Coronavirus disease-2019 (COVID-19) is impacting access to medical care and social services, potentially worsening HIV outcomes among vulnerable populations . Beginning 17 March 2020, San Francisco implemented a ‘shelter-in-place’ public health order and required delay or shift to telemedicine for nonessential medical care in response to the COVID-19 pandemic [2,3]. The pandemic's impact on access to medical care , housing resources and social services  combined with challenges of those with unstable housing engaging with telemedicine  may disproportionately affect people living with HIV who experience homelessness .
Independent of barriers to care introduced by COVID-19, homelessness and unstable housing (HUH) substantially worsen viral suppression for people living with HIV (PWH) [7–10]. In January 2019, Ward 86, a large publicly funded HIV clinic in San Francisco, initiated POP-UP, a low-barrier, multicomponent model of care to address barriers to care for PWH experiencing homelessness . Throughout the COVID-19 pandemic, POP-UP remained open to drop-in, in-person visits because of challenges patients faced with telehealth. We hypothesized that care engagement and HIV viral suppression would not worsen during the COVID-19 pandemic among PWH experiencing HUH enrolled in this low-barrier, clinic-based supportive care model.
POP-UP is based at the Ward 86 HIV clinic at San Francisco General Hospital. Ward 86 provides care for approximately 2400 publicly insured patients, one-third of whom experience HUH . Eligibility criteria for POP-UP include: current homelessness or unstable housing, current unsuppressed HIV viral load (≥200 copies/ml), and at least one missed primary care visit and at least two unscheduled drop-in visits in the prior 12 months.
The POP-UP intervention includes drop-in (i.e. no appointment needed), incentivized HIV primary care and behavioral health services, on-site medication pick-up, and a care navigator to conduct patient outreach by phone or physically in the community. During shelter-in-place, POP-UP visits were deemed ‘essential’ and the program remained fully open for in-person clinic visits.
We conducted an interrupted time series analysis, comparing care engagement and viral suppression during monthly intervals in the 5 months before and after San Francisco's shelter-in-place order (hereafter pre-COVID 10/17/2019 to 3/16/2020, post-COVID 3/17/2020 to 8/16/2020). Patients enrolled in POP-UP for at least 1 month during this period were included in analysis. As enrollees, by definition, had a visit and were not virally suppressed at POP-UP enrollment, we included newly enrolled patients in this analysis beginning the month after initial enrollment. Care engagement was defined as a clinic visit within each 1-month period. Viral suppression was defined as most recent HIV viral load less than 200 copies/ml within the past 3 months; patients were otherwise considered unsuppressed. We also conducted sensitivity analysis including only viral loads measured during each monthly interval. Mixed effects logistic regression models compared outcomes over each 1-month period, pre-COVID and post-COVID, accounting for clustering by individual. Odds ratios were reported comparing the ‘level-change’ from the last month of the pre-COVID period to the first month of the post-COVID period .
Eighty-five patients were included, 44 who enrolled before 17 October 2019, 27 who enrolled pre-COVID, and 14 who enrolled during the post-COVID period. Thirteen patients discontinued POP-UP during follow-up, five pre-COVID and eight post-COVID. Unenrollment reasons include transfer to another provider/clinic (n = 7), moving out of San Francisco (n = 3), and death (n = 3). Patient characteristics include median age 40 years (IQR 26--58), 52% nonwhite, 85% cisgender men, 53% experiencing street homelessness (with others unstably or temporarily housed). Additionally, 98% had a substance use disorder, among whom 89% used methamphetamines.
The mean number of visits per patient-month was 1.6 pre-COVID (range 0--9) and 1.7 post-COVID (range 0--13). The proportion of POP-UP patients with a visit each month remained similar during the COVID-19 pandemic [64% pre-COVID vs. 58% post-COVID, odds ratio (OR) 0.66, 95% CI 0.39--1.11]. Viral suppression remained similar during the COVID-19 pandemic (48% pre-COVID, 47% post-COVID, OR 1.19, 95% CI 0.0.78–1.82, Fig. 1). Limiting analysis to only viral loads measured during each monthly interval yielded similar results.
Despite concerns that COVID-19 might worsen outcomes, particularly among people experiencing homelessness, care engagement, and viral suppression did not worsen among 85 patients enrolled in a supportive, low-barrier care model for unstably housed patients in an urban HIV clinic. Continued access to comprehensive, in-person medical and social services during the pandemic likely contributed to maintenance of HIV care outcomes during this period. In addition, 15% of POP-UP patients were temporarily housed in shelter-in-place hotels funded by San Francisco during COVID-19, providing shelter and a location where the POP-UP navigator could conduct phone-based and in-person outreach.
Our results are notable given that there were 31% higher odds of viral nonsuppression post-COVID among patients accessing traditional primary care services at Ward 86, with further worsening among people experiencing homelessness . This is noteworthy as people experiencing homelessness have faced additional structural barriers and competing needs during COVID-19, including disrupted access to shelter and decreased access to substance use and mental health treatment [5,13]. By minimizing barriers to accessing medical care during this critical time, POP-UP's drop-in, incentivized, comprehensive services may have been protective against the impact of additional structural barriers introduced by COVID-19. However, although virologic suppression rates did not drop further during COVID-19 for patients in the POP-UP program, viral suppression in this population remains inadequate and reflects ongoing need for additional interventions.
Limitations of our study include the nonrandomized prepost design and the limited sample size. We utilized an interrupted time series analysis to help isolate the effects of shelter-in-place implementation from other temporal trends. We also used an open cohort design to reduce the likelihood that improvements in viral suppression were simply because of greater duration of engagement with the POP-UP intervention. Though we did not observe large changes in outcomes during the pandemic, the limited sample size may have reduced our ability to detect smaller changes.
In conclusion, care engagement and viral suppression did not worsen during COVID-19 among people experiencing homelessness engaged in a supportive program to provide low-barrier, comprehensive HIV primary care, in contrast to worsening viral suppression rates observed among patients accessing traditional care models. Multicomponent interventions, such as POP-UP may help maintain viral suppression despite introduction of unanticipated structural barriers, such as those associated with COVID-19.
Author contributions: M.D.H. and E.I. conceived the analysis. M.D.H. wrote the first draft of the article. M.D.H. and D.V.G. performed the statistical analysis. J.B.D.R. and M.C. conducted data collection and maintained the clinical database. A.C.A and J.O. contributed to interpretation of results. M.G., E.R., and D.V.H. contributed to interpretation of results and provided supervision of clinical research activities. All authors reviewed the manuscript and provided critical edits. We would like to thank David Fischer and Janet Q. Nguyen for assistance with obtaining electronic medical record data and the POP-UP clinical team for supporting implementation of POP-UP.
Funding source: The ‘Ward 86’ HIV program in the Division of HIV, ID and Global Medicine received an unrestricted investigator-initiated grant from the Gilead Foundation to support implementation and evaluation of the ‘POP-UP’ program, a clinical program for people with HIV experiencing homelessness or unstable housing (Grant # IN-US-985-5691). Gilead had no role in the interpretation or presentation of these results. This research was also supported by an Ending the HIV Epidemic Supplemental grant from the National Institutes of Health to the UCSF-Gladstone Center for AIDS Research P30 AI027763 (M.G.) and a U.S. National Institute of Allergy and Infectious Diseases (NIAID) training grant 5T32AI060530-13 (M.D.H.).
Conflicts of interest
There are no conflicts of interest.
1. Spinelli M, Hickey M, Glidden D, Nguyen JQ, Oskarsson JJ, Havlir D, Gandhi M. Viral suppression rates in a safety-net HIV clinic in San Francisco destabilized during COVID-19
2. Order of the Health Officer No. C19-07 (Directing All Individuals Living in the County to Shelter at Their Place of Residence). San Francisco Department of Public Health; 2020. Available at: https://www.sfdph.org/dph/alerts/files/HealthOrderC19-07-%20Shelter-in-Place.pdf
. (Accessed 2 April 2020)
3. Order of the Health Officer No. C19-08 (Moratorium on Routine Medical Appointments and Elective Survery, Encouraging Remote Appointments). San Francisco Department of Public Health; 2020. Available at: https://www.sfdph.org/dph/alerts/files/Order-C19-08-ProhibitingElectiveSurgery.pdf
. (Accessed 18 May 2020)
4. Mehrotra A, Chernew M, Linetsky D, Hatch H, Cutler D. The impact of the COVID-19 pandemic on outpatient visits: a rebound emerges. The Commonwealth Fund; 2020. Available at: https://www.commonwealthfund.org/publications/2020/apr/impact-covid-19-outpatient-visits
. (Accessed 13 July 2020)
5. Imbert E, Kinley PM, Scarborough A, Cawley C, Sankaran M, Cox SN, et al. SF COVID-19 Response Team. Coronavirus disease 2019 (COVID-19) outbreak in a San Francisco homeless shelter
. Clin Infect Dis
6. Rhoades H, Wenzel S, Rice E, Winetrobe H, Henwood B. No digital divide? Technology use among homeless adults
. J Soc Distress Homeless
7. Doshi RK, Milberg J, Jumento T, Matthews T, Dempsey A, Cheever LW. For many served by the Ryan White HIV/AIDS Program, disparities in viral suppression decreased, 2010-14
. Health Aff (Millwood)
8. Griffin A, Dempsey A, Cousino W, Avery L, Phillips H, Egwim E, Cheever L. Addressing disparities in the health of persons with HIV attributable to unstable housing in the United States: the role of the Ryan White HIV/AIDS Program
. PLOS Med
9. Clemenzi-Allen A, Geng E, Christopoulos K, Hammer H, Buchbinder S, Havlir D, Gandhi M. Degree of housing instability shows independent ‘dose-response’ with virologic suppression rates among people living with human immunodeficiency virus
. Open Forum Infect Dis
10. Riley ED, Vittinghoff E, Koss CA, Christopoulos KA, Clemenzi-Allen A, Dilworth SE, Carrico AW. Housing first: unsuppressed viral load among women living with HIV in San Francisco
. AIDS Behav
11. Imbert E, Hickey MD, Clemenzi-Allen A, Conte M, Riley ED, Havlir DV, Gandhi M. POP-UP clinic: a multicomponent model of care for people living with HIV (PLHIV) who experience homelessness or unstable housing (HUH). Oral Presentation, Abstract OAEO4 presented at the: AIDS 2020 Virtual; 6 July 2020.
12. Bernal JL, Cummins S, Gasparrini A. Interrupted time series regression for the evaluation of public health interventions: a tutorial
. Int J Epidemiol
13. Okonkwo NE, Aguwa UT, Jang M, Barré IA, Page KR, Sullivan PS, et al. COVID-19 and the US response: accelerating health inequities
. BMJ Evid Based Med
2020; Published Online First: 03 June 2020. doi:10.1136/bmjebm-2020-111426.