible E.
  coli within 1 week of TMP-SXT
  
• Arm C vs. Arm A p0.007
• Arm C vs. Arm B p0.020

23
Antiretroviral drug resistance and adherence
study (ADAR)

• How common is virologic failure in patients
  receiving ART in Tanzania and who fails and why

24
Methods

• Retrospective cohort study
• 150 HIV-infected adult patients June-August
  2005
  
• FDC D4T/3TC/NVP 6 months
• Consecutive patients presenting for follow up
• Standardized questionnaire
• Sociodemographics
• Economic conditions
• HIV and ART knowledge beliefs disclosure
• Adherence
• Access to care
• Plasma
• HIV RNA quantitation

25
Risk Factors for Virologic FailureMultivariable
Analysis
Model controlled for age and gender and included
variables with pRamadhani HO et al. World AIDS Conference 2006
abstract ThLb0213
26
Self-Funded ART and Virologic Failure

• Persons paying for ART were more likely to be
  maladherent
  
• r0.54 p

27
Tuberculosis and HIV Immune Reconstitution
Syndrome Trial (THIRST)

• Is it better to start ART immediately or to defer
  ART in patients co-infected with tuberculosis

28
Methods

• Randomized controlled trial
• 70 patients with HIV infection and smear-positive
  pulmonary tuberculosis
  
• Initiate TB treatment then FDC ZDV/LMV/ABC
  after
  
• 2 weeks
• 8 weeks
• Follow-up
• 104 weeks

29
CD4-positive lymphocyte responses
CD4 count (cells/mm3)
p Entry
Week 12
Week 36
Week 48
Week 24
n 70 69
68 67
66
Shao HJ et al. CROI 2006 Abstract 796
30
Outcomes

• 3 study subjects deaths were not thought to be
  related to study medications
  
• ZDV/LMV/ABC was discontinued in 6 subjects
• - 2 with dose-limiting anemia requiring
  substitution of stavudine for ZDV
  
• - 4 with suspected ABC hypersensitivity
  reactions
  
• At week 48 64 (96) of 67 had HIV RNA copies/mL
  
• To date no cases of TB-associated immune
  reconstitution syndrome have been observed
  

31
HIV seroprevalence
Kibongoto Tuberculosis
THIRST
Hospitalized patients
Mawenzi Regional
ADAR
KCMC
HIV inpatient characteristics
Sociodemographics (VCT)
Tuberculosis symptoms
HIV VCT clients
HIV staging
TMP-SXT
Community-based subjects
Cost-effectiveness
HIV-seroincidence
HIV HBC clients
Sociodemographics (HBC)
Morbidity and survival
32
KCMC BIOTECHNOLOGY LABORATORY
Microbiology
Molecular Virology
Hematology and Chemistry
Cryopreservation
Immunology
33
Grant support for researchDuke-KCMC
Collaboration 2002-present
Clinical Research Site
Abbott (LPV/RTV)
CHAVI
CFAR (Clin Core)
ISAAC
Grant support (thousands of USD)
AITRP
GSK (THIRST)
CIPRA R03
CFAR (ADAR)
Roche Laboratories (VCT)
CFAR (Cervical Ca)
Year
34
Adult HIV treatment studies
35
Pediatric HIV treatment studies
36
HIV vaccine-related research
37
Community studies
38
HIV co-infection studies
39
HIV co-infection studies
40
Febrile illness in northern Tanzania

• Febrile illness accounts for 10-30 of admissions
  to hospital in northern Tanzania
  
• HIV co-infection is common among inpatients
• Community 7 (2004)
• Medical inpatients 21 (2000)
• Tuberculosis inpatients 41 (2000)
• Laboratory capacity is limited
• Fever is often managed empirically with
  antimalarials even whennegative
  
• Causes of fever differ from those in the west
• Little work has been done on prevention
  diagnosis and treatment of leading causes of
  fever

41
Severe febrile illness hospital management
northern Tanzania
Reyburn HG et al. Brit Med J 2004 329 1212
42
Muhimbili National Hospital Dar es
SalaamTanzania 1995
Archibald LK et al. Clin Infect Dis 1998 26
290
43
Muhimbili National Hospital Dar es
SalaamTanzania 1995
Archibald LK et al. Clin Infect Dis 1998 26
290
44
KCMC Moshi Tanzania 2007
45
Nested studies and goals

• Causes of febrile illness in children admitted to
  hospital in a low transmission area of
  P.falciparum
  
• To impact on Integrated Management of Childhood
  Illness (IMCI) algorithms
  
• Disseminated tuberculosis diagnostics study
• To improve the clinical and laboratory diagnosis
  of disseminated tuberculosis
  
• Non-Typhi Salmonella in HIV case-control study
• To inform prevention guidelines for
  HIV-associated non-Typhi Salmonella bacteremia in
  Africa

46
Future directions

• Continue to work on health problems of importance
  in Tanzania
  
• Descriptive interventional studies
• Opportunistic focused research plan
• Expand the training program
• Long-term training
• Advanced degrees
• Assist to foster a critical mass of researchers
  at KCMC
  
• Independent investigators
• Life-long collaborators

47
Conclusions

• Platform for HIV and infectious diseases
  research
  
• Strong collaborative relationships with hospital
  community and other researchers
  
• Emphasis on training
• Track record of research with service
• Growing personnel and infrastructure
• High quality laboratory facilities
• Potential to underpin ambitious and growing
  research agenda
  
• HIV treatment and prevention research
• HIV co-infection research
• Community studies

48
Duke-KCMC Collaboration Team Moshi April 2007
49
Acknowledgements

• Duke University Medical Center
• John D. Hamilton MD
• John A. Bartlett MD
• Nathan M. Thielman MD MPH
• Charles Muiruri
• L. Barth Reller MD DTMH
• G. Ralph Corey MD
• Barton F. Haynes MD
• Michael Merson MD
• Anne B. Morrissey MS
• Jean Gratz MS
• Julia Giner RN
• Jan Ostermann PhD
• Gary M. Cox MD
• Carol Dukes Hamilton MD
• Coleen K. Cunningham MD
• KIWAKKUKI
• Rehema A. Kiwera AdvDipClinMed

• Kilimanjaro Christian Medical Centre
• John F. Shao MD PhD
• Mark E. Swai MD
• Humphrey J. Shao MD
• Habib O. Ramadhani MD
• Florida P. Muro MD
• Bahati P. Msaki MD
• Emmanual Balandya MD
• Venance P. Maro MD
• Grace D. Kinabo MD
• Levina Msuya MD
• Werner Schimana MD
• Moses W. Sichangi MSc
• Francis P. Karia MBA
• Ahaz T. Kulanga MBA

50
(No Transcript)