Structural Barriers to Continuity of HIV Care: Characterizing and Evaluating the Impact of Patient Discontinuation from AIDS Drug Assistance Programs

The Ryan White HIV/AIDS Drug Assistance programs (ADAPs) are the largest source of medical care for people living with HIV (PLWH) in the United States. They pay for insurance and medical care for 20% of people living with HIV in the United States, and people on the program have high rates of viral suppression, the central measure of successful HIV treatment. In spite of the benefits of the program, clients have difficulty staying enrolled in the program and subsequently lose benefits. Until recently, federal policy required that ADAP clients provide documentation of their eligibility every 6 months or be removed from the program. In October
of 2021, these restrictions were relaxed, and states were given the authority to set their own recertification procedure.
Although this policy change has the potential to reduce the burden of recertification, there is a lack of information about how recertification affects clients and the cost of the program. This dissertation sought to fill this knowledge gap by quantifying the effect of disenrollment on the ADAP clients and building a model to project the effect of extending the recertification timeline to every 12-months.
First, we used Ryan White data from Washington State to describe the prevalence of ADAP disenrollment and identify factors associated with being removed from the program. We categorized all PLWH enrolled in ADAP clients by the success of their recertification applications as continuously enrolled, ruled ineligible, disenrolled if they failed to recertify. We compared individuals who were disenrolled to those who were continuously enrolled by demographic and socioeconomic characteristics and use of case management services. Next, we sought to quantify the impact of disenrollment from ADAP on viral suppression. Using this same population, we estimated the risk difference of viral suppression before and after enrollment using clients who were continuously enrolled as a comparator. We used quantitative bias analysis to identify how much of the effect of client disenrollment could be attributed to other unmeasured confounders. Lastly, we used the results of the first two aims to develop a Markov model to analyze the cost and health impact of changing the existing 6-month recertification schedule to a 12-month schedule. We predicted the change in annual program costs, program enrollment, and population viral suppression over a 5-year time horizon.
We found that disenrollment is common and disproportionately affects marginalized populations in Washington State. Over the two-year study, 26% of clients were disenrolled from the program at least once due to failure to recertify, which is much greater than the 18% of clients who were removed due to ineligibility. Compared to those who were continuously enrolled, disenrolled PLWH were more likely to be Black (prevalence ratio vs White 1.31, 95% CI 1.17-1.46), uninsured (PR vs private insurance 1.24, 95% CI 1.10-1.40), and younger (PR 25-34 vs 35-44 years 1.23 95% CI 1.08-1.41). We also found that disenrollment negatively impacts the viral suppression of PLWH who are removed from the program. Of the 1336 ADAP clients who were disenrolled, 83% were virally suppressed before disenrollment versus 69% after (RD 12%, 95%CI 9-15%). Our quantitative bias analysis suggested that unmeasured confounders are unlikely to explain the entirety of this effect. Our budget impact analysis suggested that a 12-month recertification policy would yield a program that costs 7% more per year ($40.2M vs $37.7M, 95% CI 6-8%), but produces greater health benefits (245 more individuals virally suppressed by the end of 2025).
The results of this dissertation demonstrate that the current ADAP recertification policies, which were formerly required by federal policy, are disruptive to the health of a large proportion of ADAP clients. In Washington state, a change to a 12-month recertification policy has the potential to reduce the number of virally unsuppressed PLWH by 10% at a modest cost relative to the overall cost of the program.