In this study, we reported the prevalence of health-related social needs (HRSN) among hospitalized PLWH in the two states of the U.S (Florida: 9.6%; Maryland: 12.5%). PLWH with documented HRSN had nearly twice the 90-day and one-year readmission rate compared with those without HRSN. PLWH mainly carried three domains of HRSN, including inadequate housing, unemployment, and insufficient family/social support. Our findings revealed that the social risk in PLWH is strongly associated with high hospital readmission rate.

Hospital readmission is one of the indicators of quality of care for chronic diseases, and a high readmission rate is reported to be associated with substandard care during the index hospitalization²² Although HIV infection is now a manageable chronic disease, PLWH have a higher rate of readmission than those without HIV. A multi-site study has reported that the 30-day readmission rate of PLWH in the U.S. is about 19.3%, compared with only 13.3% for non-PLWH for the same follow-up time²³. In a large study from all hospitals in the state of New York, the 30-day readmission of PLWH has been reported to be 21.8%.²⁴ In addition, the findings from previous studies indicate that a web of factors associated with 30-day readmission for PLWH include AIDS-defining illness, low CD4 cell count, laboratory abnormalities, psychoses, multiple comorbidities, but also social factors such as insurance status, housing instability (including homelessness), distance from the hospital, and prior emergency department visits and hospitalizations^24,25 These findings indicate that non-medical social factors play an important role in hospital readmission. It has also been proposed that about half of 30-day readmissions in PLWH are potentially preventable via access to HAART at discharge, linkage to follow-up care, management of chronic conditions, and appropriate timing of discharge²⁵. Our study along with the aforementioned studies, indicates that addressing HRSN is essential for reducing hospital readmission, therefore controlling healthcare expenditures in this population. Housing instability, unemployment, and family/social instability are key factors contributing high readmission for PLWH and underscore the necessity of incorporating social needs into the care of PLWH.

At index hospitalization, PLWH without HRSN have a significantly lower prevalence of several comorbidities compared with those with HRSN, including alcohol use disorder, drug use disorder, and psychoses (Table 2, Supp. Table 2). In the multivariable models for readmission, HRSN is the predictor with highest odds ratio for readmission, aligning with our study hypothesis.

Compared with those with private insurance, PLWH insured by Medicare or Medicaid had higher odds of being readmitted. This may reflect, in part, the greater complexity of their health and health-related social needs, and the critical role these programs play in providing access to healthcare and support services for millions of people. Surprisingly, however, the symptomatic status (i.e., presenting with AIDS-defining cancer or opportunistic infection) at index admission was weakly associated with high readmission rate, reflecting both the effectiveness of HAART therapy on HIV syndrome, and HRSN as critical drivers of serious illness requiring hospitalization. For race/ethnicity, the odds for 90-day readmission [aOR = 0.99(0.94–1.05)] and one-year readmission [aOR = 1.11(1.05–1.17)] for Black PLWH in Florida was comparable to or slightly higher than that of White counterparts. In contrast, Black PLWH in Maryland had lower odds of readmission than White PLWH, both for 90-day readmission [aOR = 0.85 (0.75–0.95)] and one-year readmission [aOR = 0.87 (0.77–0.98)]. These differences may be attributed to diverging health policy environments with Maryland (but not Florida) having expanded Medicaid services.

In both Florida and Maryland, housing needs were the most prevalent HRSN in PLWH, followed by unemployment and family/social needs (Table 1). Several previous studies have found that homeless individuals and those facing housing instability, coupled with food insecurity, are more likely to have lower CD4 (T-cell) counts, poorer medication adherence, and incomplete suppression of HIV replication^26,27,28. More recent work has shown that food insecurity was also associated with limited access to health care and inconsistent care, and lower quality of life^29,30,31. Our findings, combined with those reported in the literature support the notion that these social risks significantly impact PLWH’s ability to manage their HIV treatment, subsequently contributing to the high risk of hospital readmission.

The high readmission of PLWH with HRSN implies that PLWH with documented social needs face specific HIV management challenges, likely related to both difficulties in adhering to HIV medication and the added physical and emotional stress caused by these social needs. Stress is increasingly recognized as a driver of disease, and day-to-day experiences of homelessness, lack of employment, and social isolation are exceptionally stressful. This highlights the importance of addressing the HRSNs in PLWH to improve health outcomes and reduce healthcare costs associated with frequent hospital readmissions, and point to the necessity of expanded federal programs and other resources to address these drivers of morbidity and increased healthcare costs.

Federal programs such as Housing Opportunities for People with AIDS (HOPWA), the Ryan White Program, and the Ending the HIV Epidemic (EHE) initiative play vital role in supporting access to diagnosis and management for population in regions with highest prevalence of disease. HOPWA provides essential housing assistance to PLWH, while the EHE initiative funds innovative research to address inequities in access to HIV diagnosis and treatment, focusing on high-burden area such as Maryland and Florida. By the end of 2021, the National HIV/AIDS Strategy (NHAS) had achieved notable progress in key areas, such as improving the quality of life and health outcomes for PLWH and addressing critical challenges like food insecurity, unemployment, and unstable housing or homelessness among this population³². However, despite these achievements, disparities in HIV diagnosis and treatment persist at both population and regional levels. These ongoing challenges emphasize the need for sustained and collaborative efforts by scientists and public health authorities to ensure that PLWH have equitable access to the resources and support required to effectively manage their health.

In syndemics theory, the adverse health outcomes are hypothesized as the co-occurrence of social factors that interact with and mutually influence each other³³. Poverty and social marginalization can influence co-occurrence of HIV³⁴, poor mental health, violence, and drug use, interacting in ways that adversely affect health outcomes in structurally marginalized communities^35,36 Our work affirms that syndemic analysis can uncover the complex causes of adverse outcomes in PLWH. Our study adds hospital readmission as additional evidence that teams caring for PLWH should have sufficient resources to support patients to address HRSN to ultimately reduce morbidity, mortality, and healthcare costs for PLWH.

One of the major strengths of our study is our analysis of hospital readmission from two independent states, Florida and Maryland, with contemporary data from 2016 to 2019, but with differing policy environments. The cohort size from Florida is about 4-times larger than that of Maryland, and we had comparable results from separate analyses of the two cohorts. The data from multivariable models provided the strongest evidence that HRSN are key factors associated with hospital readmission for PLWH.

We note the following limitations. First, HRSN coding varies substantially by clinician, location, hospital policy, and general clinical practice, and the prevalence of HRSN among PLWH is likely much higher than that reported in this study. Second, we did not have access to clinical data such as CD4 cell count and disease stage, which impacts disease severity and classification, and is often not translated into coding practice in our clinical experience. Third, our analysis was limited to PLWH from two states. However, it is likely the finding from these two states reflects the trend at the national level. Fourth, the identification of HIV-positive cases may be subject to potential misclassification. As previously reported,^37,38 relying on a single occurrence of an ICD-10 code to identify HIV cases can be less accurate than using multiple occurrences. potentially leading to reduced accuracy in identifying PLWH.

In conclusion, our study highlights a critical role of social needs on hospital readmission among PLWH. The high prevalence of social factors affecting PLWH is currently a central discussion in HIV care, and the combination of these social factors and comorbidities that PLWH encounter have a synergistic effect, which may explain the higher readmission rates. It is important to train clinicians to understand the impact of HRSN on health and to give them the direct financial, staffing, and wraparound support to effectively address these needs, along with adequate time to do so. Administrative structures must create systems that correctly and efficiently describe the HRSN of patients, rather than tasking individual clinicians with an endless search for coding that connects to patient reality, and to provide with the resources and support necessary to effectively address these needs. The federal programs have provided a roadmap for addressing the social factors of HIV care. To support these efforts, healthcare systems must implement administrative structures that accurately and efficiently document the HRSNs of patients. The ultimate goal is that healthcare providers, social workers, and community organizations have the support to collaboratively address HRSN in PLWH to improve health outcomes.