Estimating the Impact of Changing Out-of-Pocket Costs of Abandonment of HIV Pre-exposure Prophylaxis (PrEP) Prescription

Abstract

Oral HIV pre-exposure prophylaxis (PrEP) is highly effective for preventing HIV. Several different developments in the U.S. threaten to increase or promise to decrease PrEP out-of-pocket costs and access in the coming years. In a sample of 58,529 people with a new insurer-approved PrEP prescription, we estimated risk-adjusted percentages of patients who abandoned (did not fill) their initial prescription, across six out-of-pocket cost categories. We then simulated the percentage of patients who would abandon PrEP under hypothetical changes to out-of-pocket costs, ranging from $0 to >$500. PrEP abandonment rates of 5.5% at $0 rose to 42.6% at >$500; even a small increase from $0 to $10 doubled abandonment. Conversely, abandonment rates of 48% at >$500 dropped to 7.3% at $0. HIV diagnoses were 2-3 times higher among patients who abandoned than those who filled PrEP. Results imply that recent legal challenges to PrEP with no cost-sharing could substantially increase PrEP abandonment and HIV rates, upending progress on the HIV/AIDS epidemic.

INTRODUCTION

Oral HIV pre-exposure prophylaxis (PrEP) is highly effective at reducing the risk of HIV (1), but PrEP uptake, adherence, and persistence remain suboptimal in the United States.(2, 3) The US Centers for Disease Control and Prevention (CDC) estimate that 1.2 million adults in the US have an indication for PrEP; however, fewer than one-third of that number received a PrEP prescription in 2021.(4) PrEP uptake is an ongoing challenge despite most private and state Medicaid plans covering PrEP(5) and an increasing number of private health insurers having expanded coverage for PrEP. PrEP out-of-pocket costs vary widely and even patients with prescription drug coverage for PrEP through their insurance plans could face high out-of-pocket costs for PrEP medications and its associated medical visits and laboratory testing.(6) In fact, patient out-of-pocket costs due to insurance cost-sharing requirements such as deductibles, coinsurance, and/or copayments have been cited as one of the major barriers to accessing PrEP.(2, 7–11)

Several different developments in the U.S. threaten to increase or promise to decrease PrEP out-of-pocket costs in the coming years. The United States Preventive Services Task Force (USPSTF) gave oral HIV PrEP an “A” rating, recommending that starting in 2021, all private insurers (except those maintaining “grandfathered” status) and Medicaid expansion programs comply with the Affordable Care Act (ACA) requirement for payment of preventive services, noting that all PrEP medications and related services should be covered with no patient cost-sharing.(12) However, in March 2023, a federal district court judge issued a ruling in Braidwood Management, Inc. v. Becerra which struck down the no patient cost-sharing mandate for preventive services that receive a USPSTF grade of A or B.(13, 14) This court ruling jeopardizes all cost-free preventive care under the ACA including “first dollar coverage” (i.e., $0 prescriptions) for PrEP and could threaten patient access to and use of PrEP, possibly increasing HIV incidence. While the federal government subsequently appealed this decision, and an administrative stay has been issued, concerns remain about PrEP out-of-pocket costs potentially increasing for privately insured individuals in the future until the court case unfolds.

At the same time, many Medicare beneficiaries are likely to see reductions in their PrEP out-of-pocket costs with the implementation of a $2,000 annual Part D out-of-pocket maximum in combination with monthly caps to distribute these costs more evenly throughout the year starting January 1, 2025.(15, 16) In the meantime, a legislative proposal has also been introduced in Congress to specifically reduce PrEP out-of-pocket costs by mandating “first dollar coverage” of PrEP prescriptions across all private and public insurers.(17)

Given these ongoing developments that are likely to increase or decrease PrEP out-of-pocket costs, it is important to understand their potential implications based on the extent to which out-of-pocket costs influence PrEP uptake and concomitant HIV incidence. Only one study to date has shown that patients who paid >$500 for a 30-day supply of PrEP had four times greater risk of abandoning PrEP than those who paid <$50.(11) Yet little quantitative evidence exists on PrEP abandonment rates if out-of-pocket costs were to increase or decrease. In this study we address this knowledge gap by first estimating the risk-adjusted association of varying levels of out-of-pocket costs with abandonment (i.e., not obtaining a fill) of a newly prescribed, insurer-approved HIV PrEP prescription. We then examined differences in the risk-adjusted rate of a new HIV diagnosis among those who abandoned PrEP vs. those who did not, across out-of-pocket cost categories. Finally, we use the risk-adjusted estimates to simulate how PrEP abandonment rates might change if patient out-of-pocket costs were hypothetically increased or decreased, both for all patients representing a range of insurance types, and for a subset of patients who are commercially insured.

METHODS

Data Source

This study used 2015-2019 data from the Symphony Health Solutions Integrated DataVerse by Source Healthcare Analytics, LLC (SHA).(18) This large, proprietary database has been used in previous studies across multiple disease areas, including HIV PrEP,(19–24) and contains medical and pharmacy claims for most patients (>274 million) in the US. SHA obtains individual-level patient data from the National Council for Prescription Drug Programs, covering 80% to 85% of all HIV PrEP prescription claims in the US.(25) Prescription claims cover the complete lifecycle of a prescription from initial adjudication (decision by insurance to approve and pay for the prescription) through final claim payment status (paid or reversed claim) and include information on the patient’s out-of-pocket liability (after application of coupons or copay assistance). Data on prescriptions were available from all pharmacy types (e.g., retail, mail-order) regardless of whether they were adjudicated and/or paid by commercial insurers, Medicaid, Medicare Part D, or point-of-sale cash paid by patients out-of-pocket. This study was deemed exempt by the John Hopkins University Institutional Review Board given the use of secondary data that were not collected directly from patients.

Study Sample

The sampling frame included individuals with an insurer-approved PrEP prescription (tenofovir disoproxil fumarate/emtricitabine; TDF/FTC) during the observation window of September 30, 2016, to June 2, 2018. Patients on tenofovir alafenamide/emtricitabine (TAF/FTC) were not included because its formulation had not yet been FDA-approved for PrEP during our observation window. Patients with an eligible claim for PrEP were identified through a multi-step process (Appendix Exhibit A1(26))based on our previously published algorithm.(21) To differentiate use of TDF/FTC for HIV PrEP versus for treatment of HIV, post-exposure prophylaxis (PEP), or Hepatitis B Virus (HBV), we only included persons who: had a claim of TDF/FTC with a supply of >28 to 91 days; had no ICD9/10 diagnosis codes for HIV or HBV, and had no claims for HIV medications during the 365 days before or 30 days after the index prescription (based on the CDC definition(27)). For persons with multiple PrEP claims with conflicting status on the same date, we assigned one final claim status, prioritizing approved claim status over reversed claim status.

The date of the first eligible claim for an insurer-approved PrEP prescription during the sample identification window was deemed as the patient’s index date. Claims were included in the analysis when patients met the following additional selection criteria: (i) evidence of at least one pharmacy claim in the 365 days pre- and post-index date, as a marker for database inclusion during the study periods of interest, (ii) no evidence of HIV diagnosis or PrEP pharmacy claims in the 365 day pre-index period to ensure the index prescription was a new PrEP prescription, (iii) no evidence of >1 PrEP prescription on the index date, (iv) no missing values for PrEP out-of-pocket cost or patient demographics (sex and state of residence), and (v) age >=18 years old.

Outcome Measures

Our key outcome of interest was PrEP prescription abandonment by the patient. Each insurer-approved PrEP claim was classified as filled (i.e., prescription was purchased by the patient) or reversed (i.e., prescription was not obtained by the patient and the claim was withdrawn by the pharmacy). Patients with a reversed PrEP claim with no evidence of pick-up within 365 days of the index claim were classified as having abandoned their PrEP prescription; the remainder of the patients were classified as having filled their PrEP prescription.

As secondary outcomes, we further segmented the “filled” group into (1) those who had filled initially (PrEP prescription was paid for and obtained by the patient within 14 days of approval); (2) those who had a delayed fill (PrEP prescription that was that was not picked up within the first 14 days of approval (reversed) but paid for and obtained within 90 days); and (3) those who had a very delayed fill (PrEP prescription that was that was not picked up within the first 14 days of approval (reversed) but later paid for and obtained between 91 and 365 days).

Another key outcome measure of interest was a new HIV diagnosis identified based on the presence of ≥1 medical claim with an ICD-10 code indicating HIV infection and/or the presence of ≥1 paid prescription claim for HIV treatment during the 365-day post-index follow-up.

PrEP Out-of-pocket Costs

Patients were assigned to one of six mutually exclusive out-of-pocket cost categories based on the final out-of-pocket cost (after application of coupons or copay assistance) of their index PrEP prescription claim standardized to 30 days of supply: $0, >$0 to ≤$10, >$10 to ≤$25, >$25 to ≤$100, >$100 to ≤$500, >$500. The cut-points for the categories were informed by previous work,(19), frequency distributions of the data, meaningful designations based on input from pharmacist co-authors, and typical cost sharing levels under Medicaid and Medicare.

Analysis

Descriptive characteristics were examined for the study sample, overall and by PrEP abandonment status. Multinomial logistic regression models were used to estimate adjusted rates of HIV PrEP prescription abandonment vs. paid fills (filled initially, delayed fill, very delayed fill) by the six out-of-pocket cost categories.

In addition, two logistic regression models were used to estimate the adjusted rates for the binary outcomes of (a) abandonment of the index PrEP prescription and (b) having any HIV diagnosis or medication within one year after the index date, by the six out-of-pocket cost categories. Covariates in all the models included patient age, sex, race and ethnicity, household income, education level, US Census region, Charlson Comorbidity Index, year of the index PrEP prescription (to control for any temporal trends), and insurance type.

Finally, we conducted a simulation (19) to explore how rates of PrEP prescription abandonment might change at different levels of out-of-pocket costs. Using the coefficients from our regression analyses, we calculated the predicted rates of PrEP abandonment in hypothetical scenarios wherein patients were shifted from their current out-of-pocket cost category to a hypothetically higher or lower out-of-pocket cost category, while holding their sociodemographic and clinical characteristics constant. For example, we report the predicted prescription abandonment rate if a patient currently paying $0 for their PrEP prescription in our data were to face a sudden increase in out-of-pocket costs (e.g., >$500). Given the recent decision in Braidwood v. Becerra, which will no longer require plans to cover preventive care such as PrEP at no cost to patients, we additionally simulated how the prescription abandonment would change specifically in the subgroup of commercially insured patients, since they might be most impacted by the ruling. Because our data spanned the period prior to the ACA mandated coverage of PrEP at no cost-sharing starting 2021, we had sufficient variation in the out-of-pocket costs in our commercially insured subgroup to be able to generate estimates to simulate how increases in out-of-pocket cost would impact prescription abandonment in this population.

Limitations

This study had several limitations. First, there is a possibility for misclassification of a PrEP prescription because the same drug formulations used for PrEP may also be used for post-exposure prophylaxis, or HIV and Hepatitis B treatment. We adapted the CDC algorithm for identifying PrEP prescriptions,(27) but another algorithm may have yielded a slightly different sample. Second, the database used includes the majority (80%-85%) of but not all (100%) of the PrEP prescriptions in the U.S.; however, its population is nationally representative of the US with respect to age, insurance, sex, and geography, and includes the vast majority of prescriptions, across all payer types, excluding only integrated delivery networks who do not make their data available. Third, there was a possibility of overestimating abandonment, which could instead have been patients moving to a pharmacy that is not captured in the dataset. We minimized the risk of this misclassification in our sample selection stage by requiring evidence of pharmacy activity in the pre- and post-period surrounding the index PrEP claim. Furthermore, the fact that we found a higher rate of new HIV diagnoses among those who abandoned vs. filled PrEP suggests that patients who are classified as abandoning PrEP are likely not getting PrEP from any other sources. Fourth, claims data may underestimate HIV cases if patients got tested for HIV outside their insurance plan (e.g., home testing) or at a health system that does not contribute claims to the dataset; however, there is no reason to believe that such occurrences will be systematically different by PrEP abandonment status. Hence, it should not affect our estimated differences in HIV incidence rates between those who abandoned vs. filled PrEP. Finally, our simulation results were based on risk-adjusted estimates obtained from cross-sectional comparisons across out-of-pocket cost categories and cannot be considered to denote causality. Nevertheless, they provide important insights into potential implications of increasing or decreasing PrEP out-of-pocket costs.

RESULTS

The final sample contained 58,529 patients with a new insurer-approved HIV PrEP prescription. Of these patients, 13.5% abandoned their index prescription(n=7,926). Nearly half the sample was less than 35 years old and 88.8% of patients were male. Insurance types included commercial (78.0%), Medicaid (14.3%), and Medicare (4.2%); 3.5% of patients paid cash. Overall, 59.2% of the patients had out-of-pocket costs of ≤$10 and 11.4% of patients had out-of-pocket costs of >$100 for a 30-day prescription. Additional sample characteristics, overall and by abandonment status, are provided in Appendix Exhibit A2.(26)

Risk-adjusted percentages of prescription abandonment by out-of-pocket cost category are displayed in Exhibit 1. Risk-adjusted rates of PrEP abandonment increased as the out-of-pocket cost category increased. Risk-adjusted abandonment rates ranged from 5.5% for patients facing $0 cost share to 42.6% for patients responsible for >$500. The proportion of delayed and very delayed prescription fills also increased as the out-of-pocket cost increased. Among patients with no out-of-pocket cost, only 2% of patients had evidence of a delayed fill compared with 6.9% of the patients whose prescriptions were in the highest out-of-pocket cost category.

EXHIBIT 1.

Adjusted Rates of Abandonment vs. Filled (Filled Initially, Delayed Fill, or Very Delayed Fill) Index PrEP prescriptions (95% CI) by Out-of-Pocket Cost Category

Source/Notes: SOURCE Authors’ analysis of data from Symphony Health Solutions Integrated DataVerse, 2015-2019. NOTES Abandoned signifies index PrEP prescriptions that were not picked up within 365 days of the index date. Filled initially signifies PrEP prescription was paid for and obtained by the patient within 14 days of approval. Delayed fill signifies PrEP prescription that was not picked up within the first 14 days of approval (reversed) but later paid for and obtained within 90 days. Very delayed fill signifies PrEP prescription that was that was not picked up within the first 14 days of approval (reversed) but later paid for and obtained between 91 and 365 days. Models adjusted for patient age, sex, race and ethnicity, household income, education level, US Census region, Charlson Comorbidity Index, year of the index PrEP prescription, and insurance type. Due to rounding, not all column sets sum to exactly 100%.

Risk-adjusted rates of a new HIV diagnosis by PrEP abandonment status and out-of-pocket cost category are displayed in Exhibit 2. Across all out-of-pocket cost categories, the HIV incidence was higher among patients who abandoned their PrEP prescription relative to those who filled their PrEP prescription: 6.1% vs. 2.1% ($0), 4.9% vs. 2.2% (>$0 to ≤$10), 3.9% vs. 1.9% (>$10 to ≤$25), 3.6% vs. 2.1% (>$25 to ≤ $100), 4.4% vs. 2.2% (>$100 to ≤$500), and 4.6% vs. 2.5% (>$500).

EXHIBIT 2.

Adjusted Rates of HIV Incidence (95% CI) for Abandoned or Filled Prescriptions by Out-of-Pocket Cost Category

Source/Notes: SOURCE Authors’ analysis of data from Symphony Health Solutions Integrated DataVerse, 2015–2019. NOTES Abandoned signifies index PrEP prescriptions that were not picked up within 365 days of the index date. Filled initially signifies PrEP prescription was paid for and obtained by the patient within 14 days of approval. Models adjusted for patient age, sex, race and ethnicity, household income, education level, US Census region, Charlson Comorbidity Index, year of the index PrEP prescription, and insurance type.

Predicted rates of PrEP prescription abandonment if out-of-pocket costs were to hypothetically increase or decrease to specified levels in our overall sample are displayed in Exhibit 3. For example, if patients in our sample who are currently paying $0 out-of-pocket for their PrEP prescription were to be hypothetically subject to even a small increase in costs, from >$0 to ≤$10, their abandonment rate would be expected to double, from an observed rate of 5.6% (95% CI: 5.1 – 6.0) to a predicted rate of 11.1% [95% CI: 10.7 – 11.5]). This is even more pronounced for patients currently facing out-of-pocket costs of >$25 to ≤$100, who had an observed abandonment rate of 11.1% (95% CI: 10.5 – 11.7). If these patients were to be subject to out-of-pocket costs of >$100 to ≤$500, their predicted abandonment rate would be expected to be thrice as high (31.9% [95% CI: 30.0 – 33.8]). Conversely, a reduction in out-of-pocket costs would likely result in meaningful reductions in the rate of prescription abandonment in most cases. For instance, nearly half of the patients (48.0% [95% CI: 46.6 – 49.5) who currently faced out-of-pocket costs >$500 were observed to abandon their prescription in our sample; if these costs were reduced to $0 (i.e., no cost-sharing for PrEP), their rate of prescription abandonment would be expected to fall to 7.3% (95% CI: 6.7 – 7.8).

EXHIBIT 3. Predicted Rates of PrEP Abandonment if Out-of-Pocket Costs were to Increase/Decrease to Specified Levels.

Source/Notes: SOURCE Authors’ analysis of data from Symphony Health Solutions Integrated DataVerse, 2015–2019. NOTES The blue highlighted sections show predicted risk-adjusted rates of PrEP abandonment if current out-of-pocket costs (shown in the rows) were to hypothetically increase to specified out-of-pocket cost category levels (shown in the columns). The green highlighted sections show predicted risk-adjusted rates of PrEP abandonment if current out-of-pocket costs were to hypothetically decrease to specified out-of-pocket cost category levels. Models adjusted for patient age, sex, race and ethnicity, household income, education level, US Census region, Charlson Comorbidity Index, year of the index PrEP prescription, and insurance type.

		Predicted PrEP Abandonment Rate (95% CI) by Hypothetical Out-of-Pocket Cost Category
Current Out-of-Pocket Cost Category	Patients (N)	$0	> $0 to ≤ $10	>$10 to ≤ $25	>$25 to ≤ $100	> $100 to ≤ $500	>$500
$0	14,155	5.6 (5.2 - 6.0)	11.1 (10.7 - 11.5)	12.9 (12.1 - 13.7)	12.8 (12.1 - 13.6)	34.7 (32.7 - 36.6)	42.9 (41.2 - 44.6)
> $0 to ≤ $10	20,505	5.7 (5.4 - 6.1)	11.3 (10.9 - 11.7)	13.1 (12.3 - 13.9)	13.0 (12.3 - 13.7)	34.9 (33.0 - 36.8)	43.1 (41.5 - 44.8)
> $10 to ≤ $25	7,437	4.8 (4.5 - 5.2)	9.8 (9.4 - 10.2)	11.4 (10.7 - 12.1)	11.3 (10.7 - 12.0)	32.3 (30.4 - 34.2)	40.5 (38.9 - 42.2)
> $25 to ≤ $100	9,792	4.7 (4.4 - 5.1)	9.6 (9.1 - 10.0)	11.1 (10.4 - 12.0)	11.1 (10.5 - 11.7)	31.9 (30.0 - 33.8)	40.1 (38.5 - 41.8)
> $100 to ≤ $500	2,282	5.6 (5.2 - 6.0)	11.1 (10.6 - 11.6)	12.9 (12.1 - 13.6)	12.8 (12.1 - 13.5)	34.5 (32.7 - 36.4)	42.7 (41.2 - 44.3)
>$500	4,358	7.3 (6.7 - 7.8)	14.0 (13.3 - 14.6)	16.0 (15.0 - 17.1)	16.0 (15.0 - 16.9)	39.8 (37.7 - 41.8)	48.0 (46.6 - 49.5)

Similar results on risk-adjusted rates of PrEP prescription abandonment were observed when limiting the sample to the 45,665 commercially insured patients (Appendix Exhibit A3 (26)). Exhibit 4 presents the predicted rates of PrEP abandonment as cost-sharing levels hypothetically increased among the subgroup of commercially insured patients currently facing $0 out-of-pocket for their PrEP prescription (n=9,096, 19.9%), who had an observed abandonment rate of 3.8% (95% CI: 3.8 – 4.1). Even a small hypothetical increase in the out-of-pocket costs for these patients (for instance, shifting from $0 to >$10 to ≤$25) could result in a tripling of the abandonment rate (11.5% [95% CI: 10.8 – 12.2). The hypothetical effects of increased cost-sharing are most pronounced among patients who would face out-of-pocket costs >$100 to ≤$500 (32.3% [95% CI: 30.4 – 34.3]) or >$500 (42.0% [95% CI: 40.2 – 43.8]), such as patients facing co-insurance and/or high-deductible health plans.

Exhibit 4.

Predicted Rates of PrEP Abandonment if Out-of-Pocket Costs were to Increase from Zero Dollars to Specified Levels for Patients who are Commercially Insured

Source/Notes: SOURCE Authors’ analysis of data from Symphony Health Solutions Integrated DataVerse, 2015–2019. NOTES Models adjusted for patient age, sex, race and ethnicity, household income, education level, US Census region, Charlson Comorbidity Index, year of the index PrEP prescription, and insurance type. See Appendix Exhibit A3 for predicted rates of PrEP abandonment for other current and hypothetical out-of-pocket cost categories.

DISCUSSION

Our analysis from over 58,000 newly prescribed PrEP patients in the U.S. found that higher out-of-pocket costs were associated with increased rate of patient abandonment of insurer-approved PrEP prescriptions. Further, within each out-of-pocket cost category, the rate of a new HIV diagnosis over a 1-year follow-up was nearly two- to three-fold higher among those who abandoned their index PrEP prescription relative to those who did not. Finally, our simulations showed that if high out-of-pocket costs are indeed a driving factor, PrEP abandonment rates for patients facing no cost-sharing are likely to double with relatively modest increases in patients’ out-of-pocket liabilities. Conversely, PrEP abandonment rates would be predicted to fall for patients who may face reductions in out-of-pocket costs.

Although our analysis cannot assert a causal link between PrEP costs and abandonments, our sample selection criteria were designed to maximize the ability to isolate the association of out-of-pocket costs with PrEP abandonment. We reduced the likelihood of patients not taking PrEP due to insurance restrictions (e.g., prior authorizations) by only studying prescriptions that were already insurer-approved. Similarly, we reduced the likelihood of patients not taking PrEP due to experience of side effects by focusing on people who had not previously prescribed PrEP. As people without a prior PrEP prescription, it is also unlikely that patients would be aware of the out-of-pocket costs of PrEP until they arrive at the pharmacy point-of-delivery, which is typically the first time that patients are confronted with the out-of-pocket cost of PrEP.(28) Hence, this rules out reverse causality due to patients self-selecting into insurance plans with lower out-of-pocket costs since they were already interested in filling a PrEP prescription before knowing the cost. Thus, our study still provides strong evidence of the likely role of cost changes on PrEP abandonment.

Collectively, these findings have implications for the recent policy developments and ongoing legal challenges that will shape future out-of-pocket costs for PrEP. While the 5^th Circuit Court of Appeals has issued a temporary administrative stay of the Braidwood decision, the possible end of the ACA’s no-cost coverage for preventive services such as PrEP still looms large for privately insured patients. Our results in the subgroup of commercially insured patients (i.e., the group that will be most impacted by the Braidwood decision) suggests that PrEP abandonment rates would likely increase even if out-of-pockets costs increase to as little as even $10 from the current $0 cost-sharing. However, PrEP abandonment rates may be substantially higher if out-of-pocket costs increase to >$100 given the ongoing trend towards higher deductibles and cost-sharing in the form of coinsurance (i.e., a percentage of the drug’s list price rather than a fixed copayment) and increasing enrollments in high-deductible health plans and coinsurance rather than copayment for specialty tier drugs.(29, 30) Our simulation suggests that around 1 in 3 commercially-insured patients currently receiving PrEP at no cost might abandon their prescriptions if their cost-sharing increased to between >$100 to ≤$500, and 42% may abandon if out-of-pocket costs increased to >$500. Given our findings of higher HIV diagnoses among patients who abandon PrEP compared to those who filled their prescriptions, increases in cost-sharing may culminate in increases in HIV cases. Although alternatives such as drug assistance programs and industry-sponsored safety net programs still exist, patient eligibility for many of those programs has become increasingly more stringent, and they may not fully cover out-of-pocket costs that patients experience.(6, 31)

While the federal government has appealed the Braidwood ruling, alternatives to maintain no-cost and low-cost access to PrEP are needed. For instance, Congress could pass the PrEP Access and Coverage Act (32) and mandate “first dollar coverage” across all sources of insurance coverage. Critically, this legislation would also remove cost-sharing for PrEP-related screenings, diagnostic procedures, and clinical follow-ups, a major barrier to PrEP access even when $0 cost-sharing was in effect.(33) Despite a promise that there will be “no immediate disruption in care or coverage(34)” as a result of the Braidwood decision, insurers have previously used various tactics such as limiting formularies, or listing PrEP drugs in the wrong tiers, to discourage PrEP use even with the ACA’s preventive coverage mandate in place.(35) In a guidance published shortly after the Braidwood decision, officials in the Departments of Labor, Health and Human Services, and the Treasury(36) were careful to note that state laws, like those in Delaware and New York,(29, 37) would not be impacted by the ruling, suggesting that this may be one possible route for ensuring continued coverage for preventive services such as PrEP.

Nevertheless, even when out-of-pocket costs were $0, 5.5% abandoned their PrEP prescription. This reinforces that out-of-pocket cost, while important, is not the only driving factor in PrEP initiation. While our sample selection criteria ruled out side effects and prior authorizations as reasons for not filling PrEP, other factors such as personal preferences, stigma, and transportation may also influence PrEP uptake (2, 3, 6, 7, 28) Hence, in addition to keeping PrEP at no cost, support around these other needs and concerns is also warranted. Policymakers should consider creation of a national PrEP program (38) (akin to the Ryan White HIV care program) that not only eliminates financial barriers for PrEP care but also offers navigation and other services for people at high risk for HIV to retain them in preventive care and remain HIV-negative.

In conclusion, our study results highlight the critical need for federal and state policymakers to identify strategies to ensure no-cost or low-cost access to PrEP with the goal to prevent widespread abandonment of PrEP medications and risk sacrificing the progress made in combating the HIV epidemic over the last 30 years in the U.S. Ongoing research and monitoring of upcoming policy developments and legal challenges will be needed to assess their impact on access and adherence to PrEP.