American Journal of Public Health | December 2007, Vol 97, No. 122238 | Research and Practice | Peer Reviewed | Kidder et al.
RESEARCH AND PRACTICE
Objectives. We sought to compare health status, health care use, HIV anti-
retroviral medication use, and HIV medication adherence among homeless and
housed people with HIV/AIDS.
Methods. Data were obtained from a cross-sectional, multisite behavioral sur-
vey of adults (N=7925) recently reported to be HIV positive.
Results. At the time interviews were conducted, 304 respondents (4%) were
homeless. Self-ratings of mental, physical, and overall health revealed that the
health status of homeless respondents was poorer than that of housed respon-
dents. Also, homeless respondents were more likely to be uninsured, to have
visited an emergency department, and to have been admitted to a hospital. Home-
less respondents had lower CD4 counts, were less likely to have taken HIV anti-
retroviral medications, and were less adherent to their medication regimen.
Homeless respondents needed more HIV social and medical services, but nearly
all respondents in both groups had received needed services. Housing status re-
mained a significant predictor of health and medication outcomes after we con-
trolled for potential confounding variables.
Conclusions. Homeless people with HIV/AIDS are at increased risk of negative
health outcomes, and housing is a potentially important mechanism for im-
proving the health of this vulnerable group. (Am J Public Health. 2007;97:
Health Status, Health Care Use, Medication Use,
and Medication Adherence Among Homeless and
Housed People Living With HIV/AIDS
| Daniel P . Kidder, PhD, Richard J. Wolitski, PhD, Michael L. Campsmith, DDS, MPH, and Glenn V. Nakamura, PhD
coverage—may be compounded among home-
less people living with the disease.30,34
People with HIV/AIDS also may have diffi-
culty adhering to prescribed HIV antiretrovi-
ral medication regimens.35,36These regimens
can be complex and often involve restrictions
on when and how the medications should be
taken and stored.31,34In addition, these med-
ications can have side effects, such as recur-
ring diarrhea, that are especially problematic
for homeless individuals. Medical providers
may believe that homeless individuals will not
be adherent, and thus they may be reluctant
to prescribe antiretroviral medications for
these individuals37given that inadequate ad-
herence can lead to drug resistance.34Despite
its importance, few studies have investigated
the issue of adherence to antiretroviral med-
ication regimens in this population.35,36,38–42
Overall, minimal research has been con-
ducted on the health of homeless people liv-
ing with HIV/AIDS.15,43,44We used data from
a large, multisite investigation to (1) assess
differences between homeless and housed
persons living with HIV/AIDS regarding
sociodemographic, health care, and medica-
tion adherence variables and (2) examine as-
sociations between housing status and health,
and medication adherence outcomes after
controlling for potential confounding factors.
We derived data for this study from the
Centers for Disease Control and Prevention’s
Supplement to HIV/AIDS Surveillance
(SHAS) project, a cross-sectional, multisite
study that collected behavioral surveillance
data from adults with HIV.45Recruitment
took place at 19 surveillance sites: the state
health departments of Arizona, Colorado,
Connecticut, Delaware, Florida, Georgia,
Illinois, Kansas, Maryland, Michigan, Min-
nesota, New Jersey, New Mexico, South Car-
olina, Texas, and Washington and the city
health departments of Houston, Tex; Los
Homeless people are at a disproportionate
risk for negative health consequences. For
instance, they typically have more chronic
diseases and more physical and mental
health problems than do the general popula-
tion, and they are at greater risk for infec-
tious diseases.1–13Homelessness is often co-
incident with poverty, mental illness, and
alcohol and drug use, compounding the
other health problems experienced by these
Homeless people are also more likely than
other groups to engage in behaviors that
place them at risk for HIV infection, includ-
ing risky sexual practices, injection drug use
and needle sharing, and performing sexual
acts in exchange for money, drugs, or a place
to stay.18–25Perhaps not surprisingly, previ-
ous research has shown that HIV is 3 to 9
times more prevalent among homeless indi-
viduals than among individuals in stable
It may be difficult for homeless people,
who are often faced with immediate subsis-
tence needs (e.g., finding adequate food and
shelter), to obtain medical care and adhere to
treatment regimens.30,31As a result, homeless
individuals are less likely than are the general
population to have stable sources of care, and
they often rely on emergency departments or
ambulatory care settings for their health care
needs.32,33Delayed medical care or lack of
care has negative effects such as delayed HIV
diagnoses and higher rates of serious oppor-
People who are living with HIV/AIDS and
are homeless face additional burdens not
faced by homeless people without HIV/AIDS.
For instance, individuals with HIV/AIDS
need greater access to comprehensive health
care, and barriers to care—including lack of
financial resources, lack of transportation, and
insufficient (or nonexistent) health insurance
December 2007, Vol 97, No. 12 | American Journal of Public Health Kidder et al. | Peer Reviewed | Research and Practice | 2239
RESEARCH AND PRACTICE
Angeles, Calif; and Philadelphia, Pa. Eligible
respondents were recruited from individuals
recently reported to have HIV or AIDS.
Recruitment was either facility based (i.e.,
from selected health care facilities; 9 sites) or
population based (i.e., from defined geograph-
ical areas; 10 sites). Potential participants had
been reported to local HIV/AIDS surveil-
lance systems within the previous 2 years. Re-
spondents were recruited using printed mate-
rials (e.g., flyers in clinics and HIV/AIDS
service organizations), provider referrals, and
reviews of clinic lists and HIV/AIDS surveil-
lance registries. Participant compensation was
determined individually at the different sites;
a majority offered participants up to $25.
Trained interviewers conducted individual
interviews with people 18 years or older. The
interviews were approximately 45 minutes in
duration and were conducted in either Eng-
lish or Spanish. Questionnaire modules in-
cluded demographic information and data on
alcohol and drug use, sexual behaviors, repro-
ductive history (among women), HIV testing
and medical therapy, and use of health care
and social services.
The data used in our study were collected
between May 2000 and December 2003.
During this period, 20944 men and women
were eligible for interviews; 8732 of these in-
dividuals could not be located or were de-
ceased (or there was no attempt made to
contact them). Among the 12212 eligible indi-
viduals offered enrollment, 8129 (67%) com-
pleted an interview and 4083 (33%) refused.
A total of 7925 individuals had valid data on
housing status (41% of those at population-
based sites and 59% of those at facility-based
sites) and were included in the study.
In terms of housing status, respondents
were categorized as homeless if they reported
living in a shelter or on the streets at the time
they were interviewed. Respondents were
considered housed if, at the time of their in-
terview, they were living in (1) a house or
apartment, either alone or with a spouse,
partner, friends, or family; (2) a medical care
facility; or (3) a correctional institution.
Data were gathered on gender, race/ethnicity,
age, highest level of education completed,
marital status, employment status, annual
household income, and primary source of in-
come. Primary source of income was catego-
rized as public (including public assistance
and Social Security supplemental income or
disability income) or private (including sala-
ries, savings, pensions, retirement funds, and
assistance from spouse, partner, family, or
friends). Risk categories associated with HIV
transmission or acquisition were based on life-
time behaviors (i.e., behaviors that could have
occurred either before or after HIV diagnosis).
An item focusing on drug use assessed illicit
injection or noninjection drug use (e.g., mari-
juana, heroin, cocaine, methamphetamine) in
the past 12 months. Potential problem drink-
ing was assessed with the 4-item CAGE scale
(Cut-Annoyed-Guilt-Eye Opener; score range:
0–4)46,47; scores of 2 to 4 indicated a poten-
tial lifetime drinking disorder.
Items focusing on health insurance coverage
included type of coverage and primary method
of payment for HIV-related prescription med-
ications. Respondents were asked whether
they had ever received medical care for HIV.
Those who answered “yes” were asked about
the most frequent source of that care, fre-
quency of care, number of emergency depart-
ment visits for HIV/AIDS care, and number of
hospital admissions (all in the past 12 months).
Numbers of emergency department visits and
hospital admissions were recategorized into di-
chotomous variables (0 or ≥1).
Self-rated health items were derived from
the Centers for Disease Control and Preven-
tion’s health-related quality of life measures.48
Overall health status was assessed on a 5-
point scale (possible responses ranged from
poor to excellent). Respondents also reported
the number of days in the past 30 days dur-
ing which (1) they felt very healthy and full of
energy; (2) they were not able to engage in
their usual activities as a result of poor physi-
cal or mental health; (3) their physical health
(physical illness and injury) was not good;
(4) pain made it difficult for them to engage
in their usual activities; (5) they did not rest or
sleep enough; (6) their mental health (stress,
depression, and problems with emotions) was
not good; (7) they felt sad, blue, or depressed;
and (8) they felt worried, tense, or anxious.
Items focusing on health conditions and
HIV progression asked whether participants
had ever had hepatitis, Pneumocystis carinii
pneumonia, a positive tuberculosis skin test, a
CD4 count test, and a viral load test. If they had
undergone CD4 testing, viral load testing, or
both, they were asked their most recent results.
Respondents were shown a list of 26 HIV
antiretroviral medications and asked which,
if any, they had ever taken and were cur-
rently taking. Those who had never taken
antiretrovirals were asked why they had not.
Responses to this open-ended question were
coded into predetermined categories.
Medication adherence questions were
asked only of respondents who reported cur-
rently taking HIV antiretroviral medications.
The primary HIV medication adherence item
focused on whether respondents had missed
or skipped any doses of any HIV antiretro-
viral medication in the previous 48 hours.
They were also asked how often they took
HIV medications exactly as their doctor had
told them. Respondents who did not answer
“always” were asked the main reason why
they did not do so. In addition, respondents
were asked whether they had stopped taking
any HIV medicines prescribed by a doctor
and, if so, to choose from a list of reasons
why they had done so.
Respondents were asked whether they had
needed each of 13 types of social and med-
ical services related to HIV in the past 12
months. In the case of each service needed,
respondents were asked whether they had
been able to obtain the service in the preced-
ing 12 months.
We used the χ2test (for categorical vari-
ables) and t test (for continuous variables) to
conduct bivariate analyses comparing home-
less and housed respondents. We used the
nonparametric Wilcoxon test for ordinal cate-
gorical variables. The α level was set at <.05.
To investigate the independent effects of
homelessness, we conducted separate logistic
regression analyses involving 6 self-reported
dependent variables: (1) health status (poor/
fair vs good/very good/excellent), (2) most
recent CD4 count (<200/µL vs ≥200/µL),
(3) most recent viral load (<500 viral
copies per milliliter [undetectable] vs ≥500
viral copies per milliliter), (4) emergency de-
partment use in past 12 months, (5) current
use of HIV antiretroviral medications, and
American Journal of Public Health | December 2007, Vol 97, No. 12 2240 | Research and Practice | Peer Reviewed | Kidder et al.
RESEARCH AND PRACTICE
TABLE 1—Sociodemographic Characteristics, Drug Use, and Alcohol Use Among Homeless
and Housed Respondents: Supplement to HIV/AIDS Surveillance Project, 2000–2003
Less than high school
More than high school
Annual household income,$
Source of income
No identified source
HIV risk category
Both MSM and IDU
Illicit drug usein past 12 months
Potential alcohol abuse
2770 (36.4) 85.4***
Note.MSM=men who have sex with men; IDU=injection drug user.
aThis is the percentage meeting CAGE (Cut-Annoyed-Guilt-Eye Opener) criteria for potential lifetime drinking disorder.
*P<.05; **P<.01; ***P<.001.
(6) complete adherence to HIV antiretroviral
medication regimen in preceding 48 hours.
The analyses of most recent CD4 count and
viral load count included only respondents
who had ever undergone a CD4 test and a
viral load test, respectively. The medication
adherence analysis included only respon-
dents currently taking HIV antiretroviral
medications. The remaining analyses in-
cluded all respondents.
Several separate logistic regression analyses
were conducted for each dependent variable.
In the first analysis, housing status was the
single predictor. We selected the predictor
variables used in the second set of analyses,
including housing status, sociodemographic
characteristics, drug use, and alcohol use, on
the basis of previous research suggesting asso-
ciations with the dependent variables.
In a third set of logistic regression analyses,
we repeated the previous analyses focusing
on health care use with the addition of self-
reported CD4 count and self-rated health as
markers of health status. The purpose of
these analyses was to examine the effects of
homelessness on health care use after we
controlled for individual differences in need
for health care. These analyses were con-
ducted for 3 of the dependent variables
(emergency department use, current use of
HIV antiretroviral medications, and adher-
ence to HIV antiretroviral medication regi-
men in preceding 48 hours) but were not
conducted for the other health status mea-
sures, because we hypothesized that they
would be highly related to self-reported
health and CD4 count.
Sociodemographic Characteristics, Drug
Use, and Alcohol Use
Of the 7925 respondents included in the
analyses, 304 (3.8%) were homeless. More
than 70% were male (homeless respondents
were significantly more likely to be male), and
more than half were Black (Table 1). Respon-
dents’ mean age was 40 years (SD=9.3),
and nearly two thirds had never been mar-
ried. Homeless respondents had less educa-
tion than housed respondents and were less
likely to be employed. In addition, they had
lower incomes and were more likely to re-
ceive public assistance or to have no identi-
fied source of income.
In terms of risk categories associated with
HIV transmission or acquisition, more men in
the housed group than in the homeless group
had had sexual intercourse with other men.
By contrast, the percentage of respondents
who reported injection drug use and the per-
centage of men who reported engaging in
both sexual intercourse with other men and
injection drug use were higher in the home-
December 2007, Vol 97, No. 12 | American Journal of Public HealthKidder et al. | Peer Reviewed | Research and Practice | 2241
RESEARCH AND PRACTICE
TABLE 2—Health Care Coverage and Use, Physical and Mental Health, and HIV Medication
Use and Adherence Among Homeless and Housed Respondents: Supplement to HIV/AIDS
Surveillance Project, 2000–2003
Health care coverage,no.(%)
Main method of medication payments,no.(%)
AIDS Drug Assistance Program
Personal health care coverage
None (not taking any medications)
Ever received medical care for HIV,no.(%)
Frequency of visiting facility for HIV care,bno.(%)
No receipt of medical care
Every 3 months or less
Every other month
More than monthly
Most frequent source of medical care for HIV,bno.(%)
No care received
Use of emergency room for HIV/AIDS care at least once,b
Admission to hospital for HIV/AIDS care at least once (not
including emergency department),bno.(%)
Ever had hepatitis,no.(%)
Ever had Pneumocystis carinii pneumonia,no.(%)
Ever had a positive tuberculosis skin test,no.(%)
Ever had a CD4 count test,no.(%)
Most recent CD4 count,no.(%)
Ever had a viral load test,no.(%)
110 (37.2)1597 (21.3) 42.3***
In comparison with housed respondents,
homeless respondents were more likely to
have used illicit drugs in the past 12 months
and to meet CAGE criteria for lifetime alco-
hol problems (Table 1).
Health Care, Physical and Mental Health,
and HIV Medication Use and Adherence
Homeless respondents were more likely to
be uninsured and less likely to have private
health insurance coverage (Table 2). For
about one third of the members of each
group, their main method of paying for med-
ications was the AIDS Drug Assistance Pro-
gram. A greater percentage of housed respon-
dents had medications paid for through their
own health care coverage.
Nearly all respondents (99%) had received
HIV medical care, with no difference be-
tween the homeless and housed groups
(Table 2). Frequency of HIV care differed
between the 2 groups; a higher percentage
of homeless respondents had received care at
Community clinics were the most frequent
source of HIV care among both homeless
and housed respondents. A larger percentage
of respondents in the housed group than in
the homeless group received care from pri-
vate physicians. Homeless respondents were
more likely to have used an emergency de-
partment for HIV care and to have been ad-
mitted to the hospital because of HIV in the
past 12 months.
Self-ratings of mental, physical, and overall
health status revealed that the health of home-
less respondents was poorer than that of
housed respondents (Table 2). Also, homeless
respondents reported significantly more days in
the past 30 days during which (1) physical or
mental health problems prevented their usual
activities, (2) they were in poor physical health,
(3) pain limited their activities, (4) they did not
rest or sleep enough, (5) they were in poor
mental health, (6) they were depressed, and
(7) they were worried. Finally, they reported
fewer days on which they were very healthy.
Homeless respondents were more likely than
housed respondents to have had hepatitis,
Pneumocystis carinii pneumonia, and a posi-
tive tuberculosis skin test (Table 2). More than
90% of both groups had undergone CD4
or viral load tests, with no between-group
American Journal of Public Health | December 2007, Vol 97, No. 12 2242 | Research and Practice | Peer Reviewed | Kidder et al.
RESEARCH AND PRACTICE
Most recent viral load,no.(%)
Ever used HIV antiretroviral medications,no.(%)
Was currently using HIV antiretroviral medication,no.(%)
Had skipped any pills in past 48 hours,no.(%)
Frequency of taking HIV/AIDS medication exactly as prescribed
in past 30 days,no.(%)
Rarely or never
Had stopped taking any prescription HIV/AIDS medications,
Primary reason for stopping medications,dno.(%)
Instructions from doctor
Doctor switched drugs
Went back on the street
Drugs quit working
Partner suggested stopping
No.of days physical or mental health problems prevented
usual activities,emean (SD)
No.of days physical health was not good,emean (SD)
No.of days pain made it difficult to do usual activities,e
No.of days without enough sleep or rest,emean (SD)
No.of days mental health not good,emean (SD)
No.of days sad,blue,or depressed,emean (SD)
No.of days worried,emean (SD)
No.of days very healthy,emean (SD)
5.9 ( 9.6)
Note.As a result of rounding,percentages may not sum to 100%.
aχ2values for percentages; t values for means.
bIn past 12 months.
cWilcoxon test for trend,z approximation.
dParticipants could choose multiple responses.
eIn past 30 days.
*P<.05; **P<.01; ***P<.001.
differences. However, there were differences
in self-reported test results. Homeless respon-
dents were more likely to report that their
most recent CD4 count was less than 200
and less likely to report an undetectable
A lower percentage of homeless respon-
dents reported having ever taken HIV anti-
retroviral medications, and they were also less
likely to currently be taking these medications
(Table 2). Among those taking antiretrovirals,
there was no difference between groups
(housed: 83%; homeless: 85%; χ2=0.29,
P=.59) in the percentage taking 3 or more,
which is sometimes used as a basic measure
of a sufficient treatment regimen.
In addition, among the respondents taking
HIV antiretroviral medications, self-reported
adherence was significantly lower in the
homeless group. More than 30% of homeless
respondents had missed doses in the past 48
hours, as compared with less than 20% of the
housed respondents. Homeless respondents
were also less likely to report that they had
always taken their HIV medications exactly
as prescribed in the past 30 days.
A higher percentage of homeless respon-
dents than housed respondents reported they
had stopped taking prescribed HIV/AIDS
medications. Among the respondents who
had stopped taking their medications, the
primary reason was that a doctor had told
them to do so, and a higher percentage of
housed respondents reported this reason.
However, the largest difference between
groups involved reports of going “back on
the street,” with more of the homeless respon-
dents mentioning this reason.
HIV-Related Services Needed and
Homeless respondents were more likely
than housed respondents to have needed 11
of the 13 HIV-related services assessed
(Table 3). Despite differences between groups
in need for services, there were no differ-
ences regarding service receipt; nearly all re-
spondents (more than 90%) in both groups
received services in the category in which
they reported having a need.
Logistic Regression Analyses
Housing status, as a single predictor, was
significant for all 6 dependent variables in
the bivariate logistic regression analyses
(Table 4). Homeless respondents were less
likely than housed respondents to report good
or excellent health, less likely to have a self-
reported CD4 count of 200 or above, less
likely to have a self-reported undetectable
viral load, more likely to have visited an
emergency department in the past year, less
likely to be taking HIV medications, and
less likely to have been treatment adherent
in the past 48 hours.
Multivariate analyses controlled for socio-
demographic, drug use, and alcohol use vari-
ables. In these analyses, housing status re-
mained a significant predictor of health
status, most recent viral load, emergency de-
partment use, current use of HIV medica-
tions, and HIV medication adherence. Hous-
ing status did not remain a significant
predictor for most recent CD4 count. When
most recent CD4 count and self-rated health
December 2007, Vol 97, No. 12 | American Journal of Public HealthKidder et al. | Peer Reviewed | Research and Practice | 2243
RESEARCH AND PRACTICE
TABLE 3—HIV-Related Services Needed and Received in Past 12 Months Among Homeless
and Housed Respondents: Supplement to HIV/AIDS Surveillance Project, 2000–2003
HIV case management services
Mental health counseling
Finding a doctor
Finding dental services
Finding shelter or housing
Finding meals or food
Local volunteer support services
Education or information on HIV
Home health services
TABLE 4—Housing Status as a Predictor in Logistic Regression Analyses of Health Status, Health Care
Use, and HIV Medication Use and Adherence: Supplement to HIV/AIDS Surveillance Project, 2000–2003
Used Was Adherent
Bivariate,OR (95% CI)
Multivariate,aAOR (95% CI)
Multivariate including CD4
and self-rated health
status,aAOR (95% CI)
Note.OR=odds ratio; CI=confidence interval; AOR=adjusted odds ratio.
aIn the multivariate analyses,we controlled for HIV risk group,age,gender,race/ethnicity,marital status,education,annual household income,employment status,use of illicit drugs in past 12
months,and lifetime alcohol abuse.
were included in the multivariate analyses as
additional predictors, housing status re-
mained a significant predictor of emergency
department use, current use of HIV medica-
tions, and HIV medication adherence.
Some of the respondents who were hospi-
talized (n=152) or in a correctional facility
(n=15) at the time of interview may have
been homeless before entering these facilities.
Thus, all multivariate analyses were repeated
without data from these respondents. The
findings were the same as those for the entire
Our results showed that homeless respon-
dents were more likely than those in
stable housing situations to report a wide
range of negative health outcomes. For exam-
ple, homeless respondents had more medical
and social service needs, suggesting that they
may have been sicker than housed respon-
dents. Although nearly all of the respondents
in both groups reported receiving the services
they needed, homeless respondents reported
poorer health on a variety of mental and
physical health measures, including markers
of HIV disease progression. In addition, more
homeless respondents received care from
emergency departments, they were hospital-
ized more often, and they had more frequent
visits for medical care relative to housed
Although HIV treatment appeared to be
indicated for a larger percentage of homeless
respondents, they were less likely to have
ever taken or to currently be taking HIV an-
tiretroviral medications and more likely to
American Journal of Public Health | December 2007, Vol 97, No. 122244 | Research and Practice | Peer Reviewed | Kidder et al.
RESEARCH AND PRACTICE
have stopped taking these medications. In the
case of respondents who were taking anti-
retrovirals, rates of self-reported adherence
were lower among those who were homeless.
This finding indicates a clear need to im-
prove access to antiretroviral medications
among homeless people living with HIV/
AIDS and to provide further assistance with
medication adherence.40,49This is not an
easy task, but it may be cost effective in the
long term because it will decrease the bur-
den on hospitals and emergency departments
and improve the health of homeless people
These results suggest that many homeless
people living with HIV/AIDS do not receive
the quality of care that is optimal for manag-
ing HIV. The extent to which this situation is
because of limitations in access to care, the
medication prescribing practices of physicians,
and problems in regard to medication adher-
ence is an important issue that should be ad-
dressed in future research.
Our multivariate analyses indicated that
housing status was a significant predictor of
health status, health care and emergency de-
partment use, use of HIV medications, and
HIV medication adherence. Housing status
remained significant even after we con-
trolled for demographic, drug use, and alco-
hol use variables. This is an important find-
ing because it indicates that stable housing
may improve the health of people living
with HIV/AIDS. This intuitive result is not a
new concept in the literature, given that
having a place to stay can improve people’s
mental and physical health by decreasing
their stress and fatigue and allowing them to
focus on other health care needs, such as
medical appointments and adherence to
In addition, the connection between physi-
cal health and mental health is well estab-
lished,50suggesting that providing housing
to homeless individuals might improve both
their mental and physical health. Yet, most
previous studies have examined the relation-
ship between health status and inadequate
housing rather than lack of housing. Ours is
one of the first large-scale studies to demon-
strate the association between lack of housing
and the health status of people living with
There were some limitations of this study.
For example, it was cross sectional, and thus
cause-and-effect conclusions regarding home-
lessness and health could not be drawn. It is
possible that those who reported being home-
less were in poorer health when they became
homeless or that factors associated with
poorer health (e.g., substance use, mental
illness) contributed to their homelessness.
However, other research has shown that im-
provements in housing status result in im-
provements in health status and decreased
HIV risk behaviors.2
Furthermore, although SHAS was a large,
multisite project, its participants may not be
representative of all people in the United
States with HIV/AIDS, and we do not know
how the individuals who refused to take part
may have affected the results. However, to
our knowledge, this is the largest data set that
has been used to examine housing status and
health among HIV-positive individuals. Be-
cause this study was not specifically designed
to investigate homelessness issues, certain
concepts related to homelessness, such as
mental illness, were not assessed. In addition,
data were self-reported and were thus subject
to socially desirable responding and recall bi-
ases. Also, the time periods used in assessing
drug and alcohol use behaviors (i.e., “ever”
and “in the past 12 months”) may not have
accurately reflected current behaviors.
The numbers of people living with HIV/
AIDS and experiencing homelessness were
probably underestimated in this study. SHAS
recruitment methods did not target homeless
individuals, who are typically more difficult to
locate. Thus, our study may have included a
lower percentage of homeless people with
HIV/AIDS than in the general population.
However, it is difficult to determine
whether percentages of homeless people were
actually lower in this study, given that deter-
mining the number of homeless individuals in
the general population is difficult as a result of
the episodic nature and typically short dura-
tion of homelessness.13Using a wider time-
frame than current housing status to catego-
rize homeless individuals might result in a
larger, more accurate estimate of HIV-positive
people who have experienced homelessness.
Even taking what is likely to be a conservative
estimate from this study (4%), tens of thou-
sands of the estimated 1 million people in the
United States who are HIV positive51may be
homeless at any given time.
Homeless people living with HIV/AIDS
are members of 2 marginalized groups, yet
little research has been conducted to under-
stand the health status and health behaviors
of this vulnerable population. Even less re-
search has been conducted on the effects of
interventions designed to improve the health
of homeless people with HIV/AIDS.
One intervention concept that has been
gaining empirical support is provision of hous-
ing as a structural intervention to improve
health and prevent the spread of HIV.52Pro-
viding housing to homeless people living with
HIV/AIDS not only may allow them to ad-
dress their health care needs but could also
result in improved mental and physical health
outcomes (e.g., decreased viral load) and re-
duced risk behaviors, which could reduce
their likelihood of transmitting HIV.52,53The
health of homeless people with HIV/AIDS is
an important public health and social justice
issue that requires increased attention at both
the local and national levels given the accu-
mulating evidence that these individuals are
at greater risk of numerous health problems
while they remain unhoused.
About the Authors
The authors are with the Division of HIV/AIDS Preven-
tion, National Center for HIV/AIDS, Viral Hepatitis,
STD, and TB Prevention, Centers for Disease Control and
Prevention, Atlanta, Ga.
Requests for reprints should be sent to Daniel P. Kidder,
PhD, Centers for Disease Control and Prevention, 1600
Clifton Rd NE, Mailstop E-04, Atlanta, GA 30333
This article was accepted December 13, 2006.
Note. The findings and conclusions are those of the
authors and do not necessarily represent the views of the
Centers for Disease Control and Prevention.
D.P. Kidder and R.J. Wolitski originated the analysis
idea. D.P. Kidder, R.J. Wolitski, and M.L. Campsmith
refined the article idea and interpreted findings. D.P.
Kidder led the writing. G.V. Nakamura analyzed the
Human Participant Protection
The Supplement to HIV/AIDS Surveillance project
was approved by Centers for Disease Control and
December 2007, Vol 97, No. 12 | American Journal of Public HealthKidder et al. | Peer Reviewed | Research and Practice | 2245
RESEARCH AND PRACTICE
Prevention and local institutional review boards. Partici-
pants provided informed consent.
1. Thomson H, Petticrew M, Morrison D. Health ef-
fects of housing improvement: systematic review of in-
tervention studies. BMJ. 2001;323:187–190.
2.Aidala A, Cross JE, Stall R, Harre D, Sumartojo E.
Housing status and HIV risk behaviors: implications for
prevention and policy. AIDS Behav. 2005;9:251–265.
3. Evans GW, Wells NM, Chan H-YE, Saltzman H.
Housing quality and mental health. J Community
4.Dunn JR. Housing and health inequalities: review
and prospects for research. Housing Stud. 2000;15:
5. White MC, Tulsky JP, Dawson C, Zolopa AR,
Moss AR. Association between time homeless and
perceived health status among the homeless in San
Francisco. J Community Health. 1997;22:271–282.
6.Giannetti V, Freyder P. HIV risk among the
homeless. J Health Soc Policy. 1998;10:27–38.
7. Cheung RC, Hanson AK, Maganti K, Keeffe EB,
Matsui SM. Viral hepatitis and other infectious diseases
in a homeless population. J Clin Gastroenterol. 2002;
8.Gelberg L, Linn LS. Demographic differences in
health status of homeless adults. J Gen Intern Med.
9. Martens WH. A review of physical and mental
health in homeless persons. Public Health Rev. 2001;
10. Gelberg L, Linn LS. Assessing the physical health
of homeless adults. JAMA. 1989;262:1973–1979.
11. Krieger J, Higgins DL. Housing and health: time
again for public health action. Am J Public Health.
12. Riley ED, Moss AR, Clark RA, Monk SL, Bangs-
berg DR. Cash benefits are associated with lower risk
behavior among the homeless and marginally housed
in San Francisco. J Urban Health. 2005;82:142–150.
13. Burt M, Aron LY, Lee E. Helping America’s Home-
less: Emergency Shelter or Affordable Housing? Washing-
ton, DC: Urban Institute Press; 2001.
14. Folsom DP, Hawthorne W, Lindamer L, et al.
Prevalence and risk factors for homelessness and uti-
lization of mental health services among 10,340 pa-
tients with serious mental illness in a large public men-
tal health system. Am J Psychiatry. 2005;162:370–376.
15. Riley ED, Wu AW, Perry S, et al. Depression and
drug use impact health status among marginally
housed HIV-infected individuals. AIDS Patient Care
16. Burt MR, Aron LY, Douglas T, Valente J, Lee E,
Iwen B. Homelessness: Programs and the People They
Serve. Washington, DC: Urban Institute Press; 1999.
17. Galea S, Vlahov D. Social determinants and the
health of drug users: socioeconomic status, homeless-
ness, and incarceration. Public Health Rep. 2002;
18. Culhane DP, Gollub E, Kuhn R, Shpaner M. The
co-occurrence of AIDS and homelessness: results from
the integration of administrative databases for AIDS
surveillance and public shelter utilisation in Philadel-
phia. J Epidemiol Community Health. 2001;55:515–520.
19. Wolitski RJ, Kidder DP, Fenton K. HIV,
homelessness, and public health: critical issues and a call
for increased action. AIDS Behav. 2007;11:5167–5171.
20. Fournier AM, Tyler R, Iwasko N, LaLota M, Shultz J,
Greer PJ. Human immunodeficiency virus among the
homeless in Miami: a new direction for the HIV epi-
demic. Am J Med. 1996;100:582–584.
21. Allen DM, Lehman S, Green TA, et al. HIV infec-
tion among homeless adults and runaway youth, United
States, 1989–1992. AIDS. 1994;8:1593–1598.
22. Wechsberg WM, Lam WKK, Zule W, Hall G,
Middlesteadt R, Edwards J. Violence, homelessness,
and HIV risk among crack-using African-American
women. Subst Use Misuse. 2003;38:669–700.
23. Metraux S, Metzger DS, Culhane DP. Homeless-
ness and HIV risk behaviors among injection drug
users. J Urban Health. 2004;81:618–629.
24. Rahav M, Nuttbrock L, Rivera JJ, Link BG. HIV
infection risks among homeless, mentally ill, chemical
misusing men. Subst Use Misuse. 1998;33:1407–1426.
25. Kidder DP, Wolitski RJ, Royal S, et al. Access to
housing as a structural intervention for homeless and
unstably housed people living with HIV: rationale,
methods, and implementation of the Housing and
Health Study. AIDS Behav. 2007;11:5149–5161.
26. Paris NM, East RT, Toomey KE. HIV seropreva-
lence among Atlanta’s homeless. J Health Care Poor
27. Torres RA, Mani S, Altholtz J, Brickner PW.
Human immunodeficiency virus infection among
homeless men in a New York City shelter: association
with Mycobacterium tuberculosis infection. Arch Intern
28. Zolopa AR, Hahn JA, Gorter R, Miranda J,
Wlodarczyk D. HIV and tuberculosis infection in San
Francisco’s homeless adults. JAMA. 1994;272:455–461.
29. Estebanez PE, Russell NK, Aguilar MD, Beland F,
Zunzunegui MV. Women, drugs and HIV/AIDS: results
of a multicentre European study. Int J Epidemiol. 2000;
30. Gelberg L, Gallagher TC, Andersen RM, Koegel P.
Competing priorities as barriers to medical care in
homeless adults in Los Angeles. Am J Public Health.
31. Conanan B, London K, Martinez L, et al. Adapting
Your Practice: Treatment and Recommendations for Home-
less Patients With HIV/AIDS. Nashville, Tenn: Health
Care for the Homeless Clinician’s Network; 2003.
32. Kushel MB, Perry S, Bangsberg D, Clark R, Moss AR.
Emergency department use among the homeless and
marginally housed: results from a community-based
study. Am J Public Health. 2002;92:778–784.
33. O’Toole TP, Gibbon JL, Hanusa BH, Fine MJ.
Utilization of health care services among subgroups of
urban homeless and housed poor. J Health Polit Policy
34. Song J. HIV/AIDS & Homelessness: Recommenda-
tions for Clinical Practice and Public Policy. Nashville,
Tenn: National Health Care for the Homeless Council
35. Moss AR, Hahn JA, Perry S, et al. Adherence to
highly active antiretroviral therapy in the homeless
population in San Francisco: a prospective study. Clin
Infect Dis. 2004;39:1190–1198.
36. Berg KM, Demas PA, Howard AA, Schoenbaum EE,
Gourevitch MN, Arnsten JH. Gender differences in fac-
tors associated with adherence to antiretroviral therapy.
J Gen Intern Med. 2004;19:1111–1117.
37. Wong MD, Cunningham WE, Shapiro MF, et al.
Disparities in HIV treatment and physician attitudes
about delaying protease inhibitors for nonadherent
patients. J Gen Intern Med. 2004;19:366–374.
38. Johnson MO, Catz SL, Remien RH, et al. Theory-
guided, empirically supported avenues for intervention
on HIV medication nonadherence: findings from the
Health Living Project. AIDS Patient Care STDs. 2003;
39. Wagner JH, Justice AC, Chesney M, Sinclair G,
Weissman S, Rodriguez-Barradas M. Patient- and
provider-reported adherence: toward a clinically useful
approach to measuring antiretroviral adherence. J Clin
40. Holodniy M, Charlebois ED, Bangsberg DR,
Zolopa AR, Schulte M, Moss AR. Prevalence of anti-
retroviral drug resistance in the HIV-1-infected urban
indigent population in San Francisco: a representative
study. Int J STD AIDS. 2004;15:543–551.
41. Bangsberg DR, Hecht FM, Charlebois ED, et al.
Adherence to protease inhibitors, HIV-1 viral load, and
development of drug resistance in an indigent popula-
tion. AIDS. 2000;14:357–366.
42. Waldrop-Valverde D, Valverde E. Homelessness
and psychological distress as contributors to antiretro-
viral nonadherence in HIV-positive injecting drug
users. AIDS Patient Care STDs. 2005;19:326–334.
43. Hall CS, Charlebois ED, Hahn JA, Moss AR,
Bangsberg DR. Hepatitis C virus infection in San
Francisco’s HIV-infected urban poor. J Gen Intern Med.
44. Smith MY, Rapkin BD, Winkel G, Springer C,
Chhabra R, Feldman IS. Housing status and health care
service utilization among low-income persons with
HIV/AIDS. J Gen Intern Med. 2000;15:731–738.
45. Buehler JW, Diaz T, Hersh BS, et al. The Supple-
ment to HIV-AIDS Surveillance project: an approach
for monitoring HIV risk behaviors. Public Health Rep.
46. Buchsbaum DG, Buchanan RG, Centor RM, et al.
Screening for alcohol abuse using CAGE scores and
likelihood ratios. Ann Intern Med. 1991;115:774–777.
47. Ewing JA. Detecting alcoholism: the CAGE ques-
tionnaire. JAMA. 1984;252:1905–1907.
48. Moriarty D, Zack M, Kobau R. The Centers for
Disease Control and Prevention’s Healthy Days Mea-
sures—population tracking of perceived physical and
mental health over time. Health Qual Life Outcomes.
49. Bamberger JD, Unick J, Klein P, Fraser M,
Chesney M, Katz MH. Helping the urban poor stay
with antiretroviral HIV drug therapy. Am J Public
50. Jones D, Beach S, Forehand R. Disease status in Af-
rican American single mothers with HIV: the role of de-
pressive symptoms. Health Psychol. 2001;20:417–423.
51. Glynn M, Rhodes P. Estimated HIV prevalence in
the United States at the end of 2003. Paper presented
at: National HIV Prevention Conference, June 2005,
52. Sumartojo E. Structural factors in HIV prevention:
concepts, examples, and implications for research.
AIDS. 2000;14(suppl 1):S3–S10.
53. Quinn T, Wawer M, Sewankambo N, et al. Viral
load and heterosexual transmission of human immunode-
ficiency virus type 1. N Engl J Med. 2000;342:921–929.