AIDS Education and Prevention, 26(2), 122–133, 2014
© 2014 The Guilford Press
Kimberly M. Nelson, MS, MPH, is with the Department of Psychology, University of Washington, Seattle.
Hanne Thiede, DVM, MPH, is with Public Health – Seattle & King County, Seattle, Washington. Richard
A. Jenkins, PhD, is with the National Institute on Drug Abuse, Bethesda, Maryland. James W. Carey,
PhD, MPH, is with the Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention,
Atlanta, Georgia, Matthew R. Golden, MD, MPH, is with the Public Health – Seattle & King County,
Seattle, Washington, the School of Public Health and Community Medicine and the School of Medicine,
University of Washington, Seattle.
We would like to thank our participants for their help with this project. This study was funded by CDC
cooperative agreement number U64/CCU019523. Kimberly Nelson is supported by NIMH of the Na-
tional Institutes of Health under award number F31MH088851. The findings and conclusions in this re-
port are those of the authors and do not necessarily represent the views of the Centers for Disease Control
and Prevention or the National Institutes of Health.
Address correspondence to Kimberly M. Nelson, MS, MPH, University of Washington, Department of
Psychology, Box 351525, Seattle WA 98195-1525. E-mail: firstname.lastname@example.org
FACTORS RELATED TO DELAYED HIV DIAGNOSIS
NELSON ET AL.
PERSONAL AND CONTEXTUAL FACTORS
RELATED TO DELAYED HIV DIAGNOSIS
AMONG MEN WHO HAVE SEX WITH MEN
Kimberly M. Nelson, Hanne Thiede, Richard A. Jenkins,
James W. Carey, Rebecca Hutcheson, and Matthew R. Golden
Delayed HIV diagnosis among men who have sex with men (MSM) in the
United States continues to be a significant personal and public health issue.
Using qualitative and quantitative data from 75 recently tested, HIV-sero-
positive MSM (38 delayed and 37 nondelayed testers), the authors sought
to further elucidate potential personal and contextual factors that may
contribute to delayed HIV diagnosis among MSM. Findings indicate that
MSM who experience multiple life stressors, whether personal or contex-
tual, have an increased likelihood of delaying HIV diagnosis. Furthermore,
MSM who experience multiple life stressors without the scaffolding of
social support, stable mental health, and self-efficacy to engage in protective
health behaviors may be particularly vulnerable to delaying diagnosis. Inter-
ventions targeting these factors as well as structural interventions targeting
physiological and safety concerns are needed to help MSM handle their life
stressors more effectively and seek HIV testing in a timelier manner.
Men who have sex with men (MSM) remain the group most affected by HIV in the
United States, accounting for approximately 63% of all new HIV infections and
52% of all persons living with HIV (Centers for Disease Control and Prevention
[CDC], 2012a, 2012b). Among MSM living with HIV, it has been estimated that,
depending on their race/ethnicity, 19%–26% are unaware of their HIV infection
(Chen et al., 2012). The CDC has routinely suggested at least annual HIV screening
for sexually active MSM (Branson et al., 2006); however, results from the National
HIV Behavioral Surveillance System in 21 cities indicated that 53% of the 680 MSM
FACTORS RELATED TO DELAYED HIV DIAGNOSIS 123
who tested HIV-seropositive as a part of the study had previously tested more than
12 months prior to diagnosis or had never tested before (CDC, 2011).
In the past few years, there has been a growing interest in the “treatment cas-
cade” and implications for HIV prevention related to attrition at each step (Mugave-
ro, Amico, Horn, & Thompson, 2013). The treatment cascade starts with estimates
of prevalent HIV cases, with successive bars illustrating the subsequent steps of HIV
diagnosis, linkages to care, retention in care, antiretroviral therapy receipt, and plas-
ma viral suppression. People who are HIV-seropositive but remain undiagnosed are
unable to receive the care and treatment they need to slow disease progression and
prevent transmission. Early diagnosis of HIV has significant personal and public
health benefits. Diagnosis early in disease progression increases survival, improves
health outcomes, and decreases health care costs (Chadborn, Delpech, Sabin, Sinka,
& Evans, 2006; Hogg et al., 2001; Krentz, Auld, & Gill, 2004; Sterling, Chaisson,
Keruly, & Moore, 2003).
Delayed diagnosis among MSM has additional implications for disease progres-
sion and transmission prevention. MSM who are unaware of their infection are more
likely to engage in sexual risk taking (Marks, Crepaz, Senterfitt, & Janssen, 2005),
placing others at risk for acquiring HIV, putting themselves at risk for contracting
other sexually transmitted infections (STIs), weakening their already compromised
immune systems (Wiley et al., 2000), and increasing the potential for superinfection
with multiple strains of HIV (Sidat et al., 2008). Additionally, a recent study evaluat-
ing the primary drivers of the continued HIV epidemic among MSM estimated that
24%–31% of new HIV infections among MSM can be attributed to having sex with
partners who are unaware of their HIV infection (Goodreau et al., 2012).
The HIV risk-taking literature is giving increasing attention to establishing
and testing HIV risk models that address not only sociodemographic risk factors,
but also personal and contextual factors (Traube, Holloway, & Smith, 2011; Zea,
Reisen, Poppen, & Bianchi, 2009). Although this literature is primarily focused on
behavioral risks for HIV acquisition or transmission, the same theoretical consid-
erations can be taken into account when conceptualizing reasons for delayed HIV
diagnosis. A theoretical approach that addresses personal and contextual factors
that has received growing support in the HIV-risk literature is Social Action Theory
(Ewart, 1991). Applying Social Action Theory to delayed HIV testing, one could
propose that self-regulatory action (e.g., getting tested for HIV) is influenced not
only by the characteristics of the person (e.g., self-efficacy to engage in protective
health behaviors), but also by the social, physical, and affective context in which the
action occurs (e.g., social support, housing, mental health status).
Sociodemographic characteristics of people with late (commonly defined as
HIV diagnosis within 12 months of AIDS diagnosis) and delayed (commonly de-
fined as HIV diagnosis 1 year or more after HIV infection) HIV diagnosis often have
been assessed. There has been limited research addressing personal and contextual
factors that may affect delayed testing (Kozak, Zinski, Leeper, Willig, & Mugavero,
2013; MacKellar et al., 2005; Mukolo, Villegas, Aliyu, & Wallston, 2013; Nelson
et al., 2010). In this article, we use qualitative and quantitative data from the Seattle
Area MSM Study (SAMS) to conduct a preliminary assessment of potential personal
and contextual factors that may contribute to delayed HIV diagnosis among MSM.
Specifically we hypothesized that participants who reported increased personal and
contextual stressors would be more likely to have delayed diagnosis. Stressors were
conceptualized as factors that exhibit the potential to induce negative changes in the
individual’s psychological, physiological, and immunological equilibrium (Paterson
124 NELSON ET AL.
& Neufeld, 1989). To enrich our understanding of the relation between the level of
stressors and delayed status, we compare overall patterns of stressors between de-
layed versus nondelayed testers and provide prototypical case examples.
STUDY DESIGN, ELIGIBILITY, RECRUITMENT, AND DATA
SAMS was a cross-sectional survey of HIV risk factors among newly diagnosed
HIV-seropositive and HIV-seronegative MSM conducted by the Public Health-Seat-
tle & King County (PHSKC) HIV/AIDS Epidemiology Program under a cooperative
agreement with the CDC. The study collected data from July 2002 to May 2005.
Detailed descriptions of the recruitment and study procedures are provided else-
where (Thiede et al., 2009). Briefly, potential participants were passively referred to
the study after testing for HIV at the PHSKC sexually transmitted diseases (STD)
and HIV/AIDS Program (HAP) clinics, HAP community outreach testing sites, and
two university HIV clinics. MSM were eligible if they were age 18 or older, reported
sex with men during the preceding 6 months, and were able to complete the inter-
view in English. HIV-seropositive MSM were eligible if recruited within 3 months
of their first HIV-seropositive test result, and HIV-seronegative MSM were eligible if
recruited within 1 month of testing. A structured quantitative survey questionnaire
was administered to all participants via audio computer–assisted self-interviewing
(ACASI). All HIV-seropositive and half of the HIV-seronegative participants were
asked to participate in qualitative semistructured interviews.
All interviews were audio recorded with concurrent supplementary written
notes taken by the interviewer. Within a week of each interview, the interviewer
would create a write-up of the interview, including the notes. These write-ups would
include a combination of descriptions of the substantive content of the discussion
along with verbatim transcribed quotes and notes on the context of the quotes. Ver-
batim quotes were chosen based on their direct relation to the study goals, whereas
lengthy, less relevant, or off-topic sections of the interviews were summarized by the
interviewer in the write-up. This method has a long history of successful and reliable
use within traditional ethnography and is an efficient way to contain write-up costs
(Spradley, 1979). To validate their accuracy, write-ups were reviewed by two addi-
tional research team members. To establish and contrast factors associated with de-
layed and nondelayed testing, we conducted a content analysis using CDC EZ-Text
software (Carey et al., 2008). Two coders developed a codebook designed to identify
themes in the interview write-ups (Hruschka et al., 2004). Overall, final intercoder
reliability was that 85.4% of the codes had kappa values equal to or greater than
0.8. The codes were used to identify common behavioral patterns and to search the
database for the illustrative verbatim quotes presented here from the delayed and
nondelayed tester samples.
During the study period, all persons testing HIV-seropositive at the PHSKC
sites were routinely offered the less sensitive (LS) HIV-1 test, employing the sero-
logic testing algorithm to assess recent HIV seroconversion (STARHS), to estimate
recent infection in the testing population. All tests were performed by the PHSKC
Laboratory (Vironostika-LS EIA; Bio Merieux, Raleigh, NC). The estimated mean
time from seroconversion (defined using a standardized optical density cutoff of 1.0)
with the Vironostika-LS is 170 days (95% CI [145, 200 days]; Kothe et al., 2003).
FACTORS RELATED TO DELAYED HIV DIAGNOSIS 125
To be consistent with our previous article on delayed diagnosis among this popula-
tion (Nelson et al., 2010), delayed testers were defined as MSM who seroconverted
6 months or more before testing HIV-seropositive, based on a positive HIV-1 test
combined with a reactive LS HIV-1 test (according to STARHS) or, in the case of
a missing LS HIV-1 test, lack of a self-reported HIV-negative test during the past 6
months. Nondelayed testers were defined as those with infections acquired within
the preceding 6 months, based on a positive HIV-1 test combined with a nonreactive
LS HIV-1 test (according to STARHS) or self-reported HIV-seronegative test during
the past 6 months.
Among participants in whom we relied on self-reported last HIV-seronegative
test, attempts were made to verify the date of last test by contacting the participant’s
medical provider or through PHSKC HIV testing records. Through this method we
were able to verify the date of last HIV-seronegative result for three of the nonde-
layed testers. In addition, six of the nondelayed testers were verified to be have infec-
tions due to their diagnosis with a syndrome of primary infection at an HIV clinic
that specialized in diagnosis and treatment of new infections in addition to their
self-reported prior HIV-seronegative tests.
Two participants had reactive LS HIV-1 results while having a self-reported
HIV-seronegative test in the past 6 months. In these cases, recency of infection was
decided by the study investigators based on information the client provided in the
quantitative and qualitative interviews about testing history and usual frequency of
testing. Given these participants’ reported frequency of testing and the amount of
detail, including dates and locations, provided about their last HIV-seronegative test
result, these two participants were deemed to be nondelayed testers.
Overall, quantitative and qualitative analyses were conducted on 75 of the 77
participants who tested HIV-seropositive and had coded qualitative interview data,
including 38 delayed and 37 nondelayed testers. This study was approved by the
Institutional Review Boards of the University of Washington and the CDC. All par-
ticipants provided informed consent.
To create a description of our sample, chi-square and Fisher’s exact tests were
used to compare delayed and nondelayed testers with respect to sociodemographic
variables. Sociodemographic variables included recruitment site (HAP clinic, HAP
outreach, STD Clinic, and other sites), age (18–29, 30–39, 40+), race/ethnicity
(White, Black/African American, Hispanic, other), education (high school or less,
more than high school), sexual identity (gay, bisexual), extent they had revealed their
sexual practices (“out” to > 50% of the people they knew about having male-male
sex, “out” to ≤ 50% of the people they knew about having male-male sex), and
health insurance (yes/no).
To establish variables that would likely increase the level of stressors in a par-
ticipant’s life, we analyzed two questions from the semistructured interview: (a) “Be-
fore you found out your HIV results, had anything big happened in your life in the
last year? For example, did you move, lose or start a job, did a relationship end or
begin, etc.” and (b) “Do you have any history of depression or other mental health
issues?” All participants were asked these questions with additional specific probes
to further engage the participant as necessary. For example, for the question “Before
you found out your HIV results, had anything big happened in your life in the last
year?” the probes included: (a) “Was there anything from before a year ago that had
a bigger impact for you?”; (b) “What happened?”; (c) “Where did it happen?”; (d)
126 NELSON ET AL.
“When did it happen?”; (e) “How did it happen?”; (f) “Who was there?”; (g) “How
did you deal with this?”; (h) “What helped you cope?”; and (i) “How did things
Four main variables in the qualitative data were hypothesized to increase the
level of stressors in a participant’s life. These included any experiences of violence
(yes/no), relationship issues (yes/no), or substance use issues (yes/no) in the past 12
months and a self-reported history of hospitalization for mental illness (yes/no).
Participants were coded as having experienced violence if they reported: (a) doing
something violent, acting out, or being malicious; (b) being physically abused or
hurt; (c) being sexually abused, molested, or raped; or (d) being verbally or emotion-
ally abused or hurt in the past 12 months. Relationship issues were defined as par-
ticipants who self-reported experiencing significant events in their romantic, family,
or friendship relationships (e.g., starting a new relationship, ending a relationship,
relationship conflict) in the past 12 months. Participants who self-reported increases
in illicit drug or alcohol use, attempting to become sober, overdosing, and/or enter-
ing detoxification treatment or other substance use–related programs in the past 12
months were coded as having substance use issues. To supplement the qualitative
data, we also identified behavioral variables in the quantitative dataset that we hy-
pothesized would likely increase stress in participants’ lives. These included current
homelessness (yes/no) and history of incarceration (yes/no). Homelessness was de-
fined as self-reporting being homeless or having a nonpermanent residence. History
of incarceration was defined as self-reporting ever having spent a night in jail.
A count variable was created to represent the overall level of stressors experi-
enced in a participant’s life, with a 0 corresponding to not having experienced any
of the hypothesized stressors and a 6 corresponding to having experienced all of the
hypothesized stressors. Participants were considered to have encountered significant
stressors in their lives if they experienced two or more of the proposed stressors.
Fisher’s exact tests were used to compare delayed and nondelayed testers with re-
spect to the individual stressors and the overall stressor variable (0–1, 2+).
Overall patterns, case examples, and illustrative quotes were extracted from the
qualitative data to enrich our understanding of the relationship between the level
of stressors and delayed status. We established four case examples: an HIV-sero-
positive participant who was delayed and reported two or more stressors, an HIV-
seropositive participant who was nondelayed and reported two or more stressors,
an HIV-seropositive participant who was delayed and reported 0–1 stressor, and an
HIV-seropositive participant who was nondelayed and reported 0–1 stressor.
Among the 77 HIV-seropositive MSM enrolled in SAMS, 75 had coded qualita-
tive interview data, including 38 (51%) delayed and 37 (49%) nondelayed testers.
As reported previously (Nelson et al., 2010), delayed testers were more likely to be
older (18–29 years old: 21% vs. 30%; 30–39 years old: 29% vs. 54%; and 40+ years
old: 50% vs. 16%, p < .01), Black/African American (26% vs. 3%, p < .01), and
“out” to 50% or less of the people they knew about having male-male sex (45% vs.
14%, p < .01) compared to nondelayed testers (Table 1).
FACTORS RELATED TO DELAYED HIV DIAGNOSIS 127
PROPOSED STRESSORS AND THEIR RELATION WITH DELAYED
An analysis of the proposed stressors in the qualitative and quantitative data re-
vealed that delayed testers were not more likely to report having experienced violence
(26% vs. 19%, p = .58), relationship issues (53% vs. 51%, p = 1.0), or substance use
issues (34% vs. 30%, p = .81) in the past 12 months; a history of hospitalization for
mental illness (21% vs. 14%, p = .54); or a history of incarceration (50% vs. 30%,
p = .10) compared to nondelayed testers (Table 2). Delayed testers were more likely
to be homeless (32% vs. 5%, p = .01). Interestingly, although the majority of the
proposed stressors were not significantly different between delayed and nondelayed
testers, delayed testers were significantly more likely to have experienced two or
more of the stressors (68% vs. 38%, p = .01) compared to nondelayed testers.
Upon further exploration of the narratives provided by the participants, we dis-
covered multiple patterns. Specifically, delayed testers with two or more life stressors
often described a long history of substance use and untreated or only sporadically
treated mental illness. Furthermore, they reported feeling a lack of agency or control
over their own behaviors. Lastly, they reported a lack of social support and an inac-
curate appraisal of their HIV risk. In contrast, nondelayed testers with two or more
life stressors described similar issues with substance use, mental health, and mental
health treatment; however, nondelayed testers tended to report a sense of control
and agency over their own behaviors as well as an accurate appraisal of their HIV
TABLE 1. Sociodemographic and Disclosure Characteristics of the 75 HIV-Infected Seattle Area MSM
Study Participants by Delayed Testing Status, Seattle, Washington, 2002–2005
Total Delayed testersNondelayed testers
N = 75
n = 38
n = 37
Recruitment site 0.12
HAP clinic, outreach, and other sites 51 (68.0)29 (76.3) 22 (59.5)
STD clinic24 (32.0)9 (23.7)15 (40.5)
Age in years < 0.01
18–2919 (25.3)8 (21.0)11 (29.7)
30–3931 (41.3)11 (29.0)20 (54.1)
40+25 (33.3)19 (50.0)6 (16.2)
White 44 (58.7)19 (50.0)25 (67.6)
Black/African American11 (14.7)10 (26.3) 1 (2.7)
Hispanic 11 (14.7)7 (18.4)4 (10.8)
Other9 (12.0) 2 (5.3)7 (18.9)
High school or less education20 (26.7)13 (34.2)7 (18.9)0.13
Gay sexual orientation63 (88.7)32 (84.2)31 (93.9)0.27
Out to ≤ 50% about male-male sex 22 (29.3)17 (44.7) 5 (13.5)< 0.01
Health insurance 24 (32.9)13 (36.1)11 (29.7)0.56
Note. Participants were recruited from the Public Health-Seattle & King County (PHSKC) sexually transmitted
diseases (STD) and HIV/AIDS Program (HAP) clinics, HAP community outreach testing sites, and two university HIV
128 NELSON ET AL.
Delayed testers with fewer than two life stressors often described no or few
substance use issues; overall good mental health and/or a stable mental health treat-
ment history; having a sense of agency over their own behaviors; and having a steady
social support system. Similar to delayed testers with two or more life stressors, de-
layed testers with fewer than two life stressors also appeared to have an inaccurate
appraisal of their own HIV risk, often due to a false sense of security in the context
of a long-term relationship. Nondelayed testers with fewer than two life stressors, on
the other hand, also reported having few to no substance use issues, a stable mental
health or mental health treatment history, a stable social support system, and a sense
of agency over their own behaviors, but, in contrast to the delayed testers with fewer
than two life stressors, they tended to be accurately aware of their own HIV risk.
One HIV-seropositive delayed tester who experienced two or more stressors in
his life was Black/African American, homeless, 30–39 years old, had a high school
diploma or lower level of education, and did not have health insurance. He de-
scribed himself as experiencing a lifetime of suffering from substantial alcohol use
and mental health issues, including being hospitalized on multiple occasions. He
also reported having emotional and financial problems related to his alcohol use. He
described spending all of his money on alcohol and being unable to afford things like
housing or food. Furthermore, he said that his drinking isolates him from his family;
they do not want to be around him when he is drinking because he gets “obnoxious
and smart” when he drinks.
In the year prior to testing, he reported moving to the Seattle area to seek work,
but remained unemployed. He described himself as sad about being away from fam-
ily and overall feeling socially isolated and lonely; “I don’t have too many friends.”
He reported that he has sex with men not because he wants sex, but instead as a
means of having someone to talk to and be physically close with. He described meet-
ing most of his sexual partners in bars or public parks and said he is incapable of
meeting, conversing, or having sex with people without being drunk. He reported
feeling used by men and that the men he has sex with are only having sex with him
TABLE 2. Variables Hypothesized to Increase Participants’ Stressors Among the 75 HIV-Infected Seattle
Area MSM Study Participants by Delayed Testing Status, Seattle, Washington, 2002–2005
N = 75
n = 38
n = 37
Experienced violence in past 12 months1722.7 1026.3718.90.58
Experienced relationship issues in past 12 months 3952.0 2052.6 1951.41.00
Experienced substance use issues in past 12 months 2432.0 1334.2 1129.7 0.81
History of hospitalization for mental illness 13 17.38 21.15 13.5 0.54
Homeless 1418.7 12 31.62 5.40.01
History of incarceration 3040.01950.01129.70.10
2+ of the above stress variables 4053.3 2668.4 1437.80.01
Note. Violence = (1) doing something violent, acting out, or being malicious; (2) being physically abused or hurt; (3)
being sexually abused, molested, or raped; or (4) being verbally or emotionally abused or hurt; Relationship issues =
significant events in their romantic, family, or friendship relationships; Substance use issues = illicit drug or alcohol
use, attempting to become sober, overdosing, and/or entering detoxification treatment or other substance use related
programs; Homelessness = being homeless or having a non-permanent residence; History of incarceration = having
spent a night in jail.
FACTORS RELATED TO DELAYED HIV DIAGNOSIS 129
to get something from him (e.g., place to stay, money). He stated that they “are not
An HIV-seropositive nondelayed tester who reported two or more stressors was
White, homeless, 30–39 years old, had a high school diploma or lower level of edu-
cation, and no health insurance. He reported a long history of mental health issues
for which he had attempted, without much success, to get treatment through the
public health system. He reported significant drug use through his lifetime, including
pot, speed, alcohol, and heroin. He said that most people would describe him as a
“junkie, thief, liar, hypocrite, careless, liar, scared, immature…”
He reported being tested because his primary partner had recently tested HIV-
seropositive. He reported feeling that it is “the destiny of the gay population to go
out with AIDS” and that gay people deserve AIDS because “it is unnatural, un-
healthy for the gay community to identify itself with their sexual practices.” He
further reported that in the past he had reduced his risk of contracting HIV by let-
ting only “love affair” partners ejaculate inside him. He did not ask his sex partners
about their HIV status and said, “I don’t want to know.” He feels that even if he
asked them, they could be lying, so he assumes all his partners are positive. He was
a prostitute for a long time and felt “invincible” from HIV because he did not pick
it up while hustling. However, at this point in his life, he felt very “used up” and
described having to “work to make sex a positive experience.”
An HIV-seropositive delayed tester who reported 0–1 stressors was a Hispanic,
40+ year old who lived with a spouse/partner, is a college graduate, and has health
insurance. In the past, he reported being treated for depression after breaking up
with a partner of 11 years. He stopped seeking services (therapy and medication)
once he recovered, and said he had not had problems with depression since then. He
described himself as “sincere, honest, warm, generous.”
He described liking monogamous, long-term relationships, and said he has had
four 3+ year relationships/sex partners since he came out as gay. He reported that he
did not like to do drugs or party. He noted that he and his most recent partner did
not have sex until they had been dating for 3 months. After 2 years, they stopped
using condoms because his partner told him the “last time I tested, I tested negative”
before they started having sex and he felt “like he was my soul mate—that this was
pretty much it for me.” He reported not feeling at risk because of this. He did not
identify anything stressful in the year before he tested HIV-seropositive.
An HIV-seropositive nondelayed tester who reported 0–1 stressors was White,
under 25 years old, living with roommates or friends, and was a college graduate
with no health insurance. He reported ongoing medication management for an anxi-
ety disorder and a history of depression that resolved 3 years ago. He attributed his
depression to the stress he experienced after coming out. He said that all the mental
energy it was taking to appear straight was suddenly available for other things. “I
came out—and I think it was like the first time that I was using my whole brain! So
it was a little overwhelming.”
He reported routinely testing for HIV and that he met the man he believes in-
fected him at a bathhouse. He described not normally meeting men in that kind of
venue, but he was feeling curious and it was available. He was about to leave the
area after graduating from college and wanted to experience that element of gay
culture. “Why am I faced with this issue and it doesn’t seem to be there for anybody
else. The immediacy of if you’re horny and want to have sex, if you’re gay—it’s so
easy … for straight guys it’s just totally not.” He described that the existence of bath-
130 NELSON ET AL.
houses is wrong and dirty, and he has to be “tipsy” to go to them. He did not identify
anything stressful in the year before he tested HIV-seropositive.
HIV testing is one of the primary HIV prevention strategies supported by the U.S.
government. With the accessibility of HIV testing as well as the significant personal
benefits of timely HIV diagnosis, it is remarkable that a substantial proportion of
MSM delay HIV testing and remain unaware of their HIV infection (Chen et al.,
2012). While previous research in this area has attempted to understand the sociode-
mographic factors that are related to delayed HIV diagnosis, limited work has ex-
amined personal and contextual factors (Kozak et al., 2013; MacKellar et al., 2005;
Mukolo et al., 2013; Nelson et al., 2010). Our study sought to provide a preliminary
view of the complicated interactions between personal and contextual factors that
may lead to delayed testing among MSM.
In line with Social Action Theory, in our sample, getting tested for HIV ap-
pears to be influenced by a confluence of personal characteristics (e.g., self-efficacy
to engage in protective health behaviors), as well as the social, physical, and affective
context in which the testing occurs (e.g., social support, housing, mental health sta-
tus). Specifically, we found that the majority of the personal and contextual factors
we hypothesized would increase the level of stressors in a participant’s life may not
be individually associated with delayed HIV diagnosis. However, participants who
indicated that they experienced two or more stressors in the past year were signifi-
cantly more likely to have delayed testing.
Additionally, in our further exploration of the participants’ narratives, we iden-
tified some prominent patterns involved in the relationship between stressors and
delayed diagnosis. Specifically, it appears that there may be a few factors that differ-
entiate those who are able to cope with multiple stressors and continue to engage in
timely health care seeking behaviors versus those who are not. Our data indicate that
men with stable social support and mental health as well as self-efficacy to engage
in protective health behaviors may be more equipped to cope with multiple stressors
and continue to engage in timely health care seeking behaviors. On the other hand,
men who are experiencing multiple stressors but who are lacking these supportive
factors may not have the coping skills necessary to engage in timely health care seek-
ing behaviors. Interventions attempting to address delayed diagnosis among MSM
should specifically target social support, mental health, and self-efficacy to engage
in protective health behaviors as well as integrate techniques to address coping skill
Additionally, it is possible that MSM who are experiencing multiple life stress-
ors without the scaffolding of social support, stable mental health, and self-efficacy
to engage in protective health behaviors are unable to prioritize HIV testing over
more salient and pressing issues. It is likely that MSM who are experiencing multiple
stressors, such as being homeless and experiencing violence, feel a need to address
their physiological and safety concerns before they are able to prioritize HIV testing.
If this is true, it indicates the further need for structural interventions addressing
these issues in addition to interventions to increase social support, mental health
treatment, and self-efficacy to engage in protective health behaviors.
Lastly, it appears that recognition of HIV risk is an important factor in delayed
diagnosis. Delayed testers, regardless of level of stressors, predominantly indicated
FACTORS RELATED TO DELAYED HIV DIAGNOSIS 131
that they were unaware of their HIV risk. This result is in line with previous research
which found that the majority of MSM either with late diagnosis or undiagnosed
HIV infection perceived themselves to be at low risk for acquiring the virus (MacK-
ellar et al., 2005; Schwarcz et al., 2011). While obsessing about HIV risk is likely to
be maladaptive, averting attention from it may lead to delayed testing and increased
risk taking. As we found in our previous work (Nelson et al., 2010), many of the
delayed testers reported that they did not frequent MSM-specific venues or identify
strongly with the MSM community. As such, effective messaging and interventions
hoping to increase recognition of HIV risk among MSM who delay testing may need
to target men generally instead of MSM specifically and be distributed more widely
to locations that are non-MSM specific.
To our knowledge, only one previous study has explored personal and contex-
tual factors related to delaying testing for HIV (Schwarcz et al., 2011). In a recent
qualitative and quantitative study of 41 persons who developed AIDS within 12
months of their HIV diagnosis (i.e., late testers), Schwarcz et al. (2011) found that
the barriers to testing were multiple and often interwoven. Barriers in that study
included fear of receiving an HIV diagnosis, concerns about ability to pay for treat-
ment, inaccurate knowledge about treatment, inaccurate assessment of HIV risk,
and the presence of mental health issues. Because late testers are a subset of delayed
testers, it is likely that many of the same personal and contextual factors associated
with late testing may also be associated with delayed testing; these patterns were
partially supported by our previous findings (Nelson et al., 2010) and in this current
As with any study, there are limitations that should be kept in mind when con-
sidering these results. First, due to our limited sample size, and thus limited power, it
is possible that some of the individual characteristics that were assessed (e.g., history
of incarceration, history of hospitalization for mental illness) may in fact be indi-
vidually associated with delayed diagnosis. As such, we consider this a preliminary
assessment of these factors and encourage further research with a larger sample size
to fully explore the influence of these characteristics on delayed diagnosis. Second,
we recognize that our data are now approximately 10 years old and awareness of
HIV risk as well as HIV testing patterns may have changed since we collected our
data. Although this is the case, we believe that these personal and contextual fac-
tors are likely to continue to influence the timeliness of HIV testing and should be
explored further. Third, the use of self-reported behaviors and in-person qualitative
interviews may be prone to social desirability and recall bias. Fourth, due to the
demographic make-up and the recruitment locations (Thiede et al., 2009), the study
population may not be representative of all MSM in the broader community who
are recently diagnosed with HIV. Finally, there are potential limitations in our use of
a cross-sectional design and the associated classification of delayed and nondelayed
testers. Given the window period for the Vironostika-LS HIV-1 test using STARHS
(145–200 days), it is possible that while a person was reactive on the LS HIV-1 test,
they in fact seroconverted within the past 6 months (Kothe et al., 2003). Addition-
ally, because not all participants were evaluated using STARHS, in some cases we re-
lied on self-reported HIV testing history to establish whether a person seroconverted
within 6 months of testing. Thus, it is also possible that while a person did not test
within the past 6 months, he, in fact, seroconverted within the past 6 months. While
this may be the case, as reported previously (Nelson et al., 2010), the majority of
participants classified as delayed testers identified an illness associated with HIV or
an illness that persisted much longer than would normally be expected in an immu-
132 NELSON ET AL.
nocompetent person as the reason for testing. In addition to the majority not having
tested within the past 2 years, it seems likely that most delayed testers seroconverted
longer than 6 months prior to their HIV-seropositive test.
Overall our findings indicate that personal and contextual factors likely work
in concert to influence delayed HIV diagnosis among MSM. Specifically, MSM who
experience multiple life stressors, whether they are personal or contextual, have an
increased likelihood of delaying HIV diagnosis. Additionally, factors such as social
support, mental health, and self-efficacy to engage in protective health behaviors as
well as an accurate recognition of HIV risk appear to be of particular importance,
making MSM with these characteristics ideal targets for intervention. Because early
HIV diagnosis is important on both personal and community levels, interventions
targeting the enhancement of coping skills, self-efficacy to engage in protective health
behaviors, social support, and mental health treatment as well as structural interven-
tions targeting physiological and safety concerns will likely help MSM handle their
life stressors more effectively and engage in more positive health behaviors, includ-
ing seeking HIV testing in a more timely manner.
Branson, B., Handsfield, H., Lampe, M., Janssen,
R., Taylor, A., Lyss, S., & Clark, J. (2006).
Revised recommendations for HIV testing
of adults, adolescents, and pregnant women
in health-care settings. MMWR Recom-
mendations and Reports, 55, 1–17; quiz
Carey, J. W., Wenzel, P. H., Gelaude, D., Sheridan,
J., Reilly, C., & Burns, D. (2008). CDC EZ-
Text: Software for collection, management
and analysis of semi-structured qualitative
databases. Atlanta, GA: Developed by Ma-
nila Consulting Group, Inc. for the CDC.
Centers for Disease Control and Prevention.
(2011). HIV testing among men who have
sex with men—21 Cities, United States,
2008. Morbidity and Mortality Weekly Re-
port, 60(21), 694–699.
Centers for Disease Control and Prevention.
(2012a). HIV surveillance report, 2010
(No. vol. 22). Retrieved from http://www.
Centers for Disease Control and Prevention.
(2012b). Monitoring selected national HIV
prevention and care objectives by using
HIV surveillance data—United States and
6 U.S. Dependent Areas—2010. Retrieved
Chadborn, T. R., Delpech, V. C., Sabin, C. A., Sin-
ka, K., & Evans, B. G. (2006). The late di-
agnosis and consequent short-term mortal-
ity of HIV-infected heterosexuals (England
and Wales, 2000–2004). AIDS, 20(18),
Chen, M., Rhodes, P. H., Hall, H. I., Kilmarx, P.
H., Branson, B., & Valleroy, L. A. (2012).
Prevalence of undiagnosed HIV infection
among persons aged ≥13 Years—National
HIV Surveillance System, United States,
2005–2008. Morbidity and Mortality
Weekly Report, 61(2), 57–64.
Ewart, C. K. (1991). Social action theory for a
public health psychology. American Psy-
chologist, 46(9), 931–946.
Goodreau, S. M., Carnegie, N. B., Vittinghoff, E.,
Lama, J. R., Sanchez, J., Grinsztejn, B., . .
. Buchbinder, S. P. (2012). What drives the
US and Peruvian HIV epidemics in men
who have sex with men (MSM)? PLoS
One, 7(11), e50522. doi:10.1371/journal.
Hogg, R. S., Yip, B., Chan, K. J., Wood, E., Craib,
K. J., O’Shaughnessy, M. V., & Montaner,
J. S. (2001). Rates of disease progression
by baseline CD4 cell count and viral load
after initiating triple-drug therapy. JAMA,
Hruschka, D., Schwartz, D., St. John, D. C., Pi-
cone-DeCaro, E., Jenkins, R. A., & Carey, J.
W. (2004). Reliability in coding open-ended
data: Lessons learned from HIV behavioral
research. Field Methods, 16, 307–331.
Kothe, D., Byers, R., Caudill, S., Satten, G., Jans-
sen, R., Hannon, W., & Mei, J. (2003). Per-
formance characteristics of a new less sen-
sitive HIV-1 enzyme immunoassay for use
in estimating HIV seroincidence. Journal of
Acquired Immune Deficiency Syndromes,
Kozak, M., Zinski, A., Leeper, C., Willig, J. H.,
& Mugavero, M. J. (2013). Late diagnosis,
FACTORS RELATED TO DELAYED HIV DIAGNOSIS 133 Download full-text
delayed presentation and late presentation
in HIV: Proposed definitions, methodologi-
cal considerations and health implications.
Antiviral Therapy, 18(1), 17–23.
Krentz, H. B., Auld, M. C., & Gill, M. J. (2004).
The high cost of medical care for patients
who present late (CD4 <200 cells/microL)
with HIV infection. HIV Medicine, 5(2),
Lorenc, T., Marrero-Guillamón, I., Aggleton, P.,
Cooper, C., Llewellyn, A., Lehmann, A., &
Lindsay, C. (2011). Promoting the uptake
of HIV testing among men who have sex
with men: Systematic review of effective-
ness and cost-effectiveness. Sexually Trans-
mitted Infections, 87(4), 272–278.
MacKellar, D., Valleroy, L., Secura, G., Behel, S.,
Bingham, T., Celentano, D., . . . Janssen, R.
(2005). Unrecognized HIV infection, risk
behaviors, and perceptions of risk among
young men who have sex with men: Op-
portunities for advancing HIV prevention
in the third decade of HIV/AIDS. JAIDS,
Marks, G., Crepaz, N., Senterfitt, J., & Janssen,
R. (2005). Meta-analysis of high-risk sexu-
al behavior in persons aware and unaware
they are infected with HIV in the United
States: Implications for HIV prevention
programs. JAIDS, 39, 446–453.
Mugavero, M. J., Amico, K. R., Horn, T., &
Thompson, M. A. (2013). The state of en-
gagement in HIV care in the United States:
From cascade to continuum to control.
Clinical Infectious Diseases, 57, 1164–
Mukolo, A., Villegas, R., Aliyu, M., & Wallston,
K. A. (2013). Predictors of late presentation
for HIV diagnosis: A literature review and
suggested way forward. AIDS and Behav-
ior, 17(1), 5–30.
Murphy, E. L., Collier, A. C., Kalish, L. A., Ass-
mann, S. F., Para, M. F., Flanigan, T. P., . . .
Viral Activation Transfusion Study Investi-
gators. (2001). Highly active antiretroviral
therapy decreases mortality and morbidity
in patients with advanced HIV disease. An-
nals of Internal Medicine, 135(1), 17–26.
Nelson, K. M., Thiede, H., Hawes, S. E., Golden,
M. R., Hutcheson, R., Carey, J. W., . . . Jen-
kins, R. A. (2010). Why the wait? Delayed
HIV diagnosis among men who have sex
with men. Journal of Urban Health, 87(4),
Paterson, R. J., & Neufeld, R. W. J. (1989). The
stress response and parameters of stressful
situations. In R. W. J. Neufeld (Ed.), Ad-
vances in the investigation of psychological
stress (pp. 7–42). Oxford, England: John
Wiley & Sons.
Schwarcz, S., Richards, T. A., Frank, H., Wenzel,
C., Hsu, L. C., Chin, C.-S. J., . . . Dilley, J.
(2011). Identifying barriers to HIV testing:
Personal and contextual factors associated
with late HIV testing. AIDS Care, 23(7),
Sidat, M. M., Mijch, A. M., Lewin, S. R., Hoy,
J. F., Hocking, J., & Fairley, C. K. (2008).
Incidence of putative HIV superinfection
and sexual practices among HIV-infected
men who have sex with men. Sex Health,
Spradley, J. (1979). The ethnographic interview.
New York, NY: Holt, Rinehart, and Win-
Sterling, T. R., Chaisson, R. E., Keruly, J., &
Moore, R. D. (2003). Improved outcomes
with earlier initiation of highly active an-
tiretroviral therapy among human immu-
nodeficiency virus-infected patients who
achieve durable virologic suppression:
Longer follow-up of an observational co-
hort study. Journal of Infectious Diseases,
Thiede, H., Jenkins, R., Carey, J., Hutcheson,
R., Thomas, K., Stall, R., . . . Golden, M.
(2009). Determinants of recent HIV infec-
tion among Seattle-area men who have
sex with men. American Journal of Public
Health, 99(Suppl. 1), S157–S164.
Traube, D. E., Holloway, I. W., & Smith, L. (2011).
Theory development for HIV behavioral
health: Empirical validation of behavior
health models specific to HIV risk. AIDS
Care, 23(6), 663–670.
Wiley, D. J., Visscher, B. R., Grosser, S., Hoover, D.
R., Day, R., Gange, S., . . . Detels, R. (2000).
Evidence that anoreceptive intercourse with
ejaculate exposure is associated with rapid
CD4 cell loss. AIDS, 14(6), 707–715.
Zea, M. C., Reisen, C. A., Poppen, P. J., & Bianchi,
F. T. (2009). Unprotected anal intercourse
among immigrant Latino MSM: The role
of characteristics of the person and the sex-
ual encounter. AIDS and Behavior, 13(4),