Relationships Between Patient Race and Residential Race Context With Missed Human Immunodeficiency Virus Care Visits in the United States, 2010-2015

BACKGROUND

Racial inequities exist in retention in human immunodeficiency virus (HIV) care and multilevel analyses are needed to contextualize and address these differences. Leveraging data from a multisite clinical cohort of people with HIV (PWH), we assessed the relationships between patient race and residential characteristics with missed HIV care visits.

METHODS

Medical record and patient-reported outcome (PRO; including mental health and substance-use measures) data were drawn from 7 participating Center for AIDS Research Network of Integrated Clinical Systems (CNICS) sites including N = 20 807 PWH from January 2010 through December 2015. Generalized estimating equations were used to account for nesting within individuals and within census tracts in multivariable models assessing the relationship between race and missed HIV care visits, controlling for individual demographic and health characteristics and census tract characteristics.

RESULTS

Black PWH resided in more disadvantaged census tracts, on average. Black PWH residing in census tracts with higher proportion of Black residents were more likely to miss an HIV care visit. Non-Black PWH were less likely to miss a visit regardless of where they lived. These relationships were attenuated when PRO data were included.

CONCLUSIONS

Residential racial segregation and disadvantage may create inequities between Black PWH and non-Black PWH in retention in HIV care. Multilevel approaches are needed to retain PWH in HIV care, accounting for community, healthcare setting, and individual needs and resources.

Patient-Reported Outcomes Collection at an Urban HIV Clinic Associated With a Historically Black Medical College in the Southern United States: Qualitative Interview Study Among Patients With HIV

BACKGROUND

Black Americans, particularly in the southern United States, are disproportionately affected by the US HIV epidemic. Patient-reported outcome (PRO) data collection can improve patient outcomes and provide oft-overlooked data on mental health, substance use, and patient adherence to antiretroviral therapy.

OBJECTIVE

We piloted the use of an electronic tablet to collect PRO data on social and behavioral determinants of health among people with HIV at the Meharry Community Wellness Center, an HIV clinic affiliated with a Historically Black Medical College in Nashville, Tennessee. Our primary objective was to better understand patients' experiences and comfort with using an electronic PRO tool through patient interviews.

METHODS

We enrolled 100 people with HIV in care at the Meharry Community Wellness Center consecutively to completely validate PRO tools using the Research Electronic Data Capture platform on a hand-held tablet. Using a purposive sampling strategy, we enrolled 20 of the 100 participants in an in-depth interview (IDI). Interview guide development was grounded in the cognitive-behavioral model, in which thoughts, feelings, and behaviors are interrelated. IDIs were audio recorded, transcribed, deidentified, and formatted for coding. A hierarchical coding system was developed and refined using an inductive-deductive approach.

RESULTS

Among the 100 people with HIV enrolled, the median age was 50 (IQR 42-54) years; 89% (n=89) were Black, 60% (n=60) were male, and 82% (n=82) were living below 100% of the federal poverty level in 2016. Five major interview themes emerged: overall experience, question content, sensitive topics, clinic visit impact, and future recommendations. IDI participants felt that the tablet was easy to use and that the question content was meaningful. Question content related to trauma, sexual and drug use behaviors, mental health, stigma, and discrimination elicited uncomfortable or distressing feelings in some participants. Patients expressed a strong desire to be truthful, and most would complete these surveys without compensation at future visits if offered.

CONCLUSIONS

The use of an electronic tablet to complete PRO data collection was well received by this cohort of vulnerable persons in HIV care in the southern United States. Despite some discomfort related to question content, our cohort overwhelmingly believed this was a meaningful part of their medical experience and expressed a high desire for truthfulness. Future research will focus on scaling up the implementation and evaluation of PRO data collection in a contextually appropriate manner while obtaining input from providers and staff to ensure that the collected data are both applicable and actionable.

Development and Validation of a Multivariable Prediction Model for Missed HIV Health Care Provider Visits in a Large US Clinical Cohort

Background

Identifying individuals at high risk of missing HIV care provider visits could support proactive intervention. Previous prediction models for missed visits have not incorporated data beyond the individual level.

Methods

We developed prediction models for missed visits among people with HIV (PWH) with ≥1 follow-up visit in the Center for AIDS Research Network of Integrated Clinical Systems from 2010 to 2016. Individual-level (medical record data and patient-reported outcomes), community-level (American Community Survey), HIV care site–level (standardized clinic leadership survey), and structural-level (HIV criminalization laws, Medicaid expansion, and state AIDS Drug Assistance Program budget) predictors were included. Models were developed using random forests with 10-fold cross-validation; candidate models with the highest area under the curve (AUC) were identified.

Results

Data from 382 432 visits among 20 807 PWH followed for a median of 3.8 years were included; the median age was 44 years, 81% were male, 37% were Black, 15% reported injection drug use, and 57% reported male-to-male sexual contact. The highest AUC was 0.76, and the strongest predictors were at the individual level (prior visit adherence, age, CD4+ count) and community level (proportion living in poverty, unemployed, and of Black race). A simplified model, including readily accessible variables available in a web-based calculator, had a slightly lower AUC of .700.

Conclusions

Prediction models validated using multilevel data had a similar AUC to previous models developed using only individual-level data. The strongest predictors were individual-level variables, particularly prior visit adherence, though community-level variables were also predictive. Absent additional data, PWH with previous missed visits should be prioritized by interventions to improve visit adherence.