HIV-1 Incidence and HIV-1 Associated Mortality in a Cohort of Urban Factory Worker in Tanzania

Abstract

To determine HIV-1 incidence and HIV-1 associated mortality in a prospective cohort study. To determine whether the cohort is suitable for studies aiming to determine the impact of interventions on HIV-1 incidence.
The study population was a cohort of 1772 urban factory workers (1478 men and 294 women) in northwest Tanzania. The study took place from October 1991 to September 1993. Outcome measures were HIV-1 seroconversion and death.
HIV-1 incidence was 1.2 (95% CI 0.7-2.0) per 100 person-years (pyr). Crude annual mortality was 4.9 per 100 pyr in those with and 0.3 in those without HIV-1 infection, giving an age and sex adjusted mortality ratio of 12.9 (95% CI 5.4-30.7). Of all deaths, 62% were attributable to HIV-1 infection.
HIV-1 infection was a major public health problem, being the major cause of death in this adult population. At an HIV-1 incidence of 1.2 per 100 pyr, a large cohort size would be required to evaluate the impact of interventions on HIV-1 incidence.

Full text

Genitourin
Med
1995;71:212-215
HIV-
1
incidence
and
HIV-
1
associated
mortality
in
a
cohort
of
urban
factory
workers
in
Tanzania
M
W
Borgdorff,
L
R
Barongo,
A
H
Klokke,
J
N
Newell,
K
P
Senkoro,
J
P
Velema,
R
M
Gabone
Abstract
Objective-To
determine
HIV-1
inci-
dence
and
HIV-1
associated
mortality
in
a
prospective
cohort
study.
To
deterniine
whether
the
cohort
is
suitable
for
studies
aiming
to
determine
the
impact
of
inter-
ventions
on
HIV-1
incidence.
Methods-The
study
population
was
a
cohort
of
1772
urban
factory
workers
(1478
men
and
294
women)
in
northwest
Tanzania.
The
study
took
place
from
October
1991
to
September
1993.
Outcome
measures
were
HIV-1
serocon-
version
and
death.
Results-HIV-l
incidence
was
1'2
(95%
CI
0.7-2.0)
per
100
person-years
(pyr).
Crude
annual
mortality
was
4 9
per
100
pyr
in
those
with
and
0.3
in
those
without
HIV-1
infection,
giving
an
age
and
sex
adjusted
mortality
ratio
of
12*9
(95%
CI
5.4-30.7).
Of
all
deaths,
62%
were
attrib-
utable
to
HIV-1
infection.
Conclusion-HIV-1
infection
was
a
major
public
health
problem,
being
the
major
cause
of
death
in
this
adult
population.
At
an
HIV-1
incidence
of
1.2
per
100
pyr,
a
large
cohort
size
would
be
required
to
evaluate
the
impact
of
interventions
on
HIV-1
incidence.
(Genitourin
Med
1995;71:212-215)
National
Institute
for
Medical
Research,
Mwanza,
Tanzania
L
R
Barongo
J
N
Newell
K
P
Senkoro
R
M
Gabone
Bugando
Medical
Centre,
Mwanza,
Tanzania
A
Klokke
Royal
Tropical
Institute,
Amsterdam,
The
Netherlands
M
W
Borgdorff
Nijmegen
Institute
for
International
Health,
Nijmegen
University,
Nijmegen,
The
Netherlands
J
P
Velema
Correspondence
to:
Dr
M
W
Borgdorff,
National
Institute
of
Public
Health
and
Environmental
Protection,
Centre
for
Infectious
Disease
Epidemiology
(CIE),
P
0
Box
1,
3720
BA
Bilthoven,
The
Netherlands
Accepted
for
publication
1
April
1995
Keywords:
cohort;
incidence;
HIV-1;
mortality;
Tanzania
Introduction
Measuring
the
incidence
of
HIV
infection
is
necessary
to
determine
the
rate
of
HIV
trans-
mission.
In
addition,
when
studies
are
planned
to
evaluate
the
efficacy
of
interven-
tions
to
reduce
HIV
transmission,
for
instance
through
condom
promotion,
control
of
sexu-
ally
transmitted
diseases
(STDs)
or
in
future
through
immunisation
against
HIV-1,
a
rea-
sonable
estimate
of
the
incidence
of
HIV
infection
is
required
for
the
determination
of
the
size
of
the
sample
needed
for
the
study.'
However,
there
is
limited
information
on
the
incidence
of
HIV-1
infection
in
Africa,2-'5
probably
because
incidence
can
only
be
mea-
sured
in
longitudinal
studies
which
are
diffi-
cult
and
costly
to
perform.'6
HIV-1
infection
has
been
shown
to
con-
tribute
substantially
to
mortality
in
hospitals
in
Africa.'7-23
An
obvious
limitation
of
these
studies
is
their
selection
bias,
as
in
Africa
many
people
die
at
home.
Evidence
from
pop-
ulation-based
studies
is
starting
to
emerge:
in
Rwanda
and
Uganda
HIV-1
associated
mor-
tality
was
shown
to
be
substantial.'52425
Quantification
of
HIV-1-associated
mortality
is
important
for
assessment
and
prediction
of
the
demographic
and
socio-economic
impact
of
the
epidemic.2
28
A
cohort
study
among
urban
factory
workers
in
Mwanza,
Tanzania,
was
started
in
October
1991
in
order
to
determine
the
incidence
of
and
risk
factors
for
HIV-1
infection
and
other
STDs,
describe
changes
in
sexual
high
risk
behaviour
following
an
intervention,
and
describe
the
natural
history
of
HIV-1
infec-
tion.
The
prevalence
of
and
risk
factors
for
HIV-1
infection
at
intake
of
the
cohort
were
found
to
be
similar
to
those
in
the
general
urban
population2930
and
risk
behaviour
was
found
to
be
common:
having
multiple
sexual
partners
in
the
past
month
was
reported
by
22%
of
male
and
5%
of
female
workers,
while
condoms
were
used
by
3%
only.3'
In
this
paper
we
describe
the
HIV-1
inci-
dence
and
HIV-1
associated
mortality
in
this
cohort
from
October
1991
to
September
1993.
Methods
Study
population
The
study
population
comprised
workers
at
a
large
urban
factory
with
a
work
force
of
2038
workers
(1706
men,
332
women).
Enrolment
started
in
October
1991
and
continued
throughout,
both
for
new
employees
and
those
initially
declining
to
participate.
A
study
clinic
was
created
at
the
factory
to
supplement
an
existing
clinic,
as
the
latter
was
too
small
to
cope
with
additional
activities.
The
study
population
was
advised
that
the
aim
of
the
study
was
to
determine
the
health
status
of
the
workers,
with
special
interest
in
sexually
transmitted
diseases
and
HIV-1
infection
and
they
were
invited
to
present
themselves
for
registration
at
the
study
clinic,
on
a
voluntary
basis.
The
study
was
in
accordance
with
national
guidelines
on
HIV
testing
and
was
approved
by
the
ethical
committee
of
the
National
Institute
for
Medical
Research.
The
data
presented
here
cover
a
study
period
of
two
years,
from
1
October
1991
to
30
September
1993.
Data
collection
techniques
All
respondents
were
interviewed,
physically
examined
and
underwent
laboratory
investi-
gations
at
intake
and
at
four-monthly
intervals
thereafter.
Free
medical
treatment
was
provided
to
all
study
participants
at
intake,
212

HIV-1
incidence
and
mortality,
Tanzania
follow-up
visits,
and
any
time
they
presented
in
between.
Where
necessary,
patients
were
referred
to
hospital
for
further
investigation
or
treatment.
The
interview
took
place
in
a
private
room
for
30
to
45
minutes
in
Kiswahili
by
trained
interviewers,
using
a
structured
and
pre-
coded
questionnaire.
Assistant
medical
offi-
cers
carried
out
the
physical
examination,
using
a
pre-coded
form.
In
the
study
clinic
laboratory
10
ml
of
venous
blood
was
taken.
Serological
testing
for
HIV-1
antibodies
took
place
in
the
reference
laboratory
of
the
zonal
hospital
with
the
Vironostika
anti-HTLV-III
ELISA
(Organon,
Boxtel,
The
Netherlands).
All
reactive
and
weakly
reactive
samples
were
tested
with
Western
Blot
(Organon,
Epitope,
Beaverton,
Oregon,
USA).
The
Western
Blot
was
considered
positive
if
at
least
two
of
the
gp41,
gp120,
and
gp160
bands
were
pre-
sent.32
A
random
sample
of
10%
of
sera
was
retested
on
ELISA
for
internal
quality
con-
trol.
Western
Blot
band
readings
were
recorded
individually
and
strips
were
kept
for
inspection.
All
sera
of
individuals
who
were
not
HIV-1
positive
or
negative
throughout
the
study
period
were
tested
by
Western
Blot
as
well.
No
sero-retroconversions
(from
HIV-1
positive
to
HIV-1
negative)
were
observed.
Comparison
of
200
serology
results
in
1991
at
the
zonal
laboratory
with
those
of
Nijmegen
University
showed
a
100%
agreement.
Pre-
and
post-HIV-
1-test
counselling
was
offered
to
all.
After
pre-test
counselling
a
sep-
arate
blood
sample
was
taken
for
HIV-1
testing
from
those
wishing
to
know
their
HIV-1
test
result.
The
HIV-1
results
were
made
known
to
the
counsellor
and
through
the
counsellor
to
the
individual
concerned
only.
Other
staff
at
the
factory
clinic
were
blind
to
the
HIV-1
status
of
individuals.
Deaths
among
study
participants
were
identified
by
a
follow-up
worker
tracing
non-
attenders
and
through
the
welfare
officer
at
the
factory
responsible
for
paying
an
allowance
to
the
relatives
of
the
deceased.
The
latter
officer
knew
of
all
deaths
identified,
suggesting
that
his
registration
of
deaths
was
comprehensive.
The
household
of
the
deceased
was
visited
by
an
assistant
medical
officer
two
to
six
months
after
death
to
inter-
view
the
person
who
had
taken
care
of
the
deceased
before
his
or
her
death.
If
this
per-
Table
1
Incidence
of
HIV-I
infection
in
a
cohort
of
1567
non-HIV-I
infected
urban
factory
workers
Number
Number
of
Person-years
Incidence
Age
group
enrolled
of
sero-
(pyr)
of
/1
00
pyr
(years)
(HIV-)
conversions
follow-up
(95%
CI)
Males
15-25
340
1
211-7
0.5
(00-2.6)
25-34
409
5
342-8
1-5
(0
5-3.4)
35-44
411
7
433-1
1-6(07-33)
45
+
165
2
183-3
1-1
(0-1-3-9)
All
males
1325
15
1171-0
1-3
(0
7-2-1)
Females
15-25
105
0
64-7
0.0
(00-5
5)
25-34
97
2
84-3
24
(0
3-83)
35-44
31
0
38-0
00
(00-93)
45
+
9
0
8-0
00
(00-369)
Allfemales
242
2
195-0
1.0
(0
1-3.7)
Total
1567
17
1365-9
1-2
(0.7-2.0)
son
was
not
available
the
household
would
be
visited
again
up
to
a
total
of
three
visits.
Data
collection
tools
used
were
open-ended
ques-
tions,
a
structured
questionnaire,
and
a
death
notification
form
if
available.
For
those
who
died
in
hospital,
hospital
records
were
traced
in
order
to
get
the
hospital
diagnosis.
The
probable
cause
of
death
was
determined
from
hospital
records
if
available
and
otherwise
by
consensus
of
a
group
of
three
medical
officers,
including
one
specialist
in
internal
medicine,
blinded
to
HIV-1
status
of
the
deceased.
Data
analysis
Two
sets
of
person-years
of
follow-up
were
calculated.
For
HIV-1
incidence
follow-up
included
the
time
period
between
the
first
and
last
negative
HIV-1
test
result
plus
for
sero-
converters
half
the
time
period
between
the
last
negative
and
first
positive
HIV-1
result.
For
mortality
the
follow-up
period
included
the
time
from
the
first
HIV-1
result
until
the
end
of
follow-up;
the
end
of
follow-up
was
defined
as
the
date
of:
(1)
death
or
leaving
the
factory,
(2)
one
year
after
the
last
HIV-1
result
or
(3)
30
September
1993,
whichever
came
first.
Deaths
were
included
if
they
occurred
inside
the
defined
follow-up
period.
The
population
attributable
risk
was
calcu-
lated
as
the
difference
of
the
mortality
rate
in
the
total
population
and
the
mortality
rate
in
those
without
HIV-1
infection
and
the
popu-
lation
aetiological
fraction
as
the
ratio
of
the
population
attributable
risk
and
the
mortality
rate
in
the
total
population.
Results
Of
the
2038
factory
workers
1478/1706
(87%)
men
and
294/332
(89%)
women
had
enrolled
in
the
study.
A
total
of
471
(27%)
of
these
were
lost
to
follow-up
by
the
end
of
the
study
period.
Defaulting
was
not
associated
with
sex
or
HIV-1
infection
(data
not
shown),
but
was
higher
in
those
aged
below
35
years
(317/1079
=
29%)
than
in
those
aged
35
years
and
over
(154/693
=
22%)
(X2
10-7;
p
<
0.01).
The
rate
of
defaulting
was
not
associ-
ated
with
the
time
of
enrolment:
in
the
two
groups
enrolling
before
and
after
26
weeks
the
rate
was
19
per
100
person-years
of
follow-up.
At
intake
153/1478
(10.4%)
men
and
52/294
(17.7%)
women
were
HIV-1
infected.
In
the
study
period
17
seroconversions
took
place
in
1365.9
person
years
of
follow-up
giving
an
HIV-1
incidence
rate
of
1.
2
per
100
per-
son-years
of
follow-up
(95%
CI
0-7-2.0).
No
association
was
found
between
seroconversion
and
age
or
sex;
the
power
of
the
study
to
detect
such
associations
was
limited
as
the
number
of
seroconversions
was
small
(table
1).
The
crude
annual
mortality
rate
was
4.9
per
100
person-years
in
those
with
and
0.3
in
those
without
HIV-1
infection
(table
2).
The
age
and
sex
adjusted
mortality
ratio
was
12-9
(95%
CI
5.4-30
7).
The
age
and
sex
adjusted
population
attributable
risk
was
0.5
per
100
person-years,
and
the
population
etiological
fraction
62%.
Of
the
14
HIV-1
infected
people
who
died,
213

Borgdorff,
Barongo,
Klokke,
Newell,
Senkoro,
Velema,
Gabone
Table
2
Mortality
rates
in
an
urban
factory
in
205
workers
with
and
1567
workers
without
HIV-1
infection
HIV-I
infected
Non-HIV-I
infected
Person-
Mortality
Person-
Mortality
Age
group
Number*
years
of
rate
Number*
years
of
rate
(years)
Enrolled
Died
follow-up
(I
OO
pyr)
(95%
CI)
EnroUed
Died
follow-up
(/I
OO
pyr)
95%
CI
Males
15-24
17
0
21-2
0-0
(00-17-4)
340
1
421-8
0-2
(0-0-1
3)
25-34
71
3
100-8
3.0
(06-9-2)
409
0
561-8
0
0
(0
0-0
0)
35-44
54
9
78-8
11-4
(5-5-21-4)
411
2
629-2
0-3
(0-0-1
1)
45
+
11
1
12-7
7-9
(0-6-43.3)
165
1
257-9
0-4
(00-2-1)
All
males
153
13
213-6
6-1
(3-1-10-4)
1325
4
1870-7
0-2
(0-1-05)
Females
15-24
14
0
17-8
0.0
(0
0-20.7)
105
0
132-1
0-0
(0-0-2-8)
25-34
26
0
35-1
0.0
(0
0-10
5)
97
3
131-1
2-3
(0
0-2.6)
35-44
11
1
17-0
59
(0.6-32.3)
31
0
51-1
00
(00-70)
45
+
1
0
1.9
00
(00-842)
9
0
13-3
0-0
(0-0-247)
All
females
52
1
71-9
1-4
(0-1-7
8)
242
3
327-5
0.9
(02
2-27)
Total
205
14
285-5
4-9
(26-8
2)
1567
7
2198-2
0.3
(0-1-07)
*HIV-1
status
is
given
at
enrolment.
Seroconverters
contribute
person-years
of
follow-up
to
the
groups
without
and
with
HIV-1
infection
before
and
after
seroconversion,
respectively.
nine
met
the
criteria
of
the
1987
revised
CDCtWHO
AIDS
case
definition33:
one
had
cryptococcal
meningitis
and
eight
HIV
wast-
ing
syndrome.
Two
others
had
had
weight
loss
and
fever,
but
the
evidence
was
inade-
quate
to
make
or
reject
the
diagnosis
of
AIDS.
The
remaining
three
without
an
AIDS
diag-
nosis
had
pulmonary
tuberculosis,
diarrhoea
and
pyomyositis,
respectively.
Discussion
In
this
cohort
of
urban
factory
workers
with
a
prevalence
of
HIV-1
infection
of
12%,
the
annual
HIV-1
incidence
rate
was
1.2%.
It
is
striking
that
at
an
HIV-1
prevalence
of
12%,
62%
of
deaths
in
adults
were
attributable
to
HIV-1
infection,
making
it
one
of
the
most
important
public
health
problems
in
this
urban,
adult
population.
The
incidence
rate
of
HIV-1
infection
in
this
study
was
similar
to
that
in
a
cohort
of
urban
health
workers
in
Zaire,34
and
rural
population
cohorts
in
Uganda,1415
but
much
lower
than
that
of
4.7%
in
an
urban
popula-
tion
in
a
neighbouring
Region
in
Tanzania,'1
which
also
had
a
much
higher
HIV-1
preva-
lence
(of
24%)
at
the
start
of
the
study
period.
Urban
women
of
childbearing
age
in
Rwanda
and
Zambia
also
had
higher
incidence
rates
of
3-4%.6
As
might
be
expected,
the
observed
HIV-1
incidence
was
much
lower
than
that
of
high
risk
groups
in
Africa
such
as
spouses
of
HIV-1
infected
people
7
10
prostitutes8
1213
and
attenders
of
sexually
transmitted
disease
clin-
ics.5
Low
HIV-1
incidence
rates
of
approxi-
mately
2%35
and
of
11
%36
were
observed
in
prostitutes
in
low
prevalence
areas,
though
in
the
latter
study
a
rapid
increase
in
incidence
rates
over
time
was
noted.36
Mortality
rates
in
those
with
and
without
HIV-1
infection
were
similar
to
those
reported
from
Rwanda,25
while
the
population
aetiological
fraction
in
the
present
study
of
62%
was
lower
than
that
in
Rwanda
(90%),
which
is
explained
by
the
difference
in
HIV-1
prevalence
(32%
in
Rwanda
and
12%
in
the
present
study).
In
Uganda,
mortality
rates
were
approximately
twice
as
high,
both
in
those
with
and
without
HIV-1
infection,1524
resulting
in
a
population
aetiological
fraction
of
48%
at
an
HIV-1
prevalence
of
8%.'5
The
population
attributable
risk
is
deter-
mined
not
only
by
the
prevalence
of
the
risk
factor
(HIV-1
infection)
but
also
by
the
rela-
tive
risk
or
mortality
rate
ratio.
As
might be
expected,
the
mortality
rate
ratios
in
these
three
studies
were
not
significantly
different:
they
were
reported
to
be
20-8
in
Uganda
and
12.9
in
the
present
study,
and
can
be
esti-
mated
at
approximately
28
in
Rwanda.
Confidence
intervals
are
particularly
wide
in
the
present
and
the
Rwanda
study,
because
of
the
small
numbers
of
deaths
in
non-HIV-1-
infected
people.
It
seems
likely
that
a
sizeable
proportion
of
deaths
(at
least
3/14
=
21%)
occurred
in
HIV-
1
infected
people
who
had
not
yet
developed
AIDS,
as
was
shown
in
Rwanda,9
Kenya,'9
and
among
intravenous
drug
users
in
the
U.S.,'37-39
although
because
of
data
limitations
a
precise
estimate
of
this
proportion
could
not
be
made.
Tuberculosis
was
diagnosed
in
2
out
of
14
deaths;
an
autopsy
study
in
Ivory
Coast40
suggests
that
a
much
larger
proportion
of
deaths
in
the
study
population
may
have
been
attributable
to
tuberculosis
than
is
evi-
dent
from
our
data.
The
study
has
a
number
of
limitations.
Although
total
enrolment
was
reasonable
at
87%
of
those
eligible,
losses
to
follow-up
were
considerable
at
19
per
100
person-years
of
fol-
low-up.
These
losses
to
follow-up
were
largely
due
to
two
factors:
many
study
participants
did
not
like
to
submit
blood
samples
repeat-
edly,
and
owing
to
economic
difficulties
the
factory
had
to
close
down
temporarily
in
the
second
year
of
the
study,
sending
staff
on
unpaid
leave.
The
first
factor
might
be
partly
overcome
by
more
intensive
follow-up
and
education;
the
latter
factor
however,
which
affected
in
particular
those
in
younger
age
groups,
is
much
more
serious,
and
may
if
repeated
jeopardise
the
continuation
of
this
cohort
study.
The
effect
of
these
losses
to
follow-up
may
be
an
underestimate
of
the
incidence
of
HIV-
1
infection,
if
the
incidence
is
highest
in
those
less
than
35
years
old.'4
The
effect
on
the
mortality
results
is
probably
limited,
as
fami-
lies
of
those
sent
on
unpaid
leave
still
qualified
for
receiving
an
allowance
if
a
worker
died.
214

HIV-1
incidence
and
mortality,
Tanzania
For
studies
aiming
to
show
that
preventive
interventions
reduce
the
incidence
of
HIV-1
infection,
a
large
sample
size
will
be
required
to
show
an
effect
at
a
baseline
incidence
of
1.2%.
For
instance,
if
a
vaccine
against
HIV-1
reduced
the
incidence
of
HIV-1
infection
by
50%
from
1.2
to
0.6
per
100
pyr,
a
sample
size
would
be
required
in
intervention
and
control
group
of
approximately
5500
pyr
each
to
have
a
90%
probability
of
showing
that
the
vaccine
is
effective.
In
conclusion,
HIV-1
infection
was
a
major
public
health
problem,
being
the
major
cause
of
death
in
this
adult
population.
At
an
HIV-1
incidence
of
1-2
per
100
pyr,
a
large
cohort
size
would
be
required
to
evaluate
the
impact
of
interventions
on
HIV-1
incidence.
We
thank
the
Principal
Secretary,
Ministry
of
Health
and
the
Director
General,
National
Institute
for
Medical
Research
for
permission
to
carry
out
the
study
and
to
publish
its
results.
We
are
grateful
to
Prof
A
S
Muller,
Prof
R
A
Coutinho,
and
Dr
J
T
Boerma
for
their
critical
comments
on
drafts
of
this
paper.
The
research
for
this
publication
was
financed
by
the
Netherlands'
Minister
for
Development
Cooperation,
Section
for
Research
and
Technology,
P.O.
Box
20061,
2500
EB,
The
Hague,
as
part
of
the
Tanzania-Netherlands
Research
Project
on
AIDS
and
HIV
Infection
in
Mwanza
Region.
Responsibility
for
the
contents
and
for
the
opinions
expressed
rests
solely
with
the
authors;
publication
does
not
constitute
an
endorsement
by
the
Netherlands'
Minister
for
Development
Co-operation.
1
Dixon
DO,
Rida
WN,
Fast
PE,
Hoth
DF.
HIV
vaccine
trials:
some
design
issues
including
sample
size
calcula-
tion.
Jf
Acquir
Immune
Defic
Syndr 1993;6:485-96.
2
Nkowane
BJ.
Prevalence
and
incidence
of
HIV
infection
in
Africa:
a
review
of
data
published
in
1990.
AIDS
1991;5(Suppl
l):S7-S15.
3
Mann
JM,
Francis
H,
Quinn
TC,
et
al.
HIV
seroincidence
in
a
hospital
worker
population.
Ann
Soc
Belg
Med
Trop
1986;66:245-50.
4
N'galy
B,
Ryder
RW,
Bila
K,
et
al.
Human
immunodefi-
ciency
virus
infection
among
employees
in
an
African
hospital.
N
EnglJ
Med
1988;319:1123-7.
5
Cameron
DW,
Simonsen
JN,
D'Costa
U,
et
al.
Female
to
male
transmission
of
human
immunodeficiency
virus
type
1:
risk
factors
for
seroconversion
in
men.
Lancet
1989;ii:403-7.
6
Hira
SK,
Mangrola
SG,
Mwale
C,
et
al.
Apparent
vertical
transmission
of
human
immunodeficiency
virus
type
1
by
breastfeeding
in
Zambia.
Jf
Pediatr
1990;117:421-4.
7
Kamenga
M,
Ryder
RW,
Jingu
M,
et
al.
Evidence
of
marked
sexual
behaviour
change
associated
with
low
HIV-1
seroconversion
in
149
married
couples
with
dis-
cordant
HIV-1
serostatus:
experience
at
an
HIV
coun-
selling
center
in
Zaire.
AIDS
1991;5:61-7.
8
Plummer
FA,
Simonsen
JN,
Cameron
DW,
et
al.
Cofactors
in
male-female
sexual
transmission
of
human
immunodeficiency
virus
type
1.
JInfect
Dis
199
1;163:233-9.
9
Allen
S,
Serufilira
A,
Bogaerts
J,
et
al.
Confidential
HIV
testing
and
condom
promotion
in
Africa.
JAMA
1992;
268:3338-43.
10
Allen
S,
Tice
J,
Perre
P
van
de,
et
al.
Effect
of
serotesting
with
counselling
on
condom
use
and
seroconversion
among
HIV
discordant
couples
in
Africa.
BMJ
1992;
304:
1605-9.
11
Killewo
JZJ,
Sandstrom
A,
Bredberg
Raden
U,
Mhalu
FS,
Biberfeld
G,
Wall
S.
Incidence
of
HIV-1
infection
among
adults
in
the
Kagera
Region
of
Tanzania.
Int
Jf
Epidemiol
1993;22:528-36.
12
Laga
M,
Manoka
A,
Kivuvu
M,
et
al.
Non-ulcerative
sexu-
ally
transmitted
diseases
as
risk
factors
for
HIV-1
trans-
mission
in
women:
results
from
a
cohort
study.
AIDS
1993;7:95-102.
13
Zekeng
L,
Feldblum
PJ,
Oliver
RM,
Kaptue
L.
Barrier
contraceptive
use
and
HIV
infection
among
high-risk
women
in
Cameroon.
AIDS
1993;7:725-31.
14
Wawer
MJ,
Sewankambo
NK,
Berkley
S,
et
al.
Incidence
of
HIV-1
infection
in
a
rural
region
of
Uganda.
BMJ
1994;308:171-3.
15
Mulder
DW,
Nunn
AJ,
Wagner
HU,
Kamali
A,
Kengeya-
Kayondo
JF.
HIV-1
incidence
and
HIV-1-associated
mortality
in
a
rural
Ugandan
population
cohort.
AIDS
1994;8:87-92.
16
Ryder
RW,
Piot
P.
Epidemiology
of
HIV-1
infection
in
Africa.
In:
P
Piot,
JM
Main
eds
Bailliere's
Clinical
Tropical
Medicine
and
Communicable
Diseases
1988;3:
13-29.
17
De
Cock
KM,
et
al.
AIDS-the
leading
cause
of
death
in
the
west
African
city
of
Abidjan,
Ivory
Coast.
Science
1990;249:793-6.
18
De
Cock
KM,
Porter
A,
Odehouri
K,
et
al.
Rapid
emer-
gence
of
AIDS
in
Abidjan,
Ivory
Coast.
Lancet
1989;
ii:408-11.
19
Gilks
CF,
Brindle
RJ,
Otieno
LS,
et
al.
Life-threatening
bacteraemia
in
HIV-1
seropositive
adults
admitted
to
hospital
in
Nairobi,
Kenya.
Lancet
1990;336:545-9.
20
Hassig
SE,
Perriens
J,
Baende
E,
et
al.
An
analysis
of
the
economic
impact
of
HIV
infection
among
patients
at
Mama
Yemo
Hospital,
Kinshasa,
Zaire.
AIDS
1990;4:
883-7.
21
Mbaga
JM,
Pallangyo
KJ,
Bakari
M,
Aris
EA.
Survival
time
of
patients
with
acquired
immune
deficiency
syn-
drome:
experience
with
274
patients
in
Dar-es-Salaam.
East
Afr
MedJ7
1990;67:95-9.
22
Nelson
AM,
Hassig
SE,
Kayembe
M,
et
al.
HIV-1
seropos-
itivity
and
mortality
at
University
Hospital,
Kinshasa,
Zaire,
1987.
AIDS
1991;5:583-6.
23
Muller
0,
Moser
R.
HIV-1
disease
in
a
Kampala
hospital
1985-89
(letter)
Lancet
1990;335:236-7.
24
Mulder
DW,
Nunn
AJ,
Kamali
A,
Nakiyingi
J,
Wagner
HU,
Kengeya-Kayondo
JF.
Two-year
HIV-1-associated
mortality
in
a
Ugandan
rural
population.
Lancet
1994;
343:989-90.
25
Lindan
CP,
Allen
S,
Serufilira
A,
et
al.
Predictors
of
mor-
tality
among
HIV-infected
women
in
Kigali,
Rwanda.
Ann
Intern
Med
1992;116:320-8.
26
Anderson
RM,
May RM,
Boily
MC,
Garnett
GP,
Rowley
JT.
The
spread
of
HIV-1
in
Africa:
sexual
contact
pat-
terns
and
the
predicted
demographic
impact
of
AIDS.
Nature
1991;352:581-9.
27
Anderson
RM.
Some
aspects
of
sexual
behaviour
and
the
potential
demographic
impact
of
AIDS
in
developing
countries.
Soc
Sci
Med
1992;34:271-80.
28
Gregson
S,
Garnett
GP,
Anderson
RM.
Is
HIV-1
likely
to
become
a
leading
cause
of
adult
mortality
in
sub-
Saharan
Africa?
J
Acquir
Immune
Defic
Syndr
1994;7:
839-52.
29
Barongo
LR,
Borgdorff
MW,
Mosha
FF,
et
al.
The
epi-
demiology
of
HIV-1
infection in
urban
areas,
roadside
settlements
and
rural
villages
in
Mwanza
Region,
Tanzania.
AIDS
1992;6:1521-8.
30
Barongo
LR,
Borgdorff
MW,
Newell
JN,
et
al.
Intake
of
a
cohort
study
of
urban
factory
workers
in
northwest
Tanzania:
risk
factors
for
HIV-1
infection.
Trop
Geograph
Med
1994;46:157-62.
31
Borgdorff
MW,
Barongo
LR,
Newell
JN,
et
al.
Sexual
part-
ner
change
and
condom
use
among
urban
factory
workers
in
northwest
Tanzania.
Genitourin
Med
1994;31:378-83.
32
World
Health
Organization.
Acquired
Immune
Deficiency
Syndrome
(AIDS)-Proposed
WHO
criteria
for
inter-
preting
results
from
Western
blot
assays
for
HIV-1,
HIV-2,
and
HTLV-I/HTLV-II.
Wkly
Epidemiol
Rec
1990;65:281-3.
33
World
Health
Organization.
Acquired
immunodeficiency
syndrome
(AIDS)-1987
revision
of
CDC/WHO
case
definition
for
AIDS.
Wkly
Epidemiol
Rec
1988;63:1-7.
34
Leroy
V,
Van
de
Perre
P,
Lepage
P,
et
al.
Seroincidence
of
HIV-1
infection
in
African
women
of
reproductive
age:
a
prospective
cohort
study
in
Kigali,
Rwanda,
1988-1992.
AIDS
1994;8:983-6.
35
Jama
Ahmed
H,
Omar
K,
Yusuf
Adan
S,
Mohamed
Guled
A,
Grillner
L,
Bygdeman
S.
Syphilis
and
human
immunodeficiency
virus
seroconversion
during
a
6-month
follow-up
of
female
prostitutes
in
Mogadishu,
Somalia.
IntJ
STD
AIDS
1991;2:119-23.
36
Kanki
PL,
Travers
KU,
MBoup
S,
et
al.
Slower
heterosex-
ual
spread
of
HIV-2
than
HIV-1.
Lancet
1994;343:
943-6.
37
Stoneburner
RL,
Des
Jarlais
DC,
Benezra
D,
et
al.
A
larger
spectrum
of
severe
HIV-1-related
disease
in
intravenous
drug
users
in
New
York
City.
Science
1988;242:916-9.
38
Selwyn
PA.
Injection
drug
use,
mortality,
and
the
AIDS
epidemic.
Am
J7
Public
Health
199
1;81:
1247-9.
39
Selwyn
PA,
Alcabes
P,
Hartel
D,
et
al.
Clinical
manifesta-
tions
and
predictors
of
disease
progression
in
drug
users
with
human
immunodeficiency
virus
infection.
N
Engl
J
Med
1992;327:1697-703.
40
Lucas
SB,
Hounnou
A,
Peacock
C,
et
al.
The
mortality
and
pathology
of
HIV
infection
in
a
West
African
city.
AIDS
1993;7:1569-79.
215