記住我
In order to successfully implement treatment, as well as prevention, as a national strategy to end the HIV epidemic in the United States, it is crucial to address and bridge gaps in the HIV care continuum. The vast majority of people living with HIV in the United States are men who have sex with men (MSM). Among MSM, there are marked race- and ethnicity-based disparities in HIV infection rates and engagement and retention in HIV care. Compared with White MSM, Black and Hispanic or Latino (henceforth, Latino) MSM living with HIV are less likely to be on antiretroviral therapy (ART), adhere to ART, and achieve viral suppression [-]. Despite the long-standing focus on individual-level drivers, disparities in HIV acquisition and care engagement and retention among Black and Latino MSM are influenced by factors that are often beyond the individual such as the social determinants of health (SDH) []. The SDH can shape environments and behaviors that increase vulnerability to HIV acquisition and negatively affect HIV outcomes (eg, viral suppression, morbidity, and mortality rates). Yet the role of the SDH on HIV outcomes is not widely studied; thus, the mechanisms of action are not well understood.
Research on neighborhood factors and the care continuum suggests that poverty and racial segregation, as well as access to transit and distance to pharmacies and medical care, are associated with HIV care–related outcomes [-]. Similarly, social network-level factors, such as high levels of social support and HIV serostatus disclosure to social network members, are associated with improved HIV care outcomes [,-]. Network and neighborhood characteristics are reciprocally related, with age composition, concentrated poverty, and racial segregation differentially fostering advantages and disadvantages for networks and residents alongside individual-level characteristics, such as socioeconomic status [-]. Still, additional research is needed to elucidate the pathways through which networks and neighborhoods foster racial and ethnic disparities in HIV care outcomes. The results of such analyses will provide an empirical basis for interventions and policies to fill gaps in the care continuum among Black and Latino MSM. To reach that goal, a systematic, theory-based investigation of the specific and intersecting neighborhood and network characteristics that relate to retention in care, ART adherence, and viral suppression must be used.
In this longitudinal study, which was recently completed with data analysis ongoing, we enrolled MSM living with HIV in New York City to complete study visits every 6 months, up to either 12 (post–COVID-19 cohort) or 36 (pre–COVID-19 cohort) months, to determine whether and how neighborhood and network characteristics influence their HIV care. Using socioecological and intersectional stigma and discrimination conceptual frameworks, we aim to examine neighborhood- and network-level characteristics that influence different aspects of the HIV care continuum among MSM living with HIV in New York City in a longitudinal study. We focus on 3 reported activity spaces (ie, potential neighborhoods of influence)—where MSM reside (home neighborhood), where they socialize and spend the most time (social neighborhood), and where they access their HIV medical care (health facility access neighborhood). We investigate the influence of four neighborhood-level characteristic domains: (1) community violence, physical disorder, and social disorganization (eg, interpersonal crime rates, concentrated poverty, and racial segregation); (2) neighborhood alcohol and other drug (AOD) use-associated factors (eg, alcohol outlets); (3) neighborhood-level social norms (eg, homophobia, HIV stigma, racial discrimination, and health norms); and (4) community resources (eg, HIV prevention services, social services, public transit access, and greenspace) and how these play out in the different neighborhoods in which MSM live, socialize, and receive HIV care.
The Neighborhoods, Networks and HIV Care among Men Who Have Sex with Men (NNHIV) study is a longitudinal cohort study that examines neighborhood and social network factors that influence HIV care outcomes (retention in care, ART adherence, and viral suppression) using spatial analysis and multilevel modeling among MSM living with HIV in New York City. The study investigates the influence of neighborhood composition and 4 neighborhood-level domains, testing theoretical pathways of influence across the different neighborhoods in which MSM live, socialize, and receive HIV care.
Ethical ConsiderationsThis study has been reviewed and approved by the New York Blood Center institutional review board (1066143). Written informed consent was obtained for participants 18 years of age and older who took part in the longitudinal study; an assent process was in place for any participants aged 16-17 years along with the use of a minor advocate. Participants received US $100 for their time and travel at the baseline visit and at each follow-up visit.
Eligibility CriteriaEligibility criteria included (1) reporting male sex at birth, (2) self-report being HIV-positive (including being newly diagnosed with acute HIV infection), (3) engaging in anal sex with a man in the past 6 months, (4) at least 16 years of age, (5) reside within the 5 New York City boroughs, (6) able to read and respond in English, and (7) provide written informed consent (or assent if between 16 and 18 years of age, not inclusive of 18 years). Only cisgender men were included in this study. We excluded other populations, including cisgender and transgender women, because the study specifically seeks to understand the influence of neighborhoods and networks on HIV care outcomes among MSM living with HIV, and therefore, would not be optimally designed to capture the unique needs and barriers impacting women.
Sample SizeSample size, power, and effect size considerations focused on longitudinal logistic regression models with neighborhood-, network-, and individual-level variables included as covariates. We aimed to enroll at least 40% who self-identify as Black, non-Hispanic and 30% who self-identify as Hispanic or Latino. Effect size calculations were performed assuming a minimum of 40% (n=220) of Black, non-Hispanic MSM and 30% (n=165) Latino MSM. Fixing type I error rate at 0.05, a sample size of 550 MSM, within-person correlation of 0.5, statistical power of at least 80%, and 10% attrition (loss to follow-up), minimum effect sizes (in terms of odds ratios) for detecting any difference in neighborhood and network effects on HIV care outcomes over the 5-visit, 24-month period range from 1.14 to 1.23. The proposed sample size of 550 MSM was later modified to 327 given the challenges of recruitment experienced during the COVID-19 pandemic.
RecruitmentWe enrolled a total of 327 MSM. Initially, convenience sampling for recruitment was used to recruit participants, which included in-person recruitment at venues including bars clubs, gay pride events, and HIV clinics and service organizations, internet-based ad placements (eg, Craigslist [Craigslist Inc], geospatial dating applications such as Grindr [Grindr Inc]), and referrals from enrolled participants (maximum 3 referrals per participant). Later, the recruitment strategy was revised to use a modified, venue-based, time-space sampling methodology, to ensure a diverse sample to increase generalizability. For in-person and internet-based recruitment, the sampling frame of venues included a range of neighborhoods across the 5 New York City boroughs (from traditionally considered gay enclaves to those with a less visible gay presence, from places where known HIV-infected individuals frequent to lesser-known areas) and a range of venues within neighborhoods, such as bars, clubs, HIV clinics and service organizations, gay pride events, and street locations. Each month, a sample of 8 to 16 locations and day-time periods were randomly selected.
Participants who enrolled and completed a study visit were provided with referral cards (either in person, via email, or text) in which they could refer their friends to the study, a maximum 3 referrals for US $10 for each referral who successfully screened for the study and enrolled into the study (maximum US $30 for total 3 referrals); this referral system was later amended with a limit of 1 referral per participant to reduce referral bias. For internet and mobile app–based recruitment, banner ads were placed on select websites and social media apps, including Facebook (Meta Platforms), Craigslist, Instagram (Meta Platforms), and Grindr.
Participants were asked to complete a brief internet-based eligibility screener if they indicated interest in participating in the study. Potentially eligible (based on the results of the screener) participants were asked to provide contact information to schedule their study visit. At the time their study visit appointment was scheduled, the participant was asked to bring in a copy of their prior HIV diagnosis report to the visit, if available.
Study enrollment began in March 2019 and was paused in March 2020 with an enrollment of 213 participants prior to the COVID-19 pandemic lockdown in New York City (henceforth “pre-lockdown cohort”). As a result, recruitment and enrollment into the study was paused from March 2020 to August 2021. Follow-up visits for the 213 participants in the pre-lockdown cohort used a hybrid visit model initially during the beginning of the pandemic (remote and in-person procedures), followed by a transition back to all in-person visits; follow-up visits were extended from month 24 to month 36 for a longer follow-up period for the pre-lockdown cohort. The study reopened for enrollment in August 2021 with 114 participants enrolled (“post-lockdown cohort”) before enrollment finally closed in November 2022. The post-lockdown cohort participants were followed to month 12, with the last visits completed in December 2023. A total of 327 participants enrolled in the longitudinal study.
Study Visit ProceduresBaseline Visit Study ProceduresStudy visits took place at our research site in Manhattan. After written informed consent (assent for those aged 16-17 years), participants underwent a rapid HIV antibody test (Orasure Oraquick ADVANCE Rapid HIV-1/2 Antibody) to confirm their self-reported HIV status if they did not bring a copy of their prior HIV diagnosis report. Participants were also asked to bring their ART medications for a pill count by staff and to sign a medical release form, with medical provider office visit notes obtained to confirm attendance and missed visits. They also had blood collection, which was sent to a local laboratory for HIV viral load and CD4 T lymphocyte (CD4 T cell) count testing. Test results were provided to participants, and referrals for HIV medical care and risk reduction counseling were provided and documented, as needed. The study visit ended for those who did not provide an HIV status report or whose HIV status was not confirmed.
Participants were then asked to identify their home, social, and health facility access neighborhoods by a trained research counselor using a neighborhood locator questionnaire and a desktop version of Google Earth. First, the interviewer asked for the name of their home neighborhood based on a drop-down menu of all 343 New York City neighborhoods. The interviewer then assisted participants in using Google Earth to “drop a pin” at the closest intersection near their home (home neighborhood), with the latitude and longitude coordinates of the pin drop recorded. Data on the perceived activity spaces (geographical neighborhoods) were collected with the question “When you think about your home neighborhood, what area do you usually think of?” with response categories “the block you live on, the area within (5 blocks/10 blocks/larger than 10 blocks) around the place you live.” This process was repeated for the neighborhood where they most often hang out or spend the most time (social neighborhood), and where they most often went for their HIV medical care in the prior 6 months, as well as the pharmacy where they get your HIV medications filled in the last 6 months (health facility access neighborhoods).
Participants then completed an egocentric social network inventory (SNI) which provides network-level covariate data. The inventory was administered by a trained research counselor, who entered participant responses into a computer-based program (EgoWeb 2.0 for Windows). For the social network inventory, participants were asked to name persons, using a name generator, whom they could rely on for support in the domains of (1) companionship—who do you get together with to spend time talking, relaxing, or hanging out with?; (2) confidant—who do you talk to about very personal and private things?; (3) financial—who would you ask to lend you US $100 if you needed it?; (4) housing—who would provide you with housing or a place to stay a few days or more if you needed it?; (5) medical—who would go to a medical appointment, such as to an HIV provider or mental health provider, with you?; and (6) medication adherence—who encourages you to take your HIV medications regularly or who you could talk to about problems with taking your HIV medications? In order to construct a comprehensive network composition of the participants, there were no restrictions on the number of network members named. The attributes that were asked about each social network member are (1) demographics (eg, age, gender, and race or ethnicity); (2) social member relationship type (sex partner, friend, family member, coworker, etc); (3) HIV status and HIV serostatus disclosure; (4) home borough and proximity (live within 20 blocks of ego); (5) communication, including duration of relationship, frequency and content of communication, distance, conflicts, and support; (6) perceived behavioral and attitudinal norms including norms for AOD use, ART adherence, medical care follow-up, HIV serostatus disclosure, and medical mistrust; (7) whether ART drugs are shared or sold to or bought from network member; and (8) whether the person knows other members of the social network. In addition, the participants provided information about social support from each named social network member, adapted from the Inventory of Socially Supportive Behaviors [], consisting of 19 items grouped into 3 sub-domains to assess cognitive guidance (9 items, eg, [network member] suggested some action that you should take), emotional support (5 items, eg, [network member] expressed interest and concern in your well-being), and tangible assistance (4 items, eg, [network member] gave you over US $50).
Next, participants completed the computer-assisted survey instrument (CASI) questionnaire, with the first part of the CASI involving ART use and pill count completed with study staff and then the rest of the CASI completed alone in private with staff assisting as needed. The CASI included measures that tap into common (eg, depression, anxiety, and substance use perceived patient-provider interaction) and network and neighborhood-level factors that may also be related to HIV care outcomes (eg, perceived racism in health care, homophobia and HIV stigma, empowerment, and coping, racial or ethnic and gay community attachment and acculturation; ).
Table 1. Independent variables.MeasureVariables (measures)SociodemographicsaART: antiretroviral therapy.
bNYC: New York City.
Follow-Up Study Visit ProceduresEvery 6 months for up to 12 months (post–COVID-19 lockdown cohort, after August 2021) or 36 months (pre–COVID-19 lockdown cohort with participants enrolled prior to the onset of the COVID-19 pandemic in New York City in March 2020, which caused the city to go on pause or lockdown), participants completed the neighborhood locator questionnaire using Google Earth, SNI, and CASI questionnaire similar to those completed at the baseline visit; they underwent CD4 T cell and HIV viral load testing that were sent to a local laboratory. To assess network turnover, the SNI collected information on any new social network members from the prior 6 months and on any persons named previously who were no longer part of the networks. Medical records review and pill count were performed. Participants were reimbursed US $100 at each follow-up visit.
If participants were unable to come to the study site for their follow-up visit (for various reasons, including during lockdown in New York City during the COVID-19 pandemic), they were provided with the option of completing part of the follow-up visit remotely. For this remote visit, they were contacted by staff by secure videoconference or phone and completed surveys remotely on a computer via secure links. Participants then made an appointment with study staff to come into the site in person within 2 weeks for HIV viral load and CD4 T cell count testing.
Study RetentionAt the baseline visit, extensive locator information was obtained using a locator form, including the participant’s address and phone number. We also collected names, telephone numbers, and addresses of at least 2 secondary contacts, with an emphasis on stable relatives both within and outside of New York City, which allowed us to maintain contact with residentially mobile or marginally housed individuals. We asked the participant for specific instructions on how to leave messages to protect confidentiality. Retention efforts for the longitudinal study included reminders in between study visits, use of contacts, telephone reminders prior to follow-up visits, texts, and emails. If a participant missed a visit, the retention protocol was initiated by study staff, which included letters sent to participants’ homes and reaching out to the listed secondary contacts on the locator form.
Statistical AnalysisMeasuresThe primary biological outcome is viral suppression (HIV viral load <20 copies per mL based on testing). Secondary outcome measures include retention in care (ie, >1 visit with an HIV medical provider in the last year) and retention in continuous care (ie, >2 visits at least 3 months apart with an HIV medical provider in the last year) as reported by CASI and confirmed by medical records review. ART adherence is based on the AIDS Clinical Trials Group (ACTG) [] 4-question adherence instrument as reported by CASI and confirmed by pill count. Participants are categorized as adherent if they report taking ≥85% of the dispensed dosage. The geospatial database, which characterizes the demographic and socioeconomic conditions of the neighborhoods, provides data for the four neighborhood characteristic domains of interest that are (1) community violence, physical disorder, and social disorganization; (2) neighborhood AOD use-associated factors; (3) neighborhood-level social norms; and (4) community resources.
The geospatial database, which characterizes the demographic and socioeconomic conditions of the neighborhoods, provides data for the four neighborhood characteristic domains of interest that are (1) community violence, physical disorder, and social disorganization; (2) neighborhood AOD use-associated factors; (3) neighborhood-level social norms; and (4) community resources.
lists the independent variables that are collected by CASI. Data on the participants’ perceptions of neighborhoods were obtained by CASI () while lists neighborhood characteristics based on archival data. As there are no area-level indicators available for some measures, select constructs from will be aggregated up to the area level to create neighborhood-level variables used in the analyses, similar to prior work [-].
Table 2. Neighborhood characteristics assessed by participants.MeasureDefinitionsExposure to neighborhoodsLevel of exposure to both residential and other neighborhoods, including length of residence, who lives with them, assigned housing, how much time they spend in their neighborhoods, and how much time they spend in other neighborhoods [,]Social tiesNumber of friends or relatives respondents have in the neighborhood and the frequency with which the respondents see friends and relatives based on Sampson’s assessment of social ties []Social cohesion and trustPerceptions of connectedness and trust in their community [-]Social or neighborhood involvementDegree to which respondent is involved in local community or other groups (eg, local neighborhood groups or block associations), degree to which respondents interact with their neighbors will be assessed, using a sociability measure []Perceived disorder, safety, and community violencePerceived social disorder and physical decay based on Ross and Mirowsky []; Screen for Adolescent Violence Exposure (SAVE): 32-item self-report measure to assess frequency and type of violence []Neighborhood integration, sense of community, and identification with neighborhoodNeighborhood integration will be measured using Perkin’s Sense of Community Scale []. 12-item 3D Strength of Identification Scale assesses self-reported feeling of identification with residential neighborhood []Drug market presencePresence of open air drug markets, street corner sales, corner store or bodega sales and known residence-based sales will be assessed directly using adaptation of previous assessments []Social norms: drug use, sexual behaviors, violence, HIV medical care, and general healthQuestions modified from the National Survey on Drug Use and Health [] to measure behavioral and attitudinal norms around drug use and violence. Perceptions of neighborhood social norms concerning multiple partners, safer sexual behaviors, condom use, receiving HIV medical care, and general healthSocial norms: neighborhood homophobia and HIV stigma or discriminationBrief 6-item scale assesses experiences of and responses to race and sexual orientation–based discrimination in different domains [], HIV stigma will be assessed using Earnshaw’s HIV stigma framework []Table 3. Neighborhood characteristics from archival data.DimensionDataCommunity violence, physical disorder, and social disorganizationaAOD: alcohol and other drug.
Data Analysis 1: Clustering and Spatial Trends in HIV Care OutcomesFirst, we will use spatial analysis to characterize clustering and spatial trends in HIV care outcomes (retention in care, ART adherence, and viral suppression) among MSM living with HIV in the study. Mapping the data distributions of outcomes will provide visual appraisals and display potential spatial variations. We will measure and compare distances between home, social, and health facility access locations and the extent to which there is neighborhood overlapping. We will quantify patterns and relationships describing spatial variation at both the smaller scale (spatial clustering) and larger scale (spatial trends), and examine the association of sociodemographic covariates such as race or ethnicity, socioeconomic status, and neighborhood exposure with spatial clustering of the HIV care outcomes. Results will provide important insights into how social and spatial environments impact the outcomes and help inform our other regression analyses by providing baseline information on trends and variability that will be incorporated in the covariates in the regression models, both at the individual and neighborhood levels.
Data Analysis 2: Baseline Associations Between Exposure to Neighborhoods and HIV Care OutcomesSecond, we will conduct baseline analyses to assess associations between exposures to neighborhoods of potential influence (home, social, and health facility access) separately for each of the HIV care outcomes using logistic regression models. Univariate logistic regressions will be computed for the full set of independent variables describing the neighborhood of potential influence to assess the effect and contribution of each variable separately. Variables significant at the P<.10 level will be considered for multivariate analysis and longitudinal analyses to examine HIV care outcomes by race or ethnicity among MSM living with HIV in relationship to longitudinal changes in (1) exposure to neighborhoods of potential influence (home, social, and health facility access); (2) neighborhood characteristics in four conceptual domains (community violence, physical disorder, or social disorganization; neighborhood AOD use-associated factors; neighborhood-level social norms; and community resources); and (3) social networks.
Data Analysis 3: Associations in Changes Over Time Exposure to Neighborhoods and Social Networks and HIV Care OutcomesWe will use multilevel modeling to assess associations among changes over time in exposure to the 4 neighborhood characteristic domains and changes in social networks with HIV care outcomes.
The study was initially funded in August 2017, with additional multiyear funding in April 2019. The study began enrollment in March 2019 and concluded visits in December 2023, with a total of 327 participants enrolled (213 in the pre–COVID-19 pandemic lockdown cohort and 114 in the post–COVID-19 pandemic lockdown cohort). The study paused enrollment from March 2020 to August 2021 as a result of the COVID-19 pandemic, with the resumption of enrollment in August 2021 to November 2022. The participants in the prelockdown cohort completed study visits up to 36 months, while those in the postlockdown cohort completed visits up to 12 months. Results are planned for publication in the spring of 2025.
Baseline sociodemographic characteristics for 319 participants with available data are described in , with 205 participants in the prelockdown cohort and 114 in the postlockdown cohort. The median age was 44.1 (SD 11.5) years. Almost all the participants self-identified as cisgender men (n=313, 98.1%) and as gay, homosexual, or bisexual (n=301, 94.4%). Overall, 192 (60.1%) participants identified as non-Hispanic Black, 81 (25.4%) as Hispanic, 32 (10.0%) as non-Hispanic White, and 14 (4.4%) as other. More than half had any college education (n=175, 54.8%), while 216 (67.7%) reported an annual household income of US $20,000 or less and 280 (87.8%) had Medicare, Medicaid, or New York State AIDS Drug Assistance Program (ADAP) insurance coverage. Most (n=201, 63%) reported at least occasional difficulty in meeting basic needs (eg, rent and food) in the past 6 months. The mean number of years living with HIV was 15.4 (SD 10.1).
Table 4. Baseline sociodemographic characteristics of NNHIVa study cohort.aNNHIV: Neighborhoods, Networks and HIV Care among Men Who Have Sex with Men.
bADAP: AIDS Drug Assistance Program.
Significant racial or ethnic disparities in HIV care–related outcomes among MSM in the United States motivate us to understand whether and how neighborhood and network characteristics act and interact to influence these outcomes. This study has several innovative features. First, neighborhoods and networks are rarely static and the HIV care continuum is a dynamic process in which individuals with HIV move on and off steps of the continuum. Given the limited data available on neighborhood and network dynamics among MSM living with HIV, this study’s longitudinal design examines changes in exposure to neighborhoods and social networks over time and determines how these influence HIV care outcomes, elucidating the potentially unique mechanisms by which they impact the HIV care continuum for MSM living with HIV by race or ethnicity. This design positions the study team to capture differential exposure to neighborhood and network characteristics and effects of national health policy changes, including changes to local, state, and federal health and social policy that may occur over time.
Second, we focus on identifying how individual-level factors mediate the impact of neighborhood-level factors, enabling us both to identify malleable characteristics that are amenable to traditional individual-level intervention and area-level factors, which, if modified, may have important positive effects on a range of health outcomes among MSM living with HIV. In addition, we investigate multiple neighborhood contexts for MSM living with HIV in the urban environment, rather than only their residential neighborhoods, which has been the focus (and a key limitation) of the majority of neighborhood research. We use innovative spatial analysis methods that will address limitations inherent in prior neighborhood-level research. We use Google Earth to identify the location and scale of multiple neighborhoods that are potentially important to the MSM. We collect neighborhood boundaries that are defined and characterized by the participants, which will reflect their own local, small-area definitions of the size of influential space in these contexts. Furthermore, the incorporation of subjective (perceptions) and objective measures of neighborhoods is innovative and has been not used in prior research examining the disparity in HIV care outcomes among MSM.
One area of focus of our analysis is HIV stigma, which has been associated with significant negative health and social outcomes among people with HIV [,]. HIV stigma influences the HIV care cascade by reducing access to testing [,] and medical treatment [,]. It also reduces the uptake of biomedical prevention technologies, critical tools to interrupt HIV transmission [-]. There is limited data on how community-level HIV stigma influences HIV care-related outcomes. In this study, we will examine the impact of individual-level intersectional stigma (internalized, experienced, and perceived community) at various levels on HIV care outcomes and will evaluate these relations over time.
A second area of focus is on stress and coping. According to stress theory, stressful life events give rise to coping responses, which can be adaptive or maladaptive. When a coping response is adaptive and effective, the impact of the stressor is mitigated; when the response either does not exist or is maladaptive, the stressor has a negative effect on the individual [,]. Neighborhood environmental conditions, such as concentrated poverty, racial segregation, physical disorder, and violence, increase the likelihood of experiencing stressors. Physical disorders (eg, vacant buildings) can lead to stressful experiences, such as crimes and assault, as they provide spaces and cover for illegal activity (eg, drug trade) [-], increasing general arousal or anxiety, and consequent negative immune function. There is literature on community violence and adverse health outcomes [-]. We will examine the impact of neighborhood-level concentrated poverty, racial segregation, and physical disorder on stressful life events and subsequent HIV care outcomes. We will examine whether they are mediated by individual-level psychological distress and moderated by coping mechanisms, with maladaptive ones (ie, AOD use) mediating adverse neighborhood effects and positive ones (ie, positive coping skills) buffering them.
A more nuanced understanding of the mechanisms by which neighborhood and network-level factors influence the HIV care continuum steps among MSM will guide the development of urban planning policies and multilevel, neighborhood and network-level interventions to augment individual and patient provider-level interventions to improve outcomes. Neighborhood and network factors are often examined separately in research studies, and this study aims to evaluate both factors using multilevel modeling. Specific interventions can be tailored to subpopulation groups focusing on malleable characteristics that specifically influence HIV care outcomes for that group. For example, high-poverty neighborhoods with numerous health and HIV-related resources that also have high levels of HIV stigma and homophobia may be associated with inconsistent care receipt inside of that neighborhood and, subsequently, low levels of ART adherence for MSM, especially among those with high internalized stigma and social networks that offer less support. In contrast, for care receipt outside the neighborhood, network and individual-level poverty may be more important to care outcomes than local resources or neighborhood stigma. Our planned analyses will identify sets of neighborhood and network-level characteristics that constitute the varying contexts where access to care and engagement are optimized. Results will also identify how individual-level factors interact with these contexts.
Such results could inform the design of care engagement interventions for MSM who receive HIV care inside and outside their neighborhoods. Determining if the association between neighborhood resources and HIV care outcomes is modified by transit access and the neighborhood type (eg, home vs social) could inform future siting of resources for urban MSM living with HIV in underserved areas. Identifying the effects of exposure duration to stressful neighborhoods, through analysis of MSM who move, and network-level influences could inform our understanding of where to site housing for MSM living with HIV and who might benefit from moving to a different neighborhood. Social network–based interventions can be tailored to promote HIV serostatus disclosure, ART adherence, and retention in care within social networks through peer leaders to leverage their influence over HIV care-associated social norms within networks, taking into account how neighborhood-level factors condition network influences. Most importantly, the level of detail and longitudinal nature of the study addresses concerns around one-size-fits-all approaches, even at the neighborhood and network levels, such as generic antistigma media campaigns and crime reduction efforts.
The COVID-19 pandemic caused by the SARS-CoV-2 virus that emerged in New York City in March 2020 was associated with high levels of morbidity and mortality, especially among those who are immunocompromised and have underlying comorbidities, including HIV []. Our study had to pause enrollment from March 2020 to August 2021, with the resumption of enrollment in August 2021 to November 2022. This will allow us to conduct additional analysis on the effect of the COVID-19 pandemic on the mental health, substance use, and general well-being of MSM living with HIV in New York City, as well as its effect on how neighborhood and social network factors influence HIV care outcomes pre- and post-pandemic.
ConclusionsIn summary, the study uses an innovative approach combining geospatial mapping of neighborhoods with multilevel modeling, spatial statistics, and an extensive geospatial database. It will have direct implications for the design of multilevel interventions, addressing factors at the neighborhood, network, and individual levels, to improve HIV care outcomes for MSM, particularly for Black and Latino MSM. This study will have direct implications for the design of neighborhood- and network-level interventions and urban planning to improve the HIV care outcomes for MSM, particularly for Black and Latino MSM.
The authors thank the Neighborhoods, Networks and HIV Care among Men Who Have Sex with Men (NNHIV) study team, which included Michael Bianco, John Bonelli, Paris Green, Leila Noori, Jorge Soler, and Gulseren Yilmaz for their contributions. The authors thank the NNHIV study participants for their time. This work was supported by the National Institutes of Health or National Institute of Mental Health (NIMH) grants (R56 MH112401-01A1 and R01 MH118960-01A1).
The datasets generated and analyzed during this study are available from the corresponding author on reasonable request.
HVT and VAF performed the conceptualization. HVT, VAF, MW, VN, AR, FC, CW, and MRD contributed to the data curation. HVT, VN, VAF, JD, EG, FC, CW, and MRD performed the formal analysis. HVT and VAF contributed to the funding acquisition. HVT, VAF, MW, MB, JB, PG, LN, JS, and GY did the investigation. HVT, VAF, VN, JD, EG, AR, FC, and CL contributed to the methodology. HVT, VAF, and MW contributed to the project administration. VN and FC helped with the software. HVT, VAF, VN, AR, and FC helped with the resources. HVT, VAF, and MW performed the supervision. VN, FC, CW, and MRD did the validation. VN, FC, CW, and MRD did the visualization. HVT and VAF contributed to writing—original draft. HVT, VAF, JD, EG, MW, FC, MRD, CW, and CL contributed to writing—review and editing.
None declared.
Edited by T Leung; The proposal for this study was peer-reviewed by: Population and Public Health Approaches to HIV/AIDS Study Section (PPAH) - AIDS and Related Research Integrated Review Group - Center for Scientific Review (National Institutes of Health, USA). See the Multimedia Appendix for the peer-review report; submitted 18.07.24; accepted 17.08.24; published 13.11.24.
©Hong Van Tieu, Vijay Nandi, José E Diaz, Emily Greene, Melonie Walcott, Frank Curriero, Michael R Desjardins, Cara Wychgram, Carl Latkin, Andrew G Rundle, Victoria A Frye, NNHIV Study Team. Originally published in JMIR Research Protocols (https://www.researchprotocols.org), 13.11.2024.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on https://www.researchprotocols.org, as well as this copyright and license information must be included.
留言 (0)