Methods in HIV Epidemiology

1.
You have your first job out of graduate school helping in a new Internet
Study Group at Macanany
College. After your first month in the office, you
are called into the director’s office and given the following abstract. She
tells you that the first study did not generate the results they desired but did
help them receive an NIH grant and therefore
there is now money for a new
and improved version. You are asked to turn some basic ideas into her
tomorrow morning including the three questions below.
Abstract:
Youth use the Internet and computers in unprecedented numbers. We have
yet to identify
interventions that can reach and retain large numbers of
diverse youth online and demonstrate HIV prevention efficacy. We tested a
single session condom promotion Internet intervention
5
–
minute video
for
18
–
24 year olds
using stock footage from a 1992 Fri
ends episode
in
two
recruitment arms: one
recruited online and one recruited in clinics.
The
study intervention was
carried out on the Internet.
W
e
compared the
proportion of sex acts
protected by condoms in the three months before
seeing the intervention
video to the proportion of sex acts protected by
condoms in the three months after seeing the intervention video. Both
measures were determined by self
–
report.
Among sexually active youth in
the Internet sample, persons exposed to the intervention had ve
ry slight
increases in condom norms, and this was the only factor impacting
outcome
. We saw no intervention effects in the clinic sample.
a.
Why might they have compared on
–
line enrollment vs. clinic enrollment in this
study and then given everyone the
same intervention
?
(1
–
2 sentences, 1 point)
b.
You are asked to improve upon this
study and test an internet prevention
intervention video to increase the use of condoms in youth aged 18
–
24 in inner
–
city Atlanta. You can re
–
use the previous study’s vid
eo or you have a budget to
create a new one. What
three changes
do you make to the previous study
design or the study intervention?
Please defend your answer. (2
–
3 sentences
each, 2 points each; 6 points total)
c.
Based on your new study’s design, to
whom do you feel your data are
generalizable? (1
–
2 sentences, 1 point)
2.
For the next questions, c
onsider these two abstracts published in the same year
with differing conclusions regarding the risk of cardiovascular events in patients
with HIV.
Abst
ract for D:A:D study
^
Background: It remains controversial whether exposure to combination antiretroviral
treatment increases the risk of myocardial infarction.
Methods
: In this prospective observational study, we enrolled 23,468 patients from 11
previously established cohorts from December 1999 to April 2001 and collected follow
–
up data until February 2002. Data were collected on infection with the human
immunodeficienc
y virus and on risk factors for and the incidence of myocardial
infarction. Relative rates were calculated with Poisson regression models. Combination
antiretroviral therapy was defined as any combination regimen of antiretroviral drugs
that included a pro
tease inhibitor or a nonnucleoside reverse transcriptase inhibitor.
Results
: Over a period of 36,199 person
–
years, 126 patients had a myocardial
infarction. The incidence of myocardial in
farction increased with longer exposure to
combination antiretroviral therapy (adjusted relative rate per year of exposure, 1.26 [95
percent confidence interval, 1.12 to 1.41]; P<0.001). Other factors significantly
associated with myocardial infarction wer
e older age, current or former smoking,
previous cardiovascular disease, and male sex, but not a family history of coronary
heart disease. A higher total serum cholesterol level, a higher triglyceride level, and the
presence of diabetes were also associate
d with an increased incidence of myocardial
infarction.
Conclusions:
Combination antiretroviral therapy was independently associated with a 26
percent relative increase in the rate of myocardial infarction per year of exposure during
the first four to six
years of use. However, the absolute risk of myocardial infarction was
low and must be balanced against the marked benefits from antiretroviral treatment.
N Engl J Med 2003;349: 1993
–
2003.
Abstract for Bozette study
^
Background
: Metabolic abnormalities associated with human immunodeficiency vi
rus
(HIV) infection, including dysglycemia and hyperlipidemia, are increasingly prevalent,
and there is concern about the possibility of an association with accelerated
cardiovascular and cerebrovascular disease.
Methods
: We conducted a retrospective study of the risk of card
iovascular and
cerebrovascular disease among the 36,766 patients who received care for HIV infection
at Veterans Affairs facilities between January 1993 and June 2001.
Results
: For antiretroviral therapy, 70.2 percent of the patients received nucleoside
analogues, 41.6 percen
t received protease inhibitors, and 25.6 percent received
nonnucleoside reverse
–
transcriptase inhibitors for a median of 17 months, 16 months,
and 9 months, respectively. Approximately 1000 patients received combination therapy
with a protease inhibitor fo
r at least 48 months, and approximately 1000 patients
received combination therapy with a nonnucleoside reverse
–
transcriptase inhibitor for at
least 24 months. Between 1995 and 2001, the rate of admissions for cardiovascular or
cerebrovascular disease decr
eased from 1.7 to 0.9 per 100 patient
–
years, and the rate
of death from any cause decreased from 21.3 to 5.0 deaths per 100 patient
–
years.
Patient
–
level regression analyses indicated that there was no relation between the use
of nucleoside analogues, prote
ase inhibitors, or nonnucleoside reverse
–
transcriptase
inhibitors and the hazard of cardiovascular or cerebrovascular events, but the use of
antiretroviral drugs was associated with a decreased hazard of death from any cause.
Conclusions:
Use of newer ther
apies for HIV was associated with a large benefit in
terms of mortality that was not diminished by any increase in the rate of cardiovascular
or cerebrovascular events or related mortality. Fear of accelerated vascular disease
need not compromise antiretro
viral therapy over the short term. However, prolonged
survival among HIVinfected patients means that longer
–
term observation and analysis
are required.
N Engl J Med 2003;348: 702
–
10.
a.
Using only the abstracts given, please list at least three
methodological
differences between these two studies. (1 sentence each; 3 points total)
b.
Compare the use of data from the National VA medical record system versus a
multi
–
site study. Do you feel these are essentially the same modes of collecting
patie
nt data or are they appreciably different? Please defend your answer. (
3
–
4
sentences;
6 points)
c.
Please compare the two analytic endpoints used. Which do you think has less
potential for bias? (
3
–
4 sentences;
6 points)
d.
Which conclusion do you feel
is the more credible conclusion? (Note: your
opinion is not being graded but your defense of this opinion is!
(2
–
3 sentences;
5
points)
3
.
A
n HIV
discordant couple (one partner is HIV
–
positive, one partner is HIV
–
negative)
intervention trial
is designed, and i
s
to be tested in a randomized
trial
design.
The intervention is a 6
–
session, small group discussion intervention; in
each 2
–
hour group session, participants learn condom use skills, and have
activities to increase self
–
efficacy in negoti
ating condom use with sex partners.
They practice putting condoms on a banana, and role play discussing condom
use with a sex partner. The outcome of the trial is number of times having
vaginal or anal sex without a condom per month, which is collected 3
and 6
months after completion of the intervention.
You are part of the research team designing this trial.
a.
Please describe the activities you would plan for trial participants randomized
to the control condition.
Explain why you chose those activitie
s. (2
–
3
sentences; 5 points)
b.
How would you propose to measure the outcome of sex without condoms?
What are the strengths and weaknesses of your proposed way of measuring
this outcome? (
2
–
3
sentences; 5 points)
c.
When the trial results come in, the te
am is excited to see that unprotected sex
in the intervention group decreased at both 3 and 6 months post
–
intervention,
compared to baseline. However, your boss points out that unprotected sex
also decreased in the control group.
Does the finding of decr
eased risk in the attention control group surprise you?
How do yo
u explain it to your boss? (2
–
3
sentences; 5 points)
4. The following data come from a survival analysis of persons with HIV infection.
This figure shows the Kaplan Meier plot for survival among persons with and without
Hepatitis Q Virus infection (HQV, an imaginary hepatitis virus.).
The authors als
o report the following data, based on analysis of data from the same
cohort, in the results: In a Cox proportional hazards regression model of time to
death, adjusting for history of AIDS diagnosis, CD4 count, prescription of antibiotics,
alcohol consumpt
ion, and marijuana consumption, the hazard ratio associated with
Hepatitis Q infection was 1.1 (95% confidence interval, 0.9
–
1.3).
a.
Are the data presented in the Kaplan
–
Meier plot univariate or multivariate? (1
sentence, 2 points)
b.
How would you inter
pret the differences between the results presented in the
Kaplan Meier plot and the results of the Cox proportional hazards regression? (
3
–
4 sentences;
6 points)
HQV
–
negative
HQV
–
positive
c.
What 3 aIDitional questions would you have about how the analyses were done,
based
on the figure and other data presented in the question? (
3 sentences, 2
point each;
6 points
total
)
5.
You are hired as a consultant to resolve a scientific difference of opinion between
two well
–
intentioned epidemiologists. The problem being aIDressed is
engagement
of black MSM in internet prevention trial. Banner advertisements are displayed on
gay
–
themed internet sex sites, and men who click through the advertisements are
enrolled in the trial. The two behavioral epidemiologists make their cases on ho
w to
aIDress the problem:
Epidemiologist A:
We know from our work to date that black MSM actually click
through banner advertisements to enroll in our studies at a 20% lower rate than to
white MSM. Therefore, we need to invest money in ways to increase
clickthrough
rates of black MSM, so that they are comparable to clickthrough rates in white MSM.
Even if we end up with a lower number of black MSM than white MSM in our study, I
will be comfortable if the rates of clickthrough are the same.
Epidemiologi
st B:
I have a more economical solution. Further research is
expensive, and I think we should just run the banner advertisements until we enroll a
certain number of white MSM, and then stop enrollment of white MSM and continue
to show those banner advert
isements to black MSM until we enroll the same number
of black MSM. This way, when we publish our study results, everyone will see that
we did a good job of enrolling black MSM because we’ll have equal numbers of
black and white MSM.
Which epidemiologist
to you agree with, and why? (200
–
350 word count, 10 points)