Population-Based Analysis of Synergistic Effects of Functional Disability and Comorbidities on Retention in HIV Care Among People With HIV

Abstract

Objective

This cross-sectional study examined how types of disability and comorbidities influence retention in care (RIC) among people with HIV (PWH) in Houston/Harris County, Texas, USA.

Methods

Data from 1142 participants in the Houston Medical Monitoring Project (2015–2021) were analyzed. Descriptive statistics, bivariate analyses, and multivariable logistic regression were used to evaluate factors associated with RIC.

Results

Approximately 74.9% of participants were retained in care. Compared to PWH without specific comorbidities, those with dyslipidemia and kidney disease had higher odds of RIC (AOR: 2.66; 95% CI: 1.40–5.03 and AOR: 14.08; 95% CI: 1.44–138.12, respectively). PWH reporting only mobility disability had increased odds of RIC (AOR: 3.40; 95% CI: 1.33–8.71), while those reporting only visual disability had reduced odds of RIC (AOR: 0.38; 95% CI: 0.16–0.87). Durable viral suppression was strongly associated with greater RIC (AOR: 7.96; 95% CI: 5.28–11.99).

Conclusion

Comorbidities and disability types significantly influence retention in HIV care. Tailored interventions are needed to improve RIC among PWH, particularly those with visual disabilities.

Plain Language Summary

Effects of Disability and Comorbidities on Staying in HIV Care Among People with HIV

Objective:

Staying in regular HIV care is key for people with HIV (PWH) to manage their health. This study looked at how having other health conditions (comorbidities) or disabilities affects whether PWH in Houston and Harris County, Texas, stay in care.

Methods:

We used data from 1142 participants in the Houston Medical Monitoring Project collected between 2015 and 2021 cycles for this study.

Results:

We found that about three out of every four people (74.9%) stayed in regular HIV care. People with certain health conditions, such as high cholesterol or kidney disease, were more likely to stay in care than those without these conditions. People who had difficulty walking or moving around were also more likely to remain in care, while people with visual disabilities were less likely to stay in care. Keeping the virus under control was also strongly linked to staying in care.

Conclusion:

The results show that specific approaches are needed to support PWH who have vision difficulties to enable them to stay in care and maintain good health outcomes.

Keywords

HIV people with HIV comorbidities disability retention in care multimorbidity engagement in care

Introduction

Retention in Care (RIC) is an essential component of the HIV care continuum (HCC) that is vital for optimal clinical outcomes in people with HIV (PWH) because it ensures consistent receipt of antiretroviral therapy (ART), durable viral suppression, and subsequent prevention of Human Immunodeficiency Virus (HIV) transmission.^1,2 Recent research buttresses the positive contribution of RIC to durable viral suppression among both newly diagnosed and long-term patients with HIV, reaffirming its importance even in the modern ART era.³ Conversely, suboptimal RIC is associated with mortality and may increase the transmission of HIV to others secondary to unsuppressed viral loads.^1,4 Therefore, strengthening RIC is vital for improving individual health outcomes of PWH and curbing the spread of HIV. However, nationally, only 54% were retained in HIV care in 2022.⁵ This low RIC rate hinders achieving the goal of ending the HIV epidemic by 2030,⁶ underscoring the urgent need for targeted interventions to enhance care engagement among affected populations.

Due to the success of ART, the life expectancies of PWH have been tremendously extended.⁷ However, this longer life span is associated with challenges such as the acquisition of comorbidities, adverse effects of ART, and the natural effects of aging, leading to an increased likelihood of disability.^8–11 For instance, the multimorbidity burden among PWH is projected to rise from 63% in 2020 to 70% by 2030,¹² while another study showed evidence of neurocognitive impairment contributing to greater disability among PWH.¹³ In the United States, it is estimated that approximately four in ten PWH experience some form of disability.¹⁴ Notably, PWH with disabilities experience twice the risk of death compared to those without disabilities.¹⁵ Furthermore, disability can negatively affect an individual's overall health, quality of life, and ability to participate fully in society, compounding the challenges of living with HIV.^16–18

Given these problems, disability may interfere with RIC, potentially leading to poor clinical outcomes. This is especially concerning in the context of both the growing population of older PWH and the increasing burden of chronic comorbidities associated with disability.^19,20 Several individual and structural factors have been shown to negatively influence RIC, including younger age,²¹ being female,²² identification as Black,²³ poverty,²⁴ and lack of health insurance coverage.²⁵ While these factors are well-documented, there remains limited research exploring the role of specific disability types on RIC among PWH. While a geospatial analysis found that PWH residing in areas with lower disability rates were more likely to be retained in care,²⁶ the study did not assess how specific types of disability, such as cognitive, mobility, or visual impairments, may differentially impact RIC. This gap highlights the need for more nuanced, individual-level analyses to identify how various disabilities influence RIC among PWH.

In this study, we utilized data from an in-depth survey of a large, representative sample of PWH in Houston/Harris County, Texas, to examine the associations between disability status, specific types of disability, and comorbidity burden on RIC. Findings from this study will inform disability-inclusive public health strategies, promote equitable access to care, and support progress toward national goals for ending the HIV epidemic.

Methods

Data Source and Study Design

We conducted a cross-sectional study using survey data from seven consecutive cycles (2015–2021) of the Houston Medical Monitoring Project (HMMP). The Medical Monitoring Project (MMP) is an ongoing population-based national HIV surveillance system designed to produce representative estimates of behavioral and clinical characteristics of people diagnosed with HIV in the United States and Puerto Rico.²⁷ The standardized MMP questionnaires used for data collection were validated instruments developed by the Centers for Disease Control and Prevention (CDC), and undergo routine testing and refinement to ensure reliability, content validity, and consistent data collection across project areas.²⁷ Details about the MMP survey population, sampling process, and questionnaire can be found elsewhere.^27,28 The MMP utilized a two-stage sampling strategy to select a representative sample from the National HIV Surveillance System (NHSS), a census of all people with diagnosed HIV in the United States, and retrospectively linked their 2-year medical record abstraction (MRA).

In the first stage, a geographically stratified random sample of U.S. states and territories was selected in 2004 with probability proportional to the number of persons living with AIDS (based on 2002 NHSS data). Texas was among the selected states, and due to its size and surveillance structure, Houston/Harris County is separately funded from the rest of Texas, making it one of the 23 MMP project areas. In the second stage, an annual random sample of eligible individuals residing in Houston/Harris County was drawn from the National HIV Surveillance System (NHSS). The HMMP data was collected using telephone or in-person interviews and MRAs. The data contains detailed sociodemographic and behavioral characteristics, and clinical information, including diagnoses, laboratory results, and medications.

Inclusion and Exclusion Criteria

Participants were eligible for inclusion in this study if they were adults aged 18 years or older, diagnosed with HIV, alive, and residing in Houston/Harris County, Texas, as of December 31 of the year preceding the data collection cycle. All interview respondents provided informed consent and agreed to both the interview and MRA. Individuals previously sampled in an earlier MMP cycle were still eligible to be re-sampled and included in the study.

Individuals were excluded if they were younger than 18 years, deceased prior to the sampling date, or deceased on or after the sampling date but prior to recruitment, not residing in the project jurisdiction, or unable to provide informed consent. Participants were also excluded if based on interview and MRA, the status of RIC could not be determined.

Each study cycle occurs from June through May of the following year. Therefore, the study period analyzed spanned from June 2015 to May 2022. Between the 2015 and 2021 cycles, we analyzed the data of 1142 PWH, representing 24,818 PWH in Houston/Harris County, Texas, who participated in the HMMP. Data were weighed based on the known probability of selection at the city and person levels, ensuring population-level representativeness. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines²⁹ (see supplementary file 1).

Outcome variable

Retention in HIV care (RIC) was defined as having received at least two elements of outpatient HIV care at least 90 days apart during the past 12 months.³⁰ The receipt of outpatient care was determined using already abstracted medical records and identified as any documentation of the following in the past 12 months: visit with an HIV care provider, viral load test result, CD4 test result, HIV resistance test or tropism assay, ART prescription, Pneumocystis Jirovecii Pneumonia (PCP) prophylaxis, or Mycobacterium Avium Complex (MAC) prophylaxis.³⁰ The receipt of outpatient care was categorized as “yes” and “no”, indicating retention and non-retention in care, based on whether individuals met the criteria for consistent HIV care engagement.

Primary Exposure Variables

The primary exposure variables were disability status, types of disability, and a composite of comorbidity and disability. The HMMP assessed disability using six questions related to hearing, vision, cognition, mobility, self-care, and independent living. Participants were asked (1) Are you deaf or have serious difficulty hearing? (hearing disability); (2) Are you blind, or do you have serious difficulty seeing, even when wearing glasses? (vision disability); (3) Because of a physical, mental, or emotional condition, do you have serious difficulty concentrating, remembering, or making decisions? (cognitive disability); (4) Do you have serious difficulty walking or climbing stairs? (mobility disability); (5) Do you have difficulty dressing or bathing? (self-care disability); and (6) Because of physical, mental, or emotional condition, do you have difficulty doing errands alone, such as visiting a doctor's office or shopping? (independent living disability). Participants who answered “Yes” to at least one of the disability questions were classified as having a disability, and those who answered “No” to all the questions were classified as not having a disability (disability status). Next, participants were categorized into seven groups by disability types³¹: (1) no functional disability (reference group, if no to all six questions), (2) hearing limitation only (if yes to Question 1, and no to the other 5), (3) vision limitation only (if yes to Question 2, and no to the other 5), (4) cognitive limitation only (if yes to Question 3, and no to the other 5), (5) mobility limitation only (if yes to Question 4, and no to the other 5), (6) complex activity limitation only (if yes to either Question 5 or 6, and no to the other 4), and (7) two or more limitations (if yes to ≥2 questions).

In addition, comorbidity and disability status were examined as a composite exposure variable to assess the joint effect of comorbid conditions and disabilities on retention in care. This four-level categorical variable combined information on comorbidity and disability, classifying participants as: (1) no comorbidity and no disability (reference group); (2) both comorbidity and disability; (3) only disability; and (4) only comorbidity.

Covariates

The covariates include sociodemographic, behavioral, psychosocial, and clinical factors. Sociodemographic characteristics include age group (18–29, 30–39, 40–49, and ≥50 years), gender (male, female, transgender), sexual orientation (heterosexual, lesbian or gay, bisexual), education (< high school, high school diploma, >high school), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic and other/unknown), employment status (employed, not employed), poverty level (above or below), foreign-born status (yes, no), health literacy (extremely, quite a bit, somewhat, a little bit/ not at all), food insecurity (yes, no), homelessness (yes, no), and health insurance status (yes, no).

Behavioral characteristics include smoking status (never, former, and current), injection drug use in the last 12 months (yes, no), alcohol use (yes, no), heavy drinking (yes, no), and electronic cigarettes use (yes, no).

Psychosocial characteristics include anxiety (no, mild, moderate, severe), depression (no, other, major), stigma (at or above average, below average), experience of discrimination in health care setting (yes, no), experience of physical violence, sexual violence (yes, no), trust in provider (at or above average, below average), unmet need for mental health services (yes, no), unmet need for case management (yes, no), and unmet need for substance use treatment (yes, no).

Clinical characteristics were assessed from the last 12 months and included years since HIV diagnosis, ART adherence, advanced HIV disease, durable viral suppression, lowest CD4 count, number of hospital admissions, and number of emergency department visits. HIV stage 3 disease at diagnosis was also included. ART adherence was measured using a self-reported item: ‘In the past 30 days, on how many days did you miss at least one dose of any of your HIV medicines?’ Participants who reported missing two or more days were classified as nonadherent based on the commonly accepted 95% adherence threshold required for ART effectiveness.³² Durable viral suppression was defined as all viral loads undetectable or less than 200 copies/ml. The lowest CD4 count variable was classified as CD4 count ≥200 cells/mm³ and CD4 count <200 cells/mm³, with the latter threshold reflecting the clinical definition of an AIDS-defining illness in PWH.³³

The comorbidities considered in this study were substance use disorder, myocardial infarction, congestive heart failure, peripheral vascular disease, cancer, dyslipidemia, diabetes mellitus, liver diseases, obesity, kidney disease, chronic pain, chronic respiratory disease, chronic gastrointestinal disease, hypertension, and cerebrovascular disease. These comorbidities were determined through medical record abstraction using their ICD-10 code classifications. For each participant, the number of comorbidities was categorized as 0, 1, 2–3, or 4 or more.³⁴

Statistical Analysis

Descriptive statistics were performed to summarize the characteristics of the study population. Categorical variables were presented as frequencies and percentages. Bivariate analyses were conducted to examine the associations between the covariates and RIC using the Rao-Scott Chi-square test. Based on the results of the bivariate analyses, variables with a p value < 0.10 were simultaneously included in the multivariable logistic regression models. The main exposure variable, “types of disability”, was forced into the model. Three multivariable logistic regression models were developed to control for confounders, each specifying a different primary exposure: (1) overall disability status, (2) specific disability types, and (3) a combined comorbidity-disability variable. Both unadjusted and adjusted multivariable logistic regression models were used to assess the association between the selected covariates and RIC status, with odds ratios, 95% confidence intervals (CIs), and p values reported.

Missing data were present for some variables (see supplementary file 1), and we used the available case analysis approach. All analyses applied sampling weights to account for the design weight, non-response, and post-stratification, ensuring that the findings are locally representative. To protect participants’ confidentiality and in accordance with standard public health reporting practices, cell sizes with frequencies of five or fewer (n < 5) are suppressed in tables. When necessary, secondary suppression was applied to prevent the derivation of suppressed values through arithmetic inference. These procedures are consistent with confidentiality guidelines established by the CDC and HRSA. Zero values (n = 0) were not subject to suppression.

All statistical tests were two-tailed, with a <0.05 probability level used as the threshold for declaring statistical significance. Data management and statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Human Subjects Protections

The MMP has been determined by the Centers for Disease Control and Prevention to be a non-research public health surveillance activity used for disease control programs or policy purposes. As such, MMP is not subject to human subjects’ regulations, including federal institutional review board approval. All data collection at the primary stage was Health Insurance Portability and Accountability Act compliant, and informed consent was obtained from all participants after the nature of the survey and their rights were fully explained to them. Permission to use the Houston MMP data was granted by the Houston Health Department's Investigative Review Committee. However, in accordance with institutional requirements, the current study protocol was submitted to the Committee for the Protection of Human Subjects at the University of Texas Health Science Center at Houston for review, where it received Exempt Status approval (IRB reference # 259145). All relevant ethical principles and guidelines were followed in the conduct of this study, ensuring the protection of participants’ rights and confidentiality throughout the study.

Results

Sociodemographic, Behavioral, Psychosocial, and Clinical Characteristics of Participants

Among the 1142 eligible participants representing an estimated 24,818 PWH in Houston/Harris County, Texas, 74.9% were retained in care during the preceding 12 months.

Table 1 presents the sociodemographic, behavioral, psychosocial, and clinical characteristics of the study population. Most of the participants were males (73.9%), over 50 years of age (42.1%), and identified as Black non-Hispanic (52.2%). More than half of the participants had attained education beyond high school (54.3%), lived above the poverty level (58.3%), and were currently employed (50.9%). Nearly all (96.6%) had health insurance, 89.4% had stable housing, and 77.5% were food secure. Regarding substance use, the majority drank alcohol in the past 12 months (60.9%), with 4.6% classified as heavy drinkers in the past 30 days. Most of the participants (97.7%) did not use any illicit injection drugs; however, 12.9% were former smokers, 30.9% were current smokers, and 21.1% of the participants had used electronic cigarettes in the past 12 months.

Table 1.

Sociodemographic, behavioral, psychosocial, and clinical characteristics among people with HIV—Houston Medical Monitoring Project, 2015–2021.

Characteristics	Frequency (n = 1142^a)	Weighted % (95%CI^b)
Sociodemographic
Age-group (years)18–2930–3940–49≥50	137248281476	12.58 (10.44–14.73)21.56 (18.94–24.18)23.79 (21.09–26.49)42.07 (38.89–45.24)
Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic or Latino Other	17862230042	17.33 (14.80–19.85)52.20 (48.99–55.41)26.11 (23.33–28.89)4.36 (2.89–5.83)
Gender Male Female Transgender	78433915	73.94 (71.26–76.62)24.72 (22.09–27.35)1.33 (0.65–2.02)
Education <High school High school diploma >High school	217324598	17.78 (15.38–20.17)27.97 (25.12–30.83)54.25 (51.07–57.43)
Sexual orientation Lesbian or gay Heterosexual Bisexual Other	44654410631	40.87 (37.69–44.04)45.29 (42.09–48.49)10.98 (8.78–13.19)2.85 (1.77–3.92)
Foreign born No, not born in a foreign country Yes, born in a foreign country	915223	81.31 (78.92–83.69)18.69 (16.30–21.08)
Poverty level Above poverty level At or below poverty Level	587443	58.37 (55.07–61.66)41.63 (38.34–44.93)
Employment status ^c Employed Unemployed	574566	50.93(47.73–54.12)49.07 (45.88–52.27)
Health insurance No Yes	351096	3.38 (2.19–4.55)96.62 (95.45–97.80)
Food insecurity No Yes	879260	77.48 (74.85–80.11)22.52 (19.89–22.15)
Homelessness No, not homeless Yes, was homeless	1024116	89.35 (87.31–91.37)10.65 (8.63–12.68
Health literacy Extremely Quite a bit Somewhat A little bit/Not at all	642211146140	55.84 (52.64–59.04)19.56 (16.91–22.22)12.63 (10.53–14.74)11.96 (8.89–14.03)
Behavioral
Smoking status ^d Never smoker Former smoker Current smoker	650144344	56.14 (52.95–59.34)12.99 (10.82–15.16)30.87 (27.89–33.85)
Alcohol use ^d No Yes	469669	39.00 (35.93–42.08)60.99 (57.92–64.07)
Heavy drinking No Yes	107656	95.34 (93.98–96.69)4.66 (3.31–6.02)
Injection drug use^d No Yes	111323	97.69 (96.70–98.69)2.30 (1.31–3.29)
Electronic cigarettes use^d No Yes	773195	78.88 (76.00–81.76)21.12 (18.24–23.99)
Psychosocial
Depression No depression Other depression Major depression	907112108	80.93 (78.42–83.44)9.61 (7.76–11.46)9.46 (7.56–11.36)
Anxiety No anxiety Mild anxiety Moderate anxiety Severe anxiety	906646596	79.54 (76.91–82.16)5.90 (4.35–7.45)5.93 (4.40–7.46)8.63 (6.79–10.47)
HIV stigma Below average stigma At or above average stigma	350269	57.48 (53.26–61.69)42.52 (38.31–46.74)
Discrimination No discrimination Yes, had discrimination	489140	77.27 (73.71–80.83)22.73 (19.17–26.29)
Level of trust in provider Below average trust in provider At or above average trust in provider	149390	27.65 (23.58–31.72)72.35 (68.28–76.42)
Unmet need for mental health services No Yes	104594	91.75 (90.00–93.49)8.25 (6.50–9.99)
Unmet need for case management No Yes	1002134	87.87 (85.72–90.01)12.13 (9.99–14.28)
Unmet need for substance use treatment No Yes	112216	98.49 (97.67–99.31)1.51 (0.69–2.33)
Experienced physical violence No Yes	869261	77.42 (74.76–80.08)22.58 (19.92–25.24)
Experienced sexual violence No Yes	833127	87.51 (85.27–89.76)12.49 (10.24–14.73)
Clinical characteristics
Years since HIV diagnosis <5 5–9 >10	250273619	20.90 (18.39–23.41)24.85 (22.01–27.68)54.25 (51.07–57.44)
HIV stage 3 disease No Yes	530612	47.11 (43.91–50.80)52.89 (49.69–56.09)
Advanced HIV disease^d No Yes	947193	84.65 (82.45–86.86)15.35 (13.14–17.55)
Number of hospital admissions01≥2	94811078	84.33 (82.07–86.59)9.12 (7.33–10.91)6.55 (5.04–8.07)
Number of visits to the ED^e01≥2	742204191	66.78 (63.79–69.77)17.02 (14.66–19.38)16.19 (13.85–18.54)
ART adherence ART adherent ART non-adherent	731315	69.87 (66.79–72.95)30.13 (27.05–33.21)
Retention in care^d No, not retained in care Yes, Retained in care	246896	25.14 (22.18–28.09)74.86 (71.90–77.82)
Durable viral suppression (copies/ml)^dAll viral loads undetectable, <200Any viral load ≥200	679463	56.16 (52.95–59.38)43.84 (40.62–47.05)
Lowest CD4 count (cells/mm³)^d<200≥200	128780	13.06 (10.84–15.29)86.94 (84.71–89.16)
Number of comorbidities012–3≥4	45228830399	42.17 (38.97–45.38)24.69 (21.96–27.43)24.97 (22.29–27.64)8.17 (6.46–9.87)
Substance use disorder No Yes	1036106	91.16 (89.38–92.94) 8.84 (7.05–10.62)
Myocardial infarction No Yes	112517	98.23 (97.32–99.13)1.77 (0.85–2.68)
Congestive heart failure No Yes	11348	99.35 (98.87–99.83) 0.65 (0.17–1.13)
Peripheral vascular disease No Yes	1138<5^f	99.74 (99.49–99.99)0.26 (0.00–0.51)
Cancer No Yes	109151	95.60 (94.33–96.88)4.39 (3.12–5.67)
Dyslipidemia No Yes	849293	75.34 (72.67–78.02)24.65 (21.98–27.33)
Diabetes mellitus No Yes	1027115	90.84 (89.14–92.54)9.16 (7.46–10.86)
Moderate/severe liver disease No Yes	112517	98.24 (97.27–99.21)1.76 (0.79–2.73)
Obesity No Yes	1030112	91.30 (89.63–92.97)8.69 (7.03–10.37)
Kidney disease No Yes	107864	94.69 (93.35–96.03)5.31 (3.97–6.65)
Mild liver disease No Yes	105587	91.81 (89.92-93.71)8.19 (6.29-10.08)
Chronic pain No Yes	1040102	91.64 (89.93–93.36)8.36 (6.64–10.07)
Chronic respiratory disease No Yes	109646	96.51 (95.47–97.55)3.49 (2.45–4.53)
Chronic gastrointestinal disease No Yes	105983	93.44 (91.96–94.91)6.56 (5.09–8.04)
Hypertension No Yes	861281	77.33 (74.78–79.88)22.67 (20.12–25.22)
Cerebrovascular disease No Yes	11366	99.51 (99.09–99.93)0.49 (0.07–0.91)

Due to missing responses, the frequencies reported for individual characteristics may not add up to the overall sample total.

CI: confidence interval.

Unemployment status included individuals that were out of work for more than 1 year, out of work for less than 1 year, homemakers, students, retirees and those unable to work

Assessed in the past 12 months.

ED: emergency department.

Notable mental difficulties were reported by a few of the participants, with 9.5% experiencing major depression and 8.6% reporting severe anxiety. Regarding unmet service needs, 8.3% lacked access to mental health services, while 12.1% and 1.5% had unmet needs for case management and substance use treatment, respectively. Experiences of violence were notable, with 22.6% of participants reporting physical violence and 12.5% experiencing sexual violence. Among the participants, 57.5% had stigma scores below the average, while the majority (77.3%) reported no experience of discrimination. Additionally, 72.3% expressed above-average trust in their healthcare providers.

Clinically, more than half (54.3%) had been living with HIV for over a decade, and almost 85% of PWH showed no evidence of advanced HIV disease in the past 12 months. The majority (86.9%) had the lowest CD4 count of at least 200 cells/mL, and 56.2% maintained undetectable viral loads (<200 cells/mL) across all tests conducted during the same period. Adherence to ART was observed in 69.9% of PWH in the year preceding the survey.

Regarding comorbidities, 24.7% of the participants had two or three comorbidities, while 8.2% reported at least 4 comorbidities. The common comorbidities of the participants were dyslipidemia (24.7%), hypertension (22.7%), diabetes mellitus (9.2%), and substance use disorders (8.8%).

Prevalence and Patterns of Disability

Table 2 presents the distribution of disability among study participants. Mobility disability was the most reported type (19.9%), followed closely by cognitive disability (19.5%). Other reported disabilities included vision disability (11.5%), independent living disability (10.8%), hearing disability (9.3%), and self-care disability (7.2%). Overall, 39.2% of participants reported experiencing at least one form of disability. When analyzing the distribution of disability types, 20.7% of participants reported having two or more disabilities, while 6.8% reported only cognitive disabilities, 5.5% specified only mobility disabilities, 2.6% had only visual disabilities, and 2.4% reported only hearing disabilities. A smaller proportion (1.2%) experienced complex activity limitations (either independent living or self-care disability only). Among participants, 27.9% had neither comorbidity nor disability, while 24.0% had both comorbidity and disability. Approximately 14.9% reported disability only, and 33.2% had comorbidity without disability.

Table 2.

Prevalence of disability among people with HIV—Houston Medical Monitoring Project, 2015–2021.

Characteristics	Frequency ^a	Weighted% (95% CI) ^b
Hearing disability No Yes	1036104	90.74 (88.84–92.64)9.26 (7.36–11.16)
Visual disability No Yes	998141	88.46 (86.53–90.40)11.54 (9.59–13.47)
Cognitive disability No Yes	906232	80.51 (78.03–82.99)19.49 (17.01–21.97)
Mobility disability No Yes	897242	80.09 (77.58–82.59)19.91 (17.40–22.42)
Self-care disability No Yes	105485	92.79 (91.11–94.47)7.21 (5.53–8.89)
Independent living disability No Yes	1005132	89.23 (87.31–91.44)10.77 (8.86–12.69)
Any disability No Yes	678458	60.80 (57.69–63.92)39.19 (36.08–42.31)
Disability types No disability Hearing disability only Vision disability only Cognitive disability only Mobility disability only Complex activity Limitations >2 disabilities	6782433746713247	60.80 (57.69–63.92)2.39 (1.3–3.45)2.62 (1.70–3.54)6.77 (5.13–8.42)5.49 (4.12–6.86)1.24 (0.51–1.97)20.69 (18.13–23.24)
Comorbidity and disability status No comorbidity and no disability Both comorbidity and disability Only disability Only comorbidity	296300158386	27.93 (24.92–30.95)23.97 (21.35–26.59)14.85 (12.48–17.23)33.24 (30.24–36.24)

Due to missing responses, the frequencies reported for individual characteristics may not add up to the overall sample total.

CI: confidence interval.

Factors Associated with Retention in HIV Care

Table 3 displays the bivariate associations between sociodemographic, behavioral, psychosocial, and clinical characteristics and RIC status. RIC status was significantly associated with age, race, gender, foreign-born status, health insurance status, and unmet need for case management and substance use treatment (p < 0.05). A significantly (p < 0.001) higher proportion of participants aged 50 years and older were retained in care (46.3%) compared to those not retained in care (29.5%). In contrast, a significantly (p = 0.006) higher proportion of Black, non-Hispanic individuals were not retained in care (62.5%) compared to those who were retained (48.7%), indicating a potential association between Black race and lower retention in HIV care. Gender disparities in RIC were statistically significant (p = 0.025), as a higher proportion of females were not retained in care (31.3%) compared to those who were retained (22.5%). Additionally, a significantly higher proportion of U.S.-born individuals were not retained in care (86.8%) compared to those who were retained (79.5%) (p = 0.016), indicating that U.S. nativity was associated with lower HIV care retention.

Table 3.

Association between sociodemographic, behavioral, psychosocial, and clinical characteristics and retention in care among people with HIV—Houston Medical Monitoring Project, 2015–2021.

	Retained in care		Not retained in care		Test statistics
Characteristics	n^a	Weighted % (95% CI^b)	n^a	Weighted % (95% CI^b)	$χ^{2} v a l u e$ ^c	p value^d
Overall	896	74.86 (71.90–77.82)	246	25.13 (22.18–28.09)	204.61	<0.0001
Sociodemographic
Age-group (years)18–2930–3940–49≥50	97183211405	11.59 (9.26–13.93)19.69 (16.92–22.47)22.43 (19.55–25.31)46.28 (42.70–49.86)	40 65 70 71	15.55 (10.61–20.48)27.11 (20.87–33.35)27.84 (21.45–34.22)29.50 (23.12–35.89)	18.61	0.0003***
Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic or Latino Other	15146225033	19.41 (16.38–22.44)48.73 (45.16–52.30)27.49 (24.36–30.61)4.37 (2.72–6.02)	27160509	11.19 (6.90–15.34)62.54 (55.66–69.42)22.02 (15.99–28.04)4.32 (1.16–7.48)	12.61	0.0056**
Gender Male Female Transgender	62425612	76.08 (73.28–78.88)22.51 (19.79–25.23)1.41 (0.61–2.22)	16083<5^e	67.61 (61.12–74.10)31.29 (24.85–37.72)1.10 (0.00–2.41)	7.42	0.0245*
Education <High school High school diploma >High school	180258455	18.16 (15.55–20.76)29.21 (25.95–32.48)52.63 (49.07–56.19)	3766143	16.64 (11.10–22.19)24.29 (18.49–30.09)59.06 (52.16–65.97)	2.68	0.2616
Sexual orientation Lesbian or gay Heterosexual Bisexual Other	3554218423	42.85 (39.27–46.33)44.00 (40.45–47.56)10.49 (8.15–12.84)2.65 (1.55–3.75)	91123228	35.01 (28.45–41.57)49.12 (42.08–56.17)12.45 (7.15–17.75)3.42 (0.69–6.16)	3.75	0.2903
Foreign born No, not born in foreign country Yes, born in a foreign country	703190	79.46 (76.68–82.24)20.54 (17.76–23.32)	21233	86.81 (82.19–91.43)13.19 (8.57–17.80)	5.78	0.0162*
Homelessness No, not homeless Yes, was homeless	80391	88.98 (86.61–91.36)11.02 (8.4–13.39)	22125	90.43 (86.49–94.36)9.57 (5.64–13.51)	0.3540	0.5519
Poverty Level Above poverty level At or below poverty level	451358	57.91 (54.25–61.56)42.09 (38.44–45.75)	13685	59.75 (52.38–67.12)40.25 (32.88–47.62)	0.1913	0.6618
Employment status ^f Unemployed Employed	457437	50.23 (46.67–53.80)49.77 (46.19–53.33)	109137	45.63 (38.62–52.64)54.37 (47.36–61.38)	1.3047	0.2534
Health insurance No Yes	17870	1.92 (0.98–2.87)98.08 (97.13–99.02)	18226	7.68 (4.00–11.36)92.32 (88.64–96.00)	18.10	<0.0001****
Food insecurity No Yes	696197	78.72 (75.88–81.56)21.28 (18.44–24.12)	18363	73.81 (67.58–80.03)26.19 (19.97–32.42)	2.15	0.1427
Health literacy Extremely Quite a bit Somewhat A little bit/not at all	489174121109	54.12 (50.54–57.70)21.05 (17.94–24.16)13.29 (10.93–15.65)11.54 (9.34–13.73)	153372531	60.92 (54.00–67.83)15.16 (10.19–20.12)10.69 (6.18–15.21)13.23 (8.26–18.19)	5.06	0.1678
Behavioral
Smoking status Never smoker Former smoker Current smoker	505120269	56.06 (52.14–59.61)13.74 (11.26–19.22)30.20 (26.94–33.46)	1452475	56.39 (49.33–63.44)10.75 (6.29–15.21)32.86 (26.10–39.62)	1.3719	0.5036
Alcohol use ^g No Yes	379515	39.96 (36.52–43.39)60.04 (56.61–63.48)	90154	36.13 (29.43–42.83)63.87 (57.17–70.57)	0.9683	0.3251
Heavy drinking No Yes	85536	96.07 (94.65–97.49)3.93 (2.51–5.35)	22120	93.09 (89.74–96.44)6.91 (3.56–10.26)	3.40	0.0651
Injection drug use ^g No Yes	87219	97.59 (96.47–98.73)2.41 (1.27–3.53)	241<5^e	97.99 (97.83–99.73)2.01 (0.00–4.05)	0.10	0.7504
Electronic cigarettes use ^g No Yes	618148	80.21 (77.12–83.31)19.79 (16.69–22.88)	15547	74.90 (68.13–81.67)25.09 (18.33–31.86)	2.16	0.1412
Psychosocial
Depression No depression Other depression Major depression	7109186	80.71 (77.95–83.48)9.74 (7.73–11.75)9.55 (7.43–11.67)	1972122	81.57 (75.92–87.22)9.24 (4.95–13.52)9.19 (5.03–13.36)	0.07	0.9652
Anxiety No anxiety Mild anxiety Moderate anxiety Severe anxiety	711555172	79.64 (76.74–82.54)6.41 (4.61–8.20)5.82 (4.18–7.47)8.12 (6.14–10.11)	19591424	79.22 (73.33–85.11)4.37 (1.35–7.39)6.26 (2.63–9.89)10.15 (5.82–14.48)	1.75	0.6269
HIV Stigma Below average stigma At or above average stigma	271202	58.28 (53.55–63.00)41.72 (36.99–46.45)	7967	55.36 (46.38–64.35)44.64 (35.65–53.62)	0.32	0.5721
Discrimination No discrimination Yes, had discrimination	395111	77.31 (73.35–81.26)22.69 (18.73–26.65)	9429	77.14 (69.01–85.26)22.86 (14.74–30.99)	0.0014	0.9705
Level of trust in provider Below average trust in provider At or above average trust in provider	119315	27.42 (22.97–31.87)72.58 (68.13–77.03)	3075	28.45 (18.82–38.08)71.55 (61.92–81.18)	0.04	0.8471
Unmet need for mental health services No Yes	82569	92.53 (90.72–94.34)7.47 (5.66–9.28)	22025	89.42 (85.05–93.79)10.58 (6.20–14.95)	2.01	0.1559
Unmet need for case management No Yes	80390	90.88 (88.98–92.79)9.11 (7.21–11.02)	19944	78.75 (72.63–84.88)21.25 (15.12–27.37)	21.89	<0.0001****
Unmet need for substance use treatment No Yes	88310	99.03 (98.40–99.66)0.97 (0.34–1.60)	2396	96.87 (98.40–99.66)3.12 (0.46–5.78)	5.11	0.0238*
Physical violence No Yes	683204	77.58 (74.63–80.53)22.42 (19.47–25.37)	18657	76.24 (71.01–82.86)23.06 (17.14–28.99)	0.0367	0.8482
Sexual violence No Yes	66594	88.59 (86.25–90.63)11.41 (9.07–13.75)	16833	84.30 (78.71–89.88)15.70 (10.12–21.29)	2.28	0.1308
Clinical
Years since HIV diagnosis (years) <55–9>10	200208488	21.64 (18.78–24.50)23.60 (20.51–26.69)54.76 (51.21–58.30)	5065131	23.90 (13.48–23.90)28.55 (22.07–35.03)52.76 (45.75–59.77)	2.34	0.3106
HIV Stage 3 disease No Yes	398498	44.69 (41.15–48.23)55.31 (51.77–58.85)	132114	54.32 (47.33–61.31)45.68 (38.69–52.67)	5.81	0.0160*
Advanced HIV disease^g No Yes	732164	82.71 (80.06–85.38)17.28 (14.62–19.94)	21529	90.50 (86.98–94.02)9.49 (5.98–13.01)	9.42	0.0021**
Number of hospital admissions^g01≥2	7428564	84.27 (81.75–86.78)8.94 (6.98–10.89)6.79 (5.06–8.52)	2062515	84.50 (76.56–89.45)9.65 (5.56–13.74)5.85 (2.75–8.95)	0.31	0.8533
Number of visits to ED^{g, h}01≥2	589154149	67.51 (64.19–70.83)16.76 (14.09–19.41)15.73 (13.19–18.27)	1535042	64.61 (57.97–71.25)17.81 (12.74–22.89)17.57 (12.12–23.03)	0.6427	0.7252
Adherence to ART ART adherent ART non-adherent	605251	70.59 (67.25–73.95)29.40 (26.05–32.75)	12664	67.08 (59.55–74.61)32.92 (25.39–40.44)	0.73	0.3932
Durable viral suppression (copies/ml)^g All viral loads undetectable or <200Any viral load ≥200	625271	68.66 (65.28–72.03)31.34 (27.97–34.72)	54192	18.96 (13.97–23.95)81.04 (76.05–86.03)	189.31	<0.0001****
Lowest CD4 (cells/mm³)^g <200≥200	111704	12.42 (10.14–14.69)87.58 (85.30–89.86)	1776	18.61 (10.27–26.95)81.39 (73.05–89.73)	2.58	0.1082
Disability
Hearing disability No Yes	80490	89.78 (87.59–91.96)10.22 (8.04–12.41)	23214	93.59 (89.78–97.41)6.41 (2.59–10.22)	2.19	0.1381
Vision disability No Yes	776117	87.43 (85.12–89.74)12.57 (10.26–14.88)	22224	91.52 (88.11–94.93)8.48 (5.07–11.89)	3.19	0.0742
Cognitive disability No Yes	702191	79.09 (76.19–81.99)20.91 (17.90–20.91)	20441	84.73 (79.97–89.50)15.26 (10.49–20.03)	3.43	0.0641
Mobility disability No Yes	690203	78.66 (75.78–81.52)21.34 (18.47–24.22)	20739	84.35 (79.26–89.43)15.65 (10.57–20.74)	3.16	0.0756
Self-care disability No Yes	82271	92.42 (90.53–94.32)7.58 (5.68–9.47)	23214	93.89 (90.33–97.45)6.11 (2.55–9.67)	0.45	0.4484
Independent living disability No Yes	788103	89.45 (87.34–91.56)10.55 (8.44–12.65)	21729	88.57 (84.21–92.92)11.43 (7.08–15.79)	0.13	0.7149
Any disability No Yes	512379	58.69 (55.18–62.20)41.31 (37.79–44.81)	16679	67.07 (60.45–73.68)32.93 (28.32–39.55)	4.53	0.0333*
Disability types No disability Hearing disability only Vision disability only Cognitive disability only Mobility disability only Complex activity limitations only >2 disabilities	512192661599205	58.69 (55.18–62.20)2.31 (1.22–3.40)2.61 (1.59–3.63)7.11 (5.23–9.00)6.43 (4.73–8.14)1.00 (0.34–1.64)21.85 (18.93–24.77)	16657138<5^e 42	67.07 (60.45–73.68)2.63 (0.10–5.32)2.65 (0.63–4.67))5.77 (2.42–9.12)2.68 (0.76–4.59)1.98 (0.10–4.13)17.23 (11.93–22.52)	9.11	0.1678
Comorbidities
Number of comorbidities012–3≥4	26725727795	30.39 (27.08–33.69)28.88 (25.16–32.14)30.48 (27.22–33.73)10.25 (8.11–12.39)	1853126<5^e	77.26 (71.72–82.81)12.23 (7.88–16.57)8.56 (5.16–11.95)1.95 (0.00–4.05)	146.89	<0.0001****
Substance abuse disorder No Yes	796100	88.75 (86.44–91.05)11.25 (8.95–13.56)	2406	98.36 (97.01–99.71)1.64 (0.29–2.99)	29.14	<0.0001****
Myocardial infarction No Yes	88115	98.09 (97.06–99.13)1.90 (0.87–2.94)	244<5^e	98.64 (96.71–100.0)1.36 (0.00–3.29)	0.19	0.6605
Congestive heart failure No Yes	889<5^e	99.29 (98.74–99.86 0.70 (0.14–1.26)	245<5^e	99.51 (98.55–100.0)0.49 (0.00–1.45)	0.12	0.7342
Peripheral vascular disease No Yes	892<5^e	99.65 (99.31–99.99)0.35 (0.00–0.69)	2460	100 (100.00–100.00)
Cancer No Yes	85145	94.97 (93.41–96.52)5.03 (3.48–6.59)	2406	97.49 (95.42–99.57)2.50 (0.43–4.58)	2.5330	0.1115
Dyslipidemia No Yes	618278	69.14 (65.88–73.39)30.86 (27.61–34.12)	23115	93.85 (90.61–97.08)6.15 (2.92–9.39)	53.72	<0.0001****
Diabetes mellitus No Yes	788108	88.59 (86.46–90.72)11.41 (9.28–13.54)	2397	97.52 (95.41–99.64)2.48 (0.36–4.59)	15.23	<0.0001****
Moderate/severe liver disease No Yes	88016	97.81 (96.56–99.07)2.19 (0.93–3.44)	245<5^e	99.51 (98.55–100.0)0.49 (0.00–1.45)	2.53	0.1117
Obesity No Yes	793103	89.62 (87.59–91.66)10.38 (8.34–12.41)	2379	96.30 (93.67–98.92)3.70 (1.08–6.33)	8.67	0.0032**
Kidney disease No Yes	83363	93.00 (91.24–94.76)7.00 (5.24–8.76)	245<5^e	99.72 (99.18–100.0)0.27 (0.00–0.82)	24.18	<0.0001****
Mild liver disease No Yes	81878	90.09 (87.70–92.49)9.91 (7.51–12.30)	2379	96.93 (94.83–99.03)3.07 (0.97–5.17)	11.93	0.0006***
Chronic pain No Yes	79997	89.50 (87.34–91.67)10.50 (8.33–12.66)	2415	98.02 (96.01–100.0)1.98 (0.00–3.98)	13.42	0.0002***
Chronic respiratory diseases No Yes	85343	95.62 (94.28–96.96)4.38 (3.04–5.72)	243<5^e	99.16 (98.17–100.0)0.50 (0.00–1.83)	9.24	0.0024**
Chronic gastrointestinal diseases No Yes	81779	91.91 (90.09–93.74)8.09 (6.26–9.91)	242<5^e	97.96 (95.82–100.0)2.04 (0.00–4.18)	7.71	0.0055**
Hypertension No Yes	643253	73.18 (70.11–76.25)26.82 (23.75–29.89)	21828	89.68 (85.71–93.66)10.32 (6.34–14.29)	26.96	<0.0001****
Cerebrovascular disease No Yes	8915	99.5 1 (99.05–99.96)0.49 (0.04–0.9477)	245<5^e	99.51 (98.55–100.0)0.49 (0.00–1.45)	0.0000	0.9965

Due to missing responses, frequencies for individual characteristics may not sum up to the overall sample total.

CI: confidence interval.

χ² value is based on the Rao–Scott modified statistic, which provides a design-based goodness-of-fit test using survey weights.

p values are from the Rao–Scott chi-square test.

Values less than 5 are suppressed to maintain confidentiality in accordance with CDC/HRSA reporting standards.

Unemployment status includes individuals out of work for more than 1 year, out of work for less than 1 year, homemakers, students, retirees, and those unable to work.

Assessed within the past 12 months.

ED: emergency department.

Significance level: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.00001.

Most of the cohort had health insurance coverage (92.3%). Health insurance status was significantly associated with HIV care retention (p < 0.0001), as 7.7% of uninsured individuals were not retained in care, compared to only 1.9% among those who were retained. None of the reported behavioral factors, such as alcohol use, smoking status, injection drug use, and electronic cigarette use, were significantly associated with RIC (p > 0.05). Likewise, psychosocial factors such as depression, anxiety, HIV stigma, discrimination, and patient-provider trust were not associated with RIC (p > 0.05). Those not retained in care reported a higher proportion of unmet needs for case management (21.25% vs 9.11%, p < 0.0001) and substance use treatment (3.12% vs 0.91%, p = 0.024) compared to those retained.

Clinical characteristics associated with RIC include HIV stage 3 disease at diagnosis, advanced HIV disease in the past 12 months, and durable viral suppression status (p < 0.05). Compared to those not retained in care, HIV stage 3 disease at diagnosis (55.3% vs 45.7%, p = 0.016), advanced HIV disease in the past 12 months (17.3% vs 9.5%, p = 0.002), and durable viral suppression (all undetectable or low viral loads (<200 copies/mL, 68.7% vs 18.9%, p < 0.0001). However, no significant differences were observed for the lowest CD4 count (p > 0.05) with respect to RIC. A significantly higher proportion of individuals retained in care reported a disability (41.3%) compared to those not retained (32.9%, p = 0.033). However, no significant association was found between specific disability types and retention in care (p = 0.167).

PWH retained in care had a significantly higher number of comorbid conditions compared to those not retained in care. A higher proportion of those retained in care had 2–3 comorbidities (30.5% vs 8.6%) or 4 or more comorbidities (10.25% vs 1.95%; p < 0.0001). Additionally, the prevalence of specific health conditions was notably higher among individuals retained in care compared to those not retained in care, including: substance use disorder (11.3% vs 1.6%, p < 0.0001), dyslipidemia (30.9% vs 6.15%, p < 0.0001), diabetes mellitus (11.4% vs 2.5%, p < 0.0001), obesity (10.4% vs 3.7%, p = 0.003), kidney disease (7.0% vs 0.3%, p < 0.0001), mild liver disease (9.9% vs 3.1%, p < 0.001), chronic pain (10.5% vs 1.9%, p < 0.001), chronic respiratory disease (4.4% vs 0.5%), p = 0.002), chronic gastrointestinal disease (8.1% vs 2.0%, p = 0.006), and hypertension (26.8% vs 10.3%, p < 0.0001).

When comorbidity and disability statuses were combined into a composite variable, RIC differed significantly (χ² = 153.1, p < 0.0001, Figure 1). RIC was highest among participants with both comorbidity and disability (30.1%) and those with comorbidity only (39.4%), compared to those with disability only (11.2%) and those without either condition (19.3%). Conversely, non-retention was most common among individuals with no comorbidity and no disability (50.8%), followed by those with disability only (26.4%), comorbidity only (16.3%), and both conditions (6.5%). Overall, individuals with comorbidities, either alone or in combination with disability, demonstrated higher RIC compared to those without comorbid conditions.

Figure 1.

Retention in care (RIC) by comorbidity and disability status.

Multivariable Modeling of Retention in HIV Care

Table 4 presents two models with the unadjusted and adjusted odds ratios (AOR) for predictors of RIC among PWH, using disability status or disability types as the primary exposure variables. In Model A, the main exposure was disability status. Model A shows that disability status was not significantly associated with RIC (AOR: 1.40; 95% CI: 0.94–2.10; p = 0.102). However, individuals who had unmet needs for case management were less likely to be retained in care (AOR: 0.60; 95% CI: 0.37–0.99, p = 0.045) compared to their counterparts. In addition, durable viral suppression was associated with a greater likelihood of RIC (AOR: 7.64, 95% CI: 5.09–11.44, p < 0.0001) compared to PWH who were not virally suppressed. A diagnosis of substance use disorder among PWH was associated with higher odds of RIC (AOR: 7.30; 95% CI 2.69–19.82, p < 0.0001) compared to those who did not have substance use disorders. Additionally, dyslipidemia was associated with a higher odds of RIC (AOR: 2.61, 95% CI: 1.39–4.92, p = 0.003). Likewise, those with a history of kidney disease had a higher RIC (AOR: 11.13, 95% CI: 1.28–96.60, p = 0.029) compared to those without a history of kidney disease.

Table 4.

Multivariable analysis of factors associated with retention in care among people with HIV—Houston Medical Monitoring Project, 2015–2021.

			Model A^a		Model B^b
Characteristics	Unadjusted OR	p values	AOR^c (95% CI^d)	p values	AOR^c (95% CI^d)	p values
Disability status No Yes	Ref1.43 (1.03–1.99)	0.0341*	Ref1.40 (0.94–2.10)	0.1019	–	–
Disability types No disability Hearing disability only Visual disability only Cognitive disability only Mobility disability only Complex activity Limitations >2 disabilities	Ref1.00 (0.31–3.21)1.13 (0.46–2.74)1.41 (0.71–2.81)2.75 (1.23–6.11)0.57 (0.16–2.09)1.45 (0.95–2.21)	0.72860.87300.62690.0261*0.19770.4098	–		Ref1.19 (0.35–4.02)0.38 (0.16–0.87)1.75 (0.85–3.59)3.40 (1.33–8.71)0.46 (0.07–3.02)1.39 (0.83–2.36)	0.85950.00780.16930.00790.30100.3579
Age-group (years)18–2930–3940–49≥50	Ref0.97 (0.59–1.62)1.08 (0.65–1.79)2.10 (1.29–3.44)	0.10900.3839<0.0001****	Ref0.74 (0.41–1.35)0.56 (0.31–1.04)0.78 (0.41–1.47)	0.94050.07290.8770	Ref0.74 (0.41–1.35)0.54 (0.29–1.02)0.76 (0.39–1.45)	0.98130.06460.9123
Race/Ethnicity White, non-Hispanic Black, non-Hispanic Hispanic or Latino Other	Ref0.45 (0.28–0.72)0.72 (0.41–1.25)0.58 (0.22–1.51)	0.0105**0.63120.7117	Ref0.69 (0.38–1.25)0.88 (0.44–1.74)0.77 (0.31–1.91)	0.28520.78020.8163	Ref0.69 (0.38–1.25)0.88 (0.44–1.78)0.80 (0.32–1.99)	0.24920.78610.8894
Gender Male Female Transgender	Ref0.64 (0.46–0.89)1.14 (0.30–4.30)	0.16400.6024	Ref0.73 (0.48–1.12)1.76 (0.48–6.49)	0.10710.2737	Ref0.74 (0.48–1.13)1.70 (0.45–6.45)	0.12970.3090
Foreign born No, not born in a foreign country Yes, born in a foreign country	Ref1.70 (1.09–2.64)	0.0173*	Ref1.64 (0.91–2.95)	0.0983	Ref1.74 (0.96–3.15)	0.0699
Health insurance Yes No	Ref0.24 (0.12–0.48)	<0.0001****	Ref0.53 (0.25–1.12)	0.0951	Ref0.47 (0.23–0.98)	0.0444*
Heavy drinking No Yes	Ref0.55 (0.29–1.05)	0.0683	Ref0.56 (0.25–1.26)	0.1626	Ref0.52 (0.23–1.17)	0.1112
HIV Stage 3 disease^e No Yes	Ref1.47 (1.07–2.02)	0.0167*	Ref1.12 (0.75–1.67)	0.5826	Ref1.16 (0.77–1.73)	0.4837
Unmet need for case management No Yes	Ref0.37 (0.24–0.57)	<0.0001****	Ref0.60 (0.37–0.99)	0.0448*	Ref0.58 (0.35–0.95)	0.0320*
Unmet need for substance use treatment No Yes	Ref0.30 (0.10–0.91)	0.0332*	Ref0.81 (0.21–3.18)	0.7851	Ref0.78 (0.19–3.27)	0.7374
Advanced HIV disease^f No Yes	Ref1.99 (1.27–3.12)	0.0027**	Ref1.76 (0.96–3.23)	0.0658	Ref1.82 (1.00–3.32)	0.0544
Durable viral suppression (copies/ml) ^e Any viral load ≥200 All viral loads undetectable or <200	Ref9.36 (6.52–13.44)	<0.0001****	Ref7.64 (5.09–11.44)	<0.0001***	Ref7.96 (5.28–11.99)	<0.0001***
Substance use disorder No Yes	Ref7.60 (3.19–18.11)	<0.0001****	Ref7.30 (2.69–19.82)	<0.0001***	Ref7.36 (2.68–20.21)	<0.0001***
Dyslipidemia No Yes	Ref6.81 (3.80–12.19)	<0.0001****	Ref2.61 (1.39–4.92)	0.0030**	Ref2.66 (1.40–5.03)	0.0027**
Obesity No Yes	Ref3.01 (1.39–6.49)	0.0050**	Ref1.45 (0.49–4.26)	0.5002	Ref1.43 (0.48–4.27)	0.5200
Diabetes mellitus No Yes	Ref5.07 (2.06–12.49)	0.0004***	Ref2.15 (0.78–5.94)	0.1394	Ref2.04 (0.76–5.49)	0.1577
Mild liver disease No Yes	Ref3.47 (1.63–7.37)	0.0013**	Ref2.57 (1.00–6.62)	0.0497*	Ref2.63 (0.96–7.24)	0.0613
Chronic pain No Yes	Ref5.81 (2.01–16.77)	0.0012**	Ref2.27 (0.66–7.78)	0.1937	Ref2.38 (0.68–8.42)	0.1771
Chronic respiratory disease No Yes	Ref5.38 (1.59–18.17)	0.0068**	Ref1.94 (0.38–10.03)	0.4293	Ref2.37 (0.49–11.23)	0.2774
Chronic gastrointestinal diseases No Yes	Ref4.23 (1.40–12.79)	0.0105*	Ref2.73 (0.85–8.75)	0.0916	Ref2.75 (0.84–8.98)	0.0933
Kidney disease No Yes	Ref27.03 (3.71–196.81)	0.0012***	Ref11.13 (1.28–96.60)	0.0289*	Ref14.08 (1.44–138.12)	0.0232*
Hypertension No Yes	Ref3.18 (2.02–5.03)	<0.0001****	Ref1.52 (0.87–2.67)	0.1391	Ref1.51 (0.86–2.64)	0.1544

Model A: Disability status adjusted for age, race/ethnicity, gender, foreign born, health insurance, heavy drinking, HIV Stage 3 disease, unmet need for mental health services, unmet need for substance use treatment, advanced HIV disease, durable viral suppression, substance use disorder, dyslipidemia, obesity, diabetes mellitus, mild liver disease, chronic pain, chronic respiratory disease, chronic gastrointestinal disease, kidney disease, and hypertension.

Model B: Specific disability types adjusted for age, race/ethnicity, gender, foreign born, health insurance, heavy drinking, HIV Stage 3 disease, unmet need for mental health services, unmet need for substance use treatment, advanced HIV disease, durable viral suppression, substance use disorder, dyslipidemia, obesity, diabetes mellitus, mild liver disease, chronic pain, chronic respiratory disease, chronic gastrointestinal disease, kidney disease, and hypertension.

AOR: adjusted odds ratio.

CI: confidence interval.

Assessed at diagnosis.

Assessed in the past 12 months.

Significance level: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.00001.

Model B included disability types as the main exposure. Model B shows that PWH who reported only visual disability have significantly lower odds of RIC compared to those with no disability (AOR: 0.38; 95% CI: 0.16–0.87, p = 0.007). However, reporting only mobility disability was associated with higher odds of RIC (AOR: 3.40; 95% CI: 1.33–8.71, p = 0.008) compared to those with no disability. Similarly, durable viral suppression (HIV viral load undetectable or <200 copies/ml) was linked to greater odds of RIC (AOR: 7.96; 95% CI: 5.28–11.99, p < 0.0001) when compared to PWH who were not virally suppressed. However, the lack of health insurance was associated with lower odds of RIC (AOR: 0.47; 95% CI: 0.23–0.98, p = 0.044) compared to PWH who had health insurance. Also, participants who had unmet needs for case management were less likely to be retained in care (AOR: 0.58; 95% CI: 0.35–0.95, p = 0.032) compared to their counterparts. A diagnosis of substance use disorder among PWH was associated with higher odds of RIC (AOR: 7.36; 95% CI: 2.68–20.21, p < 0.0001) compared to those who did not have substance use disorder. PWH with dyslipidemia were associated with higher odds of RIC (AOR: 2.66, 95% CI: 1.40–5.03, p = 0.003). Likewise, those with a history of kidney disease were strongly associated with RIC (AOR: 14.08, 95% CI: 1.44–138.12, p = 0.023) compared to those without kidney disease.

Furthermore, to examine the complexity and potential synergistic effects of functional disability and comorbidities on RIC, we conducted a multivariable logistic regression analysis using a composite variable representing both conditions as the primary exposure (Model C). The model was adjusted for age, race, gender, foreign-born status, health insurance, HIV stage 3 disease, heavy drinking, unmet needs for case management and substance use treatment, advanced HIV disease, and durable viral suppression. The results indicated that individuals with both comorbidities and disabilities were more likely to be retained in care compared with those with neither condition (AOR: 11.16; 95% CI: 5.67–21.97; p < 0.0001). However, disability alone was not significantly associated with retention, while those with comorbidities had more than four times the odds of being retained in care (AOR = 4.59; 95% CI: 2.81–7.52; p < 0.0001) compared to those with neither condition (table not shown).

Discussion

In this population of PWH in Houston, Texas, 74.9% were retained in HIV care in the preceding year, which still falls below the recommended benchmark of 90%.³⁵ This underscores the need to identify factors that influence RIC, as higher retention is linked to improved health outcomes.^2,35

In unadjusted regression analysis, sociodemographic factors such as being 50 years or older and foreign-born status were associated with a positive RIC. This is consistent with prior research showing that older adults are more likely to be retained in care, reflecting greater health care stability,³⁶ while a younger age is negatively associated with RIC.²¹ Also, foreign-born individuals were more likely to be retained in care, and existing studies have corroborated this finding, which might be a result of better healthcare-seeking behavior.³⁷ Conversely, racial disparities were evident, with Black, non-Hispanic individuals being less likely to be retained in care compared to their White, non-Hispanic counterparts. This finding is consistent with previous research that stresses barriers encountered by Black non-Hispanics in achieving optimal engagement in care, such as HIV-related stigma and poverty.^23,38,39

In multivariable analysis, although overall disability status was not significantly associated with RIC, specific types of disability showed notable associations. PWH with only a visual disability were less likely to be retained in care compared to those without any disability. This finding suggests that people with visual impairments may encounter greater barriers to accessing and utilizing healthcare services. Supporting this, a recent study among men living with HIV reported that visual impairment was associated with a lower likelihood of seeking necessary medical care compared to those without vision difficulties.⁴⁰ Existing literature suggests that the various ways in which visual disability contributes to poor RIC include lower patient activation,⁴¹ inability to drive to appointments,⁴² difficulty reading appointment reminders,⁴³ difficulty navigating healthcare facilities without company,⁴² and problems understanding medical instructions.⁴³ Therefore, visual disability may impede activities of daily living.⁴⁴ In contrast, having only a mobility disability was associated with increased RIC. Similar to previous research among the general population, people with mobility disabilities had a higher prevalence of access to care than those reporting vision disability.¹⁶

Surprisingly, a diagnosis of substance use disorder (SUD) was positively associated with RIC, contrasting with a previous study that reported no significant relationship.⁴⁵ Moreover, robust evidence has demonstrated that alcohol and other substance use disorders can negatively impact RIC.^46–48 In this study, substance use disorders were identified using ICD-10 codes from abstractions of medical records, which may have facilitated access to substance use treatment services and care coordination, ultimately enhancing RIC. Additionally, differences in how RIC is defined and measured across studies may also account for discrepancies in findings.⁴⁹ Although this study adapted the Health Resources and Services Administration (HRSA) definition of RIC, consistent with national performance standards, alternative metrics that track missed or gap visits over specific intervals may capture different dimensions of RIC and contribute to observed variations across studies.⁵⁰ It is also plausible that integrated care models targeting high-risk populations are effectively identifying individuals with substance use disorders, linking them to HIV testing, and engaging them in care as previously recommended by research.⁵¹

Similarly, those with dyslipidemia and kidney disease had a higher likelihood of RIC. Research exploring the relationship between comorbidities and RIC has yielded mixed results.^52–55 While a study suggested that comorbid conditions increased the fatigue severity in patients receiving HIV care, which in turn hindered RIC,⁵² other studies have reported positive relationships between RIC and comorbid conditions among PWH.^54,55 When comorbidity and disability were modeled jointly, those with both conditions had higher RIC than individuals without either, while disability alone showed no significant association. The positive association between comorbidities and RIC may indicate that patients with more complex health profiles receive more intensive follow-up and multidisciplinary care, which in turn enhances RIC.^1,56

A lack of health insurance coverage and unmet needs for case management were negatively associated with RIC, consistent with previous research identifying these as structural barriers to sustained engagement in care.² Case management helps address structural and psychosocial barriers by linking PWH to needed medical and non-medical services, and studies have shown that case management interventions improve long-term RIC.^2,57 Expanding access to health insurance and case management services is crucial to achieving sustained RIC improvement.

Additionally, we found that durable viral suppression was associated with increased RIC. This finding aligns with previous studies that have assessed the effect of RIC on viral suppression.^56,58 PWH who are retained in care and consistently taking ART can achieve viral suppression in their serum to undetectable levels, effectively eliminating the risk of transmitting HIV to others, and contributing significantly to public health goals. This underscores the critical importance of strengthening RIC to improve treatment outcomes among PWH.

Our study findings highlight the need to improve RIC among PWH who have visual disabilities in Houston/Harris County, Texas. The Ryan White HIV/AIDS Program (RWHAP), which provides federal funding for low-income PWH, offers limited vision services, such as preventive eye care and low-vision training, that can help maintain functional independence and support ongoing engagement in care.⁵⁹ However, as a payer of last resort, RWHAP typically does not cover major or non-HIV-related eye procedures. In the southern United States, including Texas, limited Medicaid expansion and restrictive eligibility criteria also exacerbate gaps in access to comprehensive HIV care.⁶⁰ Therefore, expanding health insurance coverage, case management, and rehabilitative visual services could substantially improve the management of comorbidities and enhance RIC and viral suppression, ultimately reducing HIV transmission and improving quality of life for PWH.

Limitations and Strengths

This study has several limitations. First, its cross-sectional design prevents us from establishing temporal or causal relationships between disability, other associated factors, and RIC. Second, much of the data relied on self-reported measures, which may be prone to biases such as social desirability and recall errors. Despite multivariable adjustments made, unmeasured residual confounders (e.g. barriers to care, provider bias, caregiving burden, etc.) could influence both disability and RIC. Although the study disaggregates disability types, it may not fully capture the severity, onset, or duration of disability, which could affect RIC differently. Because available-case analysis assumes that data are missing completely at random, its use may introduce bias if this assumption is violated. This potential bias should be considered when interpreting the results. Additionally, although the MMP data had a relatively low response rate, weighting and statistical adjustments were applied to reduce nonresponse bias.^27,28 The findings of our study are only generalizable to the PWH in Houston/Harris County, Texas. However, any attempts to generalize these findings to other settings or populations should be approached with caution, as the current study was based on cross-sectional surveillance data from Houston/Harris County, Texas, which has unique demographic, structural, and healthcare characteristics.

Despite these limitations, the current study offers several notable strengths. The MMP's rigorous probability-based sampling methodology enabled the collection of a representative sample of PWH in Houston/Harris County, Texas, thereby enhancing the generalizability of the findings. Additionally, the MMP dataset provides rich, multidimensional data, encompassing detailed sociodemographic, behavioral, and clinical information, which allowed for a comprehensive analysis of factors associated with RIC. Importantly, this study addresses a critical gap in the literature by examining disability-specific associations with RIC, offering new insights into how different types of disability may uniquely influence care engagement. The findings have direct implications for tailoring interventions to key subgroups, such as individuals with visual impairments, and inform policy initiatives aimed at improving outcomes across the HIV care continuum.

Conclusion

As people with PWH age, the prevalence of disability is expected to increase, underscoring the growing importance of understanding how disability and comorbidity influence care outcomes. This study highlights the complex interplay between comorbidities, types of disabilities, and RIC among PWH in Houston/Harris County, Texas. While certain comorbidities and mobility disabilities were associated with higher retention, visual disabilities were linked to poorer engagement in care. Although ocular complications of HIV are well recognized,⁶¹ efforts should also focus on enrolling underinsured PWH into the RWHAP to ensure access to preventive vision care and comprehensive case management services.

These findings emphasize the need to integrate disability-sensitive and comorbidity-informed approaches into HIV care delivery. Tailored interventions that address the unique challenges faced by individuals with visual or other functional disabilities are essential to promoting equitable access and long-term engagement in care. Future research should explore the underlying mechanisms through which disabilities and comorbidities affect care engagement and assess how social determinants of health, insurance coverage, and access to assistive services shape these outcomes. Many individuals, even when insured, may lack adequate coverage for specialized services or assistive devices necessary to manage advanced functional disabilities, further widening disparities in care.

To improve retention and health outcomes, multilevel, disability-inclusive interventions that address both structural and individual-level barriers to care are critically needed. Culturally tailored strategies that expand access to supportive services, enhance provider responsiveness to disability and comorbidity, and promote equity within health systems will be vital to achieving sustained improvements in care engagement. Strengthening multidisciplinary support systems and adopting person-centered models of care may further improve retention, advance health equity, and contribute to the national goal of ending the HIV epidemic in the United States by 2030.⁶

Footnotes

Acknowledgments

We want to thank all the participants and project staff who contributed to the Medical Monitoring Project (MMP) in Houston/Harris County, Texas, during the 2015–2021 data collection cycles. We appreciate the MMP staff's diligent efforts in data collection and the valuable guidance and input of the Community and Provider Advisory Board members. Lastly, we acknowledge the leadership and support provided by the Houston Health Department's management team, whose collaboration was integral to the project's success.

ORCID iDs

Modupe Olajumoke Onigbogi

Osaro Mgbere

Charles Darkoh

Ethics Consideration

The current study protocol was submitted to the Committee for the Protection of Human Subjects at the University of Texas Health Science Center at Houston for review, where it received an Exempt Status approval [IRB reference # 259145].

Author Contributions

Conceptualization: [Modupe Olajumoke Onigbogi, Osaro Mgbere]; methodology: [Modupe Olajumoke Onigbogi, Osaro Mgbere]; formal analysis and investigation: [Modupe Olajumoke Onigbogi, Osaro Mgbere, Ruosha Li]; writing–original draft preparation: [Modupe Olajumoke Onigbogi, Osaro Mgbere]; writing–review and editing: [Modupe Olajumoke Onigbogi, Osaro Mgbere, Charles Darkoh, Ruosha Li, Paul Rowan]; supervision: [Osaro Mgbere, Charles Darkoh, Ruosha Li, Paul Rowan].

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Support for the Medical Monitoring Project was provided by the U.S. Centers for Disease Control and Prevention (CDC) through Cooperative Agreement numbers CDC-RFA-PS15-1503 and CDC-RFA-PS20-2005. This work was partly supported by the National Institute of Allergy and Infectious Diseases grants R01AI116914, R01AI150685, and R. Palmer Beasley, M.D. & Lu-Yu Hwang, M.D. Endowment.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The Houston Medical Monitoring Project data used for this study are under the custody of the Houston Health Department. Due to data use agreements and confidentiality protections, these data cannot be shared publicly. Readers seeking additional information or access to the data may contact the Houston Health Department directly.

Disclaimer

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention or the Houston Health Department.

References

Mugavero

Lin

Willig

, et al. Missed visits and mortality among patients establishing initial outpatient HIV treatment. Clin Infect Dis. 2009;48(2):248–256. doi:10.1086/595705

Yehia

Stewart

Momplaisir

, et al. Barriers and facilitators to patient retention in HIV care. BMC Infect Dis. 2015;15:246. doi:10.1186/s12879-015-0990-0

Kay

Bassler

Goebel

, et al. Retention in HIV care still matters: a longitudinal multisite analysis. J Acquir Immune Defic Syndr. 2025;100(2):123–130. doi:10.1097/QAI.0000000000003711

Skarbinski

Rosenberg

Paz-Bailey

, et al. Human immunodeficiency virus transmission at each step of the care continuum in the United States. JAMA Intern Med. 2015;175(4):588–596. doi:10.1001/jamainternmed.2014.8180

HV Surveillance Supplemental Report: Monitoring Selected National HIV Prevention and Care Objectives by Using HIV Surveillance Data United States and 6 Territories and Freely Associated States, 2022. Accessed June 27, 2025. https://stacks.cdc.gov/view/cdc/156511

Fauci

Redfield

Sigounas

Weahkee

Giroir

. Ending the HIV epidemic: a plan for the United States. JAMA. 2019;321(9):844–845. doi:10.1001/jama.2019.1343

Deeks

Lewin

Havlir

. The end of AIDS: HIV infection as a chronic disease. Lancet. 2013;382(9903):1525–1533. doi:10.1016/S0140-6736(13)61809-7

Maciel

Klück

Durand

Sprinz

. Comorbidity is more common and occurs earlier in persons living with HIV than in HIV-uninfected matched controls, aged 50 years and older: A cross-sectional study. Int J Infect Dis. 2018;70:30–35. doi:10.1016/j.ijid.2018.02.009

Marcus

Leyden

Alexeeff

, et al. Comparison of overall and comorbidity-free life expectancy between insured adults with and without HIV infection, 2000–2016. JAMA Netw Open. 2020;3(6):e207954. doi:10.1001/jamanetworkopen.2020.7954

10.

Rajasuriar

Chong

Ahmad Bashah

, et al. Major health impact of accelerated aging in young HIV-infected individuals on antiretroviral therapy. AIDS. 2017;31(10):1393–1403. doi:10.1097/QAD.0000000000001475

11.

Morgan

Iudicello

Weber

, et al. Synergistic effects of HIV infection and older age on daily functioning. J Acquir Immune Defic Syndr. 2012;61(3):341–348. doi:10.1097/QAI.0b013e31826bfc53

12.

Althoff

Stewart

Humes

, et al. The forecasted prevalence of comorbidities and multimorbidity in people with HIV in the United States through the year 2030: A modeling study. PLoS Med. 2024;21(1):e1004325. doi:10.1371/journal.pmed.1004325

13.

Johs

Tassiopoulos

, et al. Disability among middle-aged and older persons with human immunodeficiency virus infection. Clin Infect Dis. 2017;65(1):83–91. doi:10.1093/cid/cix253

14.

Chowdhury

Beer

Shu

Fagan

Luke Shouse

. Disability among adults with diagnosed HIV in the United States, 2017. AIDS Care. 2021;33(12):1611–1615. doi:10.1080/09540121.2020.1842318

15.

Westra

Giordano

Berenson

Baillargeon

Kuo

. Assessing comorbidities and survival in HIV-infected and uninfected matched Medicare enrollees. AIDS. 2021;35(10):1667–1675. doi:10.1097/QAD.0000000000002963

16.

Okoro

Hollis

Cyrus

Griffin-Blake

. Prevalence of disabilities and health care access by disability status and type among adults — United States, 2016. MMWR Morb Mortal Wkly Rep. 2018;67(32):882–887. doi:10.15585/mmwr.mm6732a3

17.

Kuper

Heydt

Davey

. A focus on disability is necessary to achieve HIV epidemic control. Lancet HIV. 2022;9(4):e293–e298. doi:10.1016/S2352-3018(21)00345-3

18.

O’Brien

Bayoumi

Strike

Young

Davis

. Exploring disability from the perspective of adults living with HIV/AIDS: Development of a conceptual framework. Health Qual Life Outcomes. 2008;6:76. doi:10.1186/1477-7525-6-76

19.

Erlandson

Schrack

Jankowski

Brown

Campbell

. Functional impairment, disability, and frailty in adults aging with HIV-infection. Curr HIV/AIDS Rep. 2014;11(3):279–290. doi:10.1007/s11904-014-0215-y

20.

Verheij

Boyd

Wit

, et al. Long-term evolution of comorbidities and their disease burden in individuals with and without HIV as they age: Analysis of the prospective AGEhIV cohort study. Lancet HIV. 2023;10(3):e164–e174. doi:10.1016/S2352-3018(22)00400-3

21.

Wester

Rebeiro

Shavor

, et al. The 2013 HIV continuum of care in Tennessee: Progress made, but disparities persist. Public Health Rep. 2016;131(5):695–703. doi:10.1177/0033354916660082

22.

Rebeiro

Gange

Horberg

, et al. Geographic variations in retention in care among HIV-infected adults in the United States. Fox MP, ed. PLoS ONE. 2016;11(1):e0146119. doi:10.1371/journal.pone.0146119

23.

Rebeiro

McPherson

Goggins

, et al. Health literacy and demographic disparities in HIV care Continuum outcomes. AIDS Behav. 2018;22(8):2604–2614. doi:10.1007/s10461-018-2092-7

24.

Kay

Lacombe-Duncan

Pinto

. Predicting retention in HIV primary care: Is there a missed visits Continuum based on patient characteristics? AIDS Behav. 2019;23(9):2542–2548. doi:10.1007/s10461-019-02508-6

25.

Pence

Bengtson

Boswell

, et al. Who will show? Predicting missed visits among patients in routine HIV primary care in the United States. AIDS Behav. 2019;23(2):418–426. doi:10.1007/s10461-018-2215-1

26.

López

Qiao

Presti

Hammer

Foraker

. The impact of neighborhood socioeconomic status on retention in care and viral suppression among people living with HIV. AIDS Care. 2022;34(11):1383–1389. doi:10.1080/09540121.2022.2040724

27.

Beer

Johnson

Fagan

, et al. A national behavioral and clinical surveillance system of adults with diagnosed HIV (the Medical Monitoring Project): Protocol for an annual cross-sectional interview and medical record abstraction survey. JMIR Res Protoc. 2019;8(11):e15453. doi:10.2196/15453

28.

Johnson

Beer

Harding

, et al. Changes to the sample design and weighting methods of a public health surveillance system to also include persons not receiving HIV medical care. PLoS One. 2020;15(12):e0243351. doi:10.1371/journal.pone.0243351

29.

Elm

Altman

Egger

, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: Guidelines for reporting observational studies. PLoS Med. 2007;4(10):e296. doi:10.1371/journal.pmed.0040296

30.

Beer

Tie

McCree

, et al. HIV Stigma among a national probability sample of adults with diagnosed HIV-United States, 2018-2019. AIDS Behav. 2022;26(Suppl 1):39–50. doi:10.1007/s10461-021-03414-6

31.

Horner-Johnson W, Dobbertin K, Lee JC, Andresen EM, the Expert Panel on Disability and Health Disparities. Disparities in health care access and receipt of preventive services by disability type: Analysis of the medical expenditure panel survey. Health Serv Res. 2014;49(6):1980–1999. doi:10.1111/1475-6773.12195

32.

O’Halloran Leach

Caballero

Thomas

Spencer

Cook

. Defining the optimal cut-point of self-reported ART adherence to achieve viral suppression in the era of contemporary HIV therapy: A cross-sectional study. AIDS Res Ther. 2021;18(1):36. doi:10.1186/s12981-021-00358-8

33.

Bento

Nguyen

Naik

. HIV-1–associated opportunistic infections. In: StatPearls [Internet]. StatPearls Publishing; 2025. Available from: https://www.ncbi.nlm.nih.gov/books/NBK539787/.

34.

Prommik

Tootsi

Saluse

Strauss

Kolk

Märtson

. Simple excel and ICD-10 based dataset calculator for the Charlson and Elixhauser comorbidity indices. BMC Med Res Methodol. 2022;22(1):4. doi:10.1186/s12874-021-01492-7

35.

Mugavero

. Elements of the HIV care continuum: Improving engagement and retention in care. Top Antivir Med. 2016;24(3):115–119.

36.

Horberg

Hurley

Klein

, et al. The HIV care cascade measured over time and by age, sex, and race in a large national integrated care system. AIDS Patient Care STDS. 2015;29(11):582–590. doi:10.1089/apc.2015.0139

37.

Levison

Regan

Khan

Freedberg

. Foreign-born status as a predictor of engagement in HIV care in a large US metropolitan health system. AIDS Care. 2017;29(2):244–251. doi:10.1080/09540121.2016.1210077

38.

Adeyemi

Livak

McLoyd

Smith

French

. Racial/ethnic disparities in engagement in care and viral suppression in a large urban HIV clinic. Clin Infect Dis. 2013;56(10):1512–1514. doi:10.1093/cid/cit063

39.

Daoud

Gladstein

Brixner

O’Brochta

Naik

. Health disparities in HIV care and strategies for improving equitable access to care. Am J Manag Care. 2025;31(1 Suppl):S3–S12. doi:10.37765/ajmc.2025.89687

40.

Abraham

Tong

Stosor

Friedman

Detels

Plankey

. Vision problems as a contributor to lower engagement in care among aging men living with HIV. Ophthalmic Epidemiol. 2025;32(2):143–152. doi:10.1080/09286586.2024.2346894

41.

Assi

Kozhaya

Swenor

Reed

. Vision impairment and patient activation among Medicare beneficiaries. Ophthalmic Epidemiol. 2022;30(2):159–165. doi:10.1080/09286586.2022.2078495

42.

Reed

Assi

Pedersen

, et al. Accompaniment to healthcare visits: The impact of sensory impairment. BMC Health Serv Res. 2020;20(1):990. doi:10.1186/s12913-020-05829-8

43.

Brown

Goldstein

Chan

Massof

Ramulu

. Characterizing functional complaints in patients seeking outpatient low-vision services in the United States. Ophthalmology. 2014;121(8):1655–1662.e1. doi:10.1016/j.ophtha.2014.02.030

44.

Relationship between vision impairment and ability to perform activities of daily living – Haymes – 2002 – Ophthalmic and Physiological Optics – Wiley Online Library. Accessed May 1, 2025. https://onlinelibrary.wiley.com/doi/full/10.1046/j.1475-1313.2002.00016.x

45.

Doshi

Byrne

Levy

, et al. Association of substance use disorders with engagement in care and mortality among a clinical cohort of people with HIV in Washington, DC. AIDS Behav. 2021;25(7):2289. doi:10.1007/s10461-021-03157-4

46.

Howe

Cole

Napravnik

, et al. The role of at-risk alcohol/drug use and treatment in appointment attendance and virologic suppression among HIV+ African Americans. AIDS Res Hum Retrovir. 2014;30(3):233–240. doi:10.1089/aid.2013.0163

47.

Monroe

Lau

Mugavero

, et al. Heavy alcohol use is associated with worse retention in HIV care. J Acquir Immune Defic Syndr. 2016;73(4):419–425. doi:10.1097/QAI.0000000000001083

48.

Hartzler

Dombrowski

Williams

, et al. Influence of substance use disorders on two-year HIV care retention in the United States. AIDS Behav. 2018;22(3):742. doi:10.1007/s10461-017-1826-2

49.

Rehman

Garcia

Jones

Mertz

Mbuagbaw

. Measures of retention in HIV care: A protocol for a mixed methods study. PLOS ONE. 2024;19(2):e0294824. doi:10.1371/journal.pone.0294824

50.

Lesko

Mugavero

Shen

, et al. Exploring definitions of retention in care for people living with HIV in the United States in the modern treatment era. AIDS. 2022;36(8):1181–1189. doi:10.1097/QAD.0000000000003232

51.

Hill

Kuo

Shenoi

Desruisseaux

Springer

. Integrated care models: HIV and substance use. Curr HIV/AIDS Rep. 2023;20(5):286–295. doi:10.1007/s11904-023-00667-9

52.

Corless

Voss

Nicholas

, et al. Fatigue in HIV/AIDS patients with comorbidities. Appl Nurs Res. 2008;21(3):116–122. doi:10.1016/j.apnr.2006.11.001

53.

Bulsara

Wainberg

Rogers

McAloon

Grove

Newton-John

TRO

. The role of comorbidity on retention in HIV care. AIDS Behav. 2021;25(5):1532–1541. doi:10.1007/s10461-020-02992-1

54.

Crawford

. Examining the relationship between multiple comorbidities and retention in HIV medical care: A retrospective analysis. AIDS Care. 2015;27(7):892–899. doi:10.1080/09540121.2015.1009361

55.

Giordano

Hartman

Gifford

Backus

Morgan

. Predictors of retention in HIV care among a national cohort of US veterans. HIV Clin Trials. 2009;10(5):299–305. doi:10.1310/hct1005-299

56.

Yehia

Stephens-Shields

Fleishman

, et al. The HIV care continuum: Changes over time in retention in care and viral suppression. PLoS ONE. 2015;10(6):e0129376. doi:10.1371/journal.pone.0129376

57.

Okeke

Ostermann

Thielman

. Enhancing linkage and retention in HIV care: a review of interventions for highly resourced and resource-poor settings. Curr HIV/AIDS Rep. 2014;11(4):376–392. doi:10.1007/s11904-014-0233-9

58.

Yehia

French

Fleishman

, et al. Retention in care is more strongly associated with viral suppression in HIV-infected patients with lower versus higher CD4 counts. J Acquir Immune Defic Syndr. 2014;65(3):333–339. doi:10.1097/QAI.0000000000000023

59.

Ginossar

Oetzel

Van Meter

Gans

Gallant

. The Ryan white HIV/AIDS program after the patient protection and affordable care act full implementation: a critical review of predictions, evidence, and future directions. Top Antivir Med. 2019;27(3):91–100.

60.

Abbas

Hallmark

McNeese

, et al. Human immunodeficiency virus in the State of Texas of the United States: past reflections, present shortcomings, and future needs of the public health response. Open Forum Infect Dis. 2020;7(10):ofaa348. doi:10.1093/ofid/ofaa348

61.

Wons

Kempen

Garweg

. HIV-induced retinitis. Ocul Immunol Inflamm. 2020;28(8):1259–1268. doi:10.1080/09273948.2020.1808225