Comparing antiretroviral treatment outcomes between a prospective community-based and hospital-based cohort of HIV patients in rural Uganda.

Abstract

BACKGROUND: Improved availability of antiretroviral therapy in sub-Saharan Africa is intended to benefit all eligible HIV-infected patients; however in reality antiretroviral services are mainly offered in urban hospitals. Poor rural patients have difficulty accessing the drugs, making the provision of antiretroviral therapy inequitable. Initial tests of community-based treatment programs in Uganda suggest that home-based treatment of HIV/AIDS may equal hospital-based treatment; however the literature reveals limited experiences with such programs. THE RESEARCH: This intervention study aimed to; 1) assess the effectiveness of a rural community-based ART program in a subcounty (Rwimi) of Uganda; and 2) compare treatment outcomes and mortality in a rural community-based antiretroviral therapy program with a well-established hospital-based program. Ethics approvals were obtained in Canada and Uganda. RESULTS AND OUTCOMES: Successful treatment outcomes after two years in both the community and hospital cohorts were high. All-cause mortality was similar in both cohorts. However, community-based patients were more likely to achieve viral suppression and had good adherence to treatment. The community-based program was slightly more cost-effective. Per capita costs in both settings were unsustainable, representing more than Uganda's Primary Health Care Services current expenditures per person per year for all health services. The unpaid community volunteers showed high participation and low attrition rates for the two years that this program was evaluated. CHALLENGES AND SUCCESSES: Key successes of this study include the demonstration that antiretroviral therapy can be provided in a rural setting, the creation of a research infrastructure and culture within Kabarole's health system, and the establishment of a research collaboration capable of enriching the global health graduate program at the University of Alberta. Challenging questions about the long-term feasibility and sustainability of a community-based ARV program in Uganda still remain. THE PARTNERSHIP: This project is a continuation of previous successful collaborations between the School of Public Health of Makerere University, the School of Public Health of University of Alberta, the Kabarole District Administration and the Kabarole Research and Resource Center.

Full text

Page 1

RESEARCH ARTICLE Open Access
Comparing antiretroviral treatment outcomes
between a prospective community-based and
hospital-based cohort of HIV patients in rural
Uganda
Walter Kipp1*†, Joseph Konde-Lule2†, Tom Rubaale3, Joa Okech-Ojony3, Arif Alibhai1, Duncan L Saunders1
Abstract
Background: Improved availability of antiretroviral therapy in sub-Saharan Africa is intended to benefit all eligible
HIV-infected patients; however in reality antiretroviral services are mainly offered in urban hospitals. Poor rural
patients have difficulty accessing the drugs, making the provision of antiretroviral therapy inequitable. Initial tests of
community-based treatment programs in Uganda suggest that home-based treatment of HIV/AIDS may equal
hospital-based treatment; however the literature reveals limited experiences with such programs.
The research: This intervention study aimed to; 1) assess the effectiveness of a rural community-based ART
program in a subcounty (Rwimi) of Uganda; and 2) compare treatment outcomes and mortality in a rural
community-based antiretroviral therapy program with a well-established hospital-based program. Ethics approvals
were obtained in Canada and Uganda.
Results and outcomes: Successful treatment outcomes after two years in both the community and hospital
cohorts were high. All-cause mortality was similar in both cohorts. However, community-based patients were more
likely to achieve viral suppression and had good adherence to treatment. The community-based program was
slightly more cost-effective. Per capita costs in both settings were unsustainable, representing more than Uganda’s
Primary Health Care Services current expenditures per person per year for all health services. The unpaid
community volunteers showed high participation and low attrition rates for the two years that this program was
evaluated.
Challenges and successes: Key successes of this study include the demonstration that antiretroviral therapy can
be provided in a rural setting, the creation of a research infrastructure and culture within Kabarole’s health system,
and the establishment of a research collaboration capable of enriching the global health graduate program at the
University of Alberta. Challenging questions about the long-term feasibility and sustainability of a community-based
ARV program in Uganda still remain.
The partnership: This project is a continuation of previous successful collaborations between the School of Public
Health of Makerere University, the School of Public Health of University of Alberta, the Kabarole District
Administration and the Kabarole Research and Resource Center.
* Correspondence: walter.kipp@ualberta.ca
† Contributed equally
1School of Public Health, University of Alberta, 3-50 University Terrace, 8303-
112 Street, Edmonton, AB, Canada, T6G 2T4
Full list of author information is available at the end of the article
Kipp et al. BMC International Health and Human Rights 2011, 11(Suppl 2):S12
http://www.biomedcentral.com/1472-698X/11/S2/S12
© 2011 Kipp et al; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.

Page 2

Background
HIV/AIDS is a major global public health problem
which particularly affects countries in sub-Saharan
Africa [1,2]. Globally, there are 33 million persons living
with HIV infection (PLWHIV). Uganda, with one mil-
lion PLWHIV, ranks as a high HIV prevalence country.
The increased availability of antiretroviral treatment
(ART) for HIV/AIDS in low income countries has chan-
ged the clinical management of PLWHIV and has
dramatically increased rates of survival.
Improved availability of ART is intended to benefit all
PLWHIV eligible for treatment. However, in reality ART
services in developing countries are mainly offered in
urban-based hospitals. Thus poor patients living in rural
areas cannot access them due to economic and geo-
graphic constraints. This inequity in providing ART
services is a crucially important issue for health care
policy makers and practitioners. In Uganda, the roll-out
of ART services to rural areas, where 80% of the popula-
tion lives, cannot be accomplished by health care workers
alone. This is due to the extreme shortage of trained
health professionals, which currently remains the biggest
single obstacle to the expansion of ART services in Ugan-
dan and all of sub-Saharan Africa. This health personnel
shortage is the most acute in rural areas of sub-Saharan
Africa. Therefore, the roll-out of ART services to rural
populations requires alternative approaches, one of
which could be the engagement of communities and
community resources.
Various treatment models involving community
resources have been developed and tested in Uganda.
These include a home-based treatment program in Tor-
oro [3], a home-based treatment program in Jinja [4] and
a home-based treatment program in Kampala, which
used PLWHIV on treatment as volunteers for outreach
[5]. All three programs have shown that home-based
treatment is as good as hospital-based treatment. In gen-
eral, however, the literature reveals that there are limited
experiences with community-based ART programs.
The research
The objectives of our intervention study were: 1) to
assess the effectiveness of a rural community-based ART
program in Rwimi subcounty, Kabarole district; and 2) to
compare the treatment outcomes and mortality in the
rural community-based ART program with a well estab-
lished hospital-based ART program offered at a best
practice urban hospital in Kabarole district. The study
was undertaken in Rwimi subcounty (population 25 000)
which is located in the Kabarole district, western Uganda,
50 km away from the nearest hospital offering ART. This
study required less material input (e.g. bicycles vs. motor-
cycles) than other studies previously done in Uganda
which may increase the chance that this model can be
replicated using available health sector resources.
The study was a cohort study with a non-randomized
intervention design [6]. The study population consisted
of a cohort of PLWHIV receiving community-based ART
established in a health centre III (run by a clinical officer)
in a rural sub-county in Kabarole district and the com-
parison population consisted of a cohort of PLWHIV
receiving ART from the best practice hospital (run by
physicians) located in an urban area in the district. In
each cohort, all HIV patients presenting themselves to
the hospital and the rural clinic for treatment were
sequentially recruited after the project commenced. Only
those patients were enrolled who were eligible for treat-
ment according to the Ugandan HIV treatment guide-
lines. Each cohort was followed up for two years to
compare mortality and ART outcomes between the
cohorts. In the community-based cohort 185 patients
were sequentially recruited and in the hospital-based
cohort 200 patients were recruited. Several sub-studies
were conducted, e.g. gender impact on treatment results,
quality of life of community-based patients on treatment,
cost-effectiveness of community-based and hospital-
based treatment models.
Forty-one lay community volunteers agreed to partici-
pate in the pilot and were provided with training on
ART. The motivation of the unpaid volunteers was based
on an intrinsic desire to support access to treatment for
their fellow community members and was reinforced
through the recognition and support they received from
the health care program and the community. Volunteers
were asked to make weekly visits to their patients, during
which time they monitored adherence to treatment,
through pill counts, as well as assessed the presence of
adverse reactions. Volunteers were asked to refer those
with clinical problems and/or adverse reactions to ARVs
to the clinical officer at the health centre. On a monthly
basis the volunteers obtained a supply of ARVs from the
local health centre and delivered these to their assigned
patients. At the time of recruitment, patients were also
asked to identify a family member/friend as their daily
treatment supporter to help with the daily intake of the
drugs. The first line antiretroviral drugs, provided free of
charge by the Ugandan national HIV/AIDS program,
consisted of stavudine, lamivudine, and nevirapine.
Outcome measures of the study were: 1) mortality, HIV-
1 RNA viral load (VL) per ml and increase in CD4 cell
count/µl which were assessed after two years of starting
ART. Statistical procedures included descriptive, univari-
ate and multivariate analysis. Logistic regression models
assessed the association between successful treatment out-
come (VL<400 copies/ml) and clinical and demographic
variables. The AIDS-related survival of hospital and
Kipp et al. BMC International Health and Human Rights 2011, 11(Suppl 2):S12
http://www.biomedcentral.com/1472-698X/11/S2/S12
Page 2 of 5

Page 3

community patients was also described. Ethics approval
was provided in Canada by the University of Alberta’s
Health Research Ethics Board and in Uganda by the
School of Public Health, Makerere University, Kampala
and the Uganda National Council of Science and Technol-
ogy, Kampala. Each participant was informed about the
study and signed a consent form. Limitations were: 1) the
study was not designed with a randomized study design
because randomization simply was not feasible; 2) the loss
of patients for follow-up, including due to mortality, was
relatively high (30% in the community-based cohort vs
29% in the hospital-base cohort).
Results and outcomes
No patient in the study was reported to have had severe
adverse reactions which required a change to second-line
treatment. Seventeen patients in the community-based
cohort developed a skin rash (6 patients in the hospital-
based cohort) and 25 community-based patients reported
suffering from mild peripheral neuropathy (50 in hospital-
based patients) which did not warrant a change in medica-
tion. Successful treatment outcomes after two years in
both the community and hospital cohorts were high. How-
ever, in the community-based cohort, HIV patients were
more likely to achieve viral suppression. CD4+ cell counts
increased similarly in both cohorts. All cause mortality
was similar in community and hospital patients after the
two year time interval. Community-based patients had
excellent overall adherence to treatment at rates through-
out the study period (adherence was not consistently mea-
sured in the hospital cohort). Almost all patients on
treatment in the community-based cohort reported a sig-
nificant increase in their overall quality of life, as measured
by a standardized health-related quality of life (HRQL)
questionnaire [7].
Women benefited more from the community-based
ART program, as more women sought treatment than
men (58.3% were females). In addition, in the first six
months, women had better treatment outcomes in terms
of successful suppression of the VL compared to men
[8,9]. However, this difference in the treatment outcome
between men and women decreased after two years and
was not statistically significant in the longer term. An
indirect benefit for women and their partners was that all
seven children born to treated women in the community
cohort were HIV negative after testing, sparing them and
their families the additional burden of caring for an HIV-
infected infant.
When program costs were calculated for both, the hos-
pital-based and community-based treatment model, the
cost-effectiveness per successful treated HIV patient was
slightly better in community-based HIV patients. How-
ever, the overall program costs per year per patient in
both programs (approximately 100 US$ including the cost
of drugs) were at a level that could not be sustained by
local Primary Health Care (PHC) Services in Uganda
which currently spends on average between 10-20 US$ per
person year for all health services. This suggests that the
expansion and long-term maintenance of ART programs
will require long-term external funding.
The ART treatment model with unpaid volunteers and
clinical officers/nurses in charge of serving patients
referred by the volunteers has shown to be sustainable
for the two year period that this program was evaluated.
This is best illustrated by the low attrition of volunteers,
as only two volunteers left the program. The recognition
efforts by our project and the community were obviously
felt by the volunteers to be valuable and very positive.
The low attrition and the effective participation of ordin-
ary community members as unpaid volunteers led us
believe that this model can be sustained, if the required
resources for their supervision, recognition and motiva-
tion can be provided. In contrast to popular belief that
volunteer programs are cheap, our cost calculations
showed that this is not the case, even though the cost-
effectiveness per successful treatment was slightly better
in the community treatment model compared to the hos-
pital model.
Challenges and successes
The main success of this study was the demonstration
that ART can be provided to rural HIV patients closer to
where they live. Our community-based treatment model
involved unpaid community volunteers, supported by
non-cash incentives, training and supervision and who
were connected to a local health centre staffed by clinical
officers and nurses. The treatment outcomes of this
model were equivalent (and even slightly superior) to the
treatment outcomes of ART in the best-practice regional
hospital in the district. As ART was delivered through
the existing rural Ugandan health care system with mini-
mal additional material and human resources, it can be a
feasible model for replication on a larger scale, e.g. dis-
trict wide. Another success was that a research infra-
structure and culture of research has been created
amongst the health system leadership in Kabarole district
and a cadre of trained and capable research assistants has
been established. This has already led to new applied
research initiatives of the Kabarole District Health Man-
agement Team (DHMT) to generate new information
required for better addressing the major health problems
in the district such as malaria, tuberculosis, and repro-
ductive health. Finally, the foundation of these collabora-
tions and the field site in Uganda was instrumental in the
establishment of the first graduate program in global
health in Canada at the University of Alberta in 2005.
Despite these successes, some challenges remain,
related to the longer term sustainability of the program.
Kipp et al. BMC International Health and Human Rights 2011, 11(Suppl 2):S12
http://www.biomedcentral.com/1472-698X/11/S2/S12
Page 3 of 5

Page 4

Questions that are raised and need to be answered
include:
1) Can the DHMT in Kabarole continue with this
program successfully on its own?
2) If this program is to be replicated, what are the fac-
tors to be considered to influence successful implemen-
tation on a larger district or region-wide scale and
which modifications have to be made from the original
smaller pilot program?
The partnership
This project is a continuation of previous successful colla-
borative HIV/AIDS work in Uganda by the School of
Public Health, Makerere University (MU), the School of
Public Health, University of Alberta (UofA), the Kabarole
Health Department (KHD) and the Kabarole Research and
Resource Center (KRC). This partnership started in 1997
between the Department of Public Health Sciences at the
UofA and the Department for Epidemiology and Biostatis-
tics, Institute of Public Health, MU. The role of MU and
UofA was to collaboratively design the study, apply for
funding and supervise the project implementation in
Uganda. The KHD’s role was to implement the grant
under the supervision from MU and UofA which was
somewhat unique to a research project, since the research
implementation team usually consists of researchers and
not health care workers. We decided to implement this
study in the context of a rural health clinic with involve-
ment of the local and district stakeholders in health care.
Our hypothesis was that if the research project was already
integrated in a practical health care setting with local
involvement, it would be much easier for the health care
workers to use the research results. Our aim from the
beginning of the study was that the intervention, if suc-
cessful, should be continued as the method of service pro-
vision to guarantee future ART to all patients enrolled.
With this study design of integrating the research project
into the local health services, we felt confident that this
project could be more easily handed over with better pro-
spects of successful continuation through the local Pri-
mary Health Care services. The special role of the KRC,
whose mandate in the district is to plan and deliver on
community-based projects for the underserved population,
was to support the role of the community volunteers and
the community itself in the ownership of the project. This
concept of integration distinguishes our research project
from many other research projects which don’t strive for a
continuation after research funding has ended.
In the past our Ugandan-Canadian collaboration has
been very successful in attracting funding from diverse
funding agencies (Canadian International Development
Agency (CIDA), the International Development and
Research Centre (IDRC), the Canadian Institutes of Health
Research (CIHR) and the Government of Alberta) with
some $3 million obtained in research and development
funding. We were diligent from the beginning to establish
a multi-disciplinary research team which would enable us
to cover a wide range of perspectives in community-
based/population-based research initiatives. Disciplines we
included were global health, health administration, public
health, nursing, epidemiology and biostatistics, clinical
(tropical) medicine, social sciences, gender equality, health
economy, adult education, communication, etc. We also
tried as much as possible to include qualitative and quali-
tative research methods in our proposals.
The basis for the successful joint work of this partner-
ship has been the mutual respect and trust among the var-
ious collaborators which was built over the years through
substantive and persistent efforts. Important was, espe-
cially at the beginning, the development of personal rela-
tionships which was only possible through on site visits
and practical collaborative work. Effective and regular
communication, which we did often in person and by
phone, is a must. It is very difficult in our opinion to
establish research collaborations only by electronic com-
munication. If personal relationships have been estab-
lished, then online communication works much better and
is feasible. Our collaboration over the years was very
unproblematic with only a few minor incidents. For our
volunteers in this project, the liaison between themselves,
their communities and an external foreign institution like
the UofA was particularly appreciated, which contributed
to their high motivation and low drop out rate (only two
volunteers have left the program to date after four years).
This strong sentiment around the value of external rela-
tions was expressed by the volunteers in many informal
encounters with us and also in formal research results.
Our experience in this and other research/development
projects in sub-Saharan Africa has shown us that the con-
tinued external links such as ours (which may include only
very limited funding) are indeed a crucial ingredient to a
sustained partnership and realized to project outcomes.
Another way to link developing and developed countries
could be the establishment of formal or informal partner-
ships between Canadian (external) and developing country
towns, schools, and other interested groups. There is a
huge potential and interest in Canada for this which has
not been fully explored and which needs to be further
developed and harvested.
This experience leads us believe that an ongoing long
term partnership model for development is what is
needed. This is in contrast to the development model in
many international development agencies which favours
a time limited collaboration and a so called “handing
over” of project activities to the developing partner with-
out further financial support. It is well known that this
short-term model has failed in many development/
research projects, since projects “handed over” to the
Kipp et al. BMC International Health and Human Rights 2011, 11(Suppl 2):S12
http://www.biomedcentral.com/1472-698X/11/S2/S12
Page 4 of 5

Page 5

local developing country partner disintegrate very fast
and often completely disappear. However, it is obvious
that an ongoing partnership between developed and
developing country institutions poses several challenges.
One is the continued necessity for a research-based orga-
nization to raise funds for this type of partnership which
absorbs a lot of working capacity and time. In our case
we have maintained our collaboration for 14 years which
is much beyond the usual duration of development pro-
jects funded by bilateral international funding agencies.
Another challenge is to tackle research topics of mutual
benefit that are directly relevant to services provided at
the grassroots level. A third collaborative partnership
challenge is always the capacity building component. For
us it involved training of local health care workers and
local researchers and Canadian graduate students. Over
time, more than 30 graduate students from UofA partici-
pated in this or other collaborative projects and success-
fully completed their thesis or their practicum. All of the
thesis students have published their study results in
respected peer reviewed journals. Many of their findings
have been used by the KHD and have influenced service
delivery and policy. Some of them are now professionals
in development organizations in Canada and other
countries.
Finally, based on our success to date in this collaborative
project as well as others, we have found academic institu-
tions to be ideal research partners for service delivery
agencies in developing countries. With shared goals and
mutually desired outcomes, both partners compliment the
skill set and contributions of each other to enhance essen-
tial services in countries where they are desperately
needed. The synergy that is created in this collaborative
process helps to overcome all the challenges.
List of abbreviations used
ART: antiretroviral treatment; CD4: cluster of differentiation 4; DHMT: district
health management team; HIV: human immunodeficiency virus; AIDS:
acquired immune deficiency syndrome; PHC: primary health care; PLWHIV:
persons living with HIV infection; RNA: ribonucleic acid; VL: viral load.
Acknowledgements
We are indebted to Jean Kipp for her comments on the first draft. We are
grateful for the tireless efforts of all the volunteers and health care workers
from Rwimi sub-county in supporting access to treatment for their fellow
community members. The study was financed through a grant from the
Canadian Institutes of Health Research (CIHR), grant No. MOP-74586.
This article has been published as part of BMC International Health & Human
Rights Volume 11 Supplement 2, 2011: Global health research case studies:
lessons from partnerships addressing health inequities. The full contents of
the supplement are available online at http://www.biomedcentral.com/1472-
698X/11?issue=S2.
Author details
1School of Public Health, University of Alberta, 3-50 University Terrace, 8303-
112 Street, Edmonton, AB, Canada, T6G 2T4.2School of Public Health,
Makerere University, Kampala, Uganda.3District Health Department, Kabarole
District Local Government, Fort Portal, Uganda.
Authors’ contributions
WK was involved in all stages of the study and wrote the article. JKL had
major input in the development of the proposal and execution of the study
as well as interpretation of study results. He also commented on the draft
manuscript. TR was involved in the design of the study and supervised the
field work in Uganda. He also provided input into the manuscript and
helped to interpret the study results within the context of the Ugandan
health care system. JO was part of the proposal writing and supervised the
field work. He was involved in the interpretation of the study findings and
suggestions on how they could be used for the expansion and scaling-up in
Kabarole District. AA was involved in all stages of the proposal development
and implementation of the study, supervision of the field work and gave
major input into the manuscript and the interpretation of the data. He
played a major part in the statistical data analysis.
Competing interests
All authors declare no conflict of interest.
Published: 8 November 2011
References
1.Hladik W, Musinguzi J, Kirungi W, et al: The estimated burden of HIV/AIDS
in Uganda, 2005-2010. AIDS 2008, 22(4):503-10.
2.Barnighausen T, Bloom DE, Humair S: Human resources for treating HIV/
AIDS: needs, capacities and gaps. AIDS Pat Care STDs 2007, 21(11):799-812.
3.Weidle PJ, Malamba S, Mwebaze R, et al: Assessment of a pilot
antiretroviral drug therapy programme in Uganda: patients’ response,
survival, and drug resistance. Lancet 2002, 360(9326):34-40.
4.Jaffar S, Amuron B, Foster S, et al: Rates of virological failure in patients
treated in a home-based versus a facility-based HIV-care model in Jinja,
southeast Uganda: a cluster-randomized equivalence trial. Lancet 2009,
374(9707):2080-89.
5.Chang L, Alamo S, Guma S, et al: Two-year virologic outcomes of an
alternative AIDS care model: Evaluation of a peer health worker and
nurse-staffed community-based program in Uganda. J Acquir Immune
Defic Syndr 2009, 50(3):276-82.
6. Kipp W, Konde-Lule J, Saunders D, et al: Antiretroviral treatment in
western Uganda: Comparing treatment outcomes and patient mortality
between a prospective community-based and hospital-based cohort.
Curr HIV Res 2010, 8(2):179-85.
7. Alibhai A, Martin L, Kipp W, Konde-Lule J, Saunders D, Rubaale T,
Houston S, Okech-Ojony J: Quality of life of HIV patients in a rural area of
western Uganda: Impact of a community-based antiretroviral treatment
program. Curr HIV Res 2010, 8:370-378.
8. Alibhai A, Kipp W, Saunders LD, Senthilselvan A, Kaler A, Houston S, Konde-
Lule J, Okech-Ojony J, Rubaale T: Gender-related mortality in a hospital
and community-based antiretroviral treatment program for HIV-infected
patients in rural Uganda. Int J Women’s Health 2010, 2:45-52.
9.Kipp W, Alibhai A, Saunders DL, Senthilselvan A, Kaler A, Konde-Lule J,
Okech-Ojony J, Rubaale T: Gender differences in antiretroviral treatment
outcomes of HIV patients in rural Uganda. AIDS Care 2010, 12(3):271-278.
doi:10.1186/1472-698X-11-S2-S12
Cite this article as: Kipp et al.: Comparing antiretroviral treatment
outcomes between a prospective community-based and hospital-based
cohort of HIV patients in rural Uganda. BMC International Health and
Human Rights 2011 11(Suppl 2):S12.
Kipp et al. BMC International Health and Human Rights 2011, 11(Suppl 2):S12
http://www.biomedcentral.com/1472-698X/11/S2/S12
Page 5 of 5