Constraining access to HIV regimens can have significant implications for patients. This study examined the economic and health impacts of restrictive HIV formulary designs.
ABSTRACTObjectives: To model the impacts of restrictive formulary designs on outcomes for patients with HIV and to demonstrate the costs of restricting access to novel HIV regimens with better safety and efficacy profiles.
Study Design: We modified an epidemiological model of HIV incidence, progression, and treatment to simulate the effects of 5 formulary scenarios on patient outcomes in the United States.
Methods: Using a cohort of HIV-susceptible individuals, we followed patients through HIV infection, disease progression, and death. Patients transitioned in and out of treatment states once infected. Treatment discontinuation, efficacy, and the rate of adverse events (AEs; renal failure and bone fracture) in each formulary scenario depended on the treatment path and regimens included. Outcomes of interest included all-cause cumulative deaths, annual rates of AEs, and costs associated with treating those AEs.
Results: All outcomes of interest were more favorable in less restrictive formulary scenarios that provided fewer barriers to appropriate treatments. By 2025, more restrictive formularies would have resulted in 171,500 more cumulative bone and renal events among treated patients with HIV compared with an open formulary. This corresponds to AE treatment costs of $3.65 billion in more restrictive formularies compared with $1.43 billion in an open formulary. Finally, compared with an open formulary, there would be an additional 16,200 cumulative deaths in more restrictive formularies.
Conclusions: Less restrictive formulary designs, which allow patients with HIV to initiate potentially safer and more efficacious regimens based on their proclivity to AEs, yield better outcomes and reduce costs.
Am J Manag Care. 2018;24(Spec Issue No. 8):SP322-SP328Takeaway Points
This study evaluated the impact of formulary designs on HIV patient outcomes, specifically the frequency and costs of certain adverse events and excess mortality. Findings are relevant for policy makers developing formulary management approaches for HIV treatments.
The past 20 years have seen a rapid development of innovative antiretroviral therapies allowing patients with HIV to live longer and healthier lives and reducing the risk of transmission.1 Molecules approved since the early 2000s are more effective and associated with fewer and less serious adverse events (AEs) than earlier HIV therapies.2 The combination of these therapies into single-tablet regimens (STRs) has also simplified treatment administration, enhancing adherence and viral suppression.3-6 Despite the availability of these improved treatment options, fewer than half of the 1.1 million Americans living with HIV in 2015 were engaged in care and achieved viral suppression.7,8
Many factors impact engagement in care and viral suppression. Formulary and utilization management policies, such as tiering and step therapy, that require patients to initiate therapy on cheaper regimens before moving to potentially more effective alternatives can reduce access to drugs that receive less desirable formulary placement.9-11 For patients with HIV, formulary restrictions may have a number of repercussions, such as reduced adherence and viral suppression.12-14 Further, requiring patients to initiate one regimen before accessing others, without consideration of patient heterogeneity, may adversely impact those with comorbid conditions.1 Healthcare costs and utilization may also increase because of reduced viral suppression or increased AEs.
New therapies, such as those based on tenofovir alafenamide (TAF), continue to reduce the likelihood and severity of AEs, decreasing the possibility of a patient switching or discontinuing treatment.15-23 Although common AEs like nausea and headaches still occur across all HIV regimens, more serious events like renal failure or bone fractures are mitigated with the newest options.18,19,23 Still, despite demonstrated improvements in reducing AEs, new regimens are not always immediately available for patients because of formulary and utilization management policies.
We aimed to model the impacts of restrictive formulary designs on outcomes for patients with HIV and to demonstrate the costs of restricting access to novel HIV regimens that have better efficacy and safety profiles.
We modified a previously developed epidemiological model of HIV incidence, progression, and treatment to simulate the effects of 5 HIV formulary scenarios on patient outcomes in the United States (eAppendix A [eAppendices available at ajmc.com]).24,25 Our model incorporated incoming cohorts of HIV-susceptible individuals, natural progression of HIV for treated and untreated individuals, and death. Individuals transitioned to different disease stages at rates based on the HIV-infected population size and infectivity in treated and untreated populations. The HIV natural history progression was defined according to the CDC’s clinical classifications of HIV disease stages26: Patients in stage 1 had CD4+ cell counts greater than 500 cells/mcL; stage 2, 350 to 499 cells/mcL; stage 3, 200 to 349 cells/mcL; and stage 4, fewer than 200 cells/mcL. We used data from the CDC to estimate the distribution of recently infected individuals by stage.26 We also included transitions between treated and untreated states according to treatment initiation and discontinuation rates and the possibility of switching HIV regimens (described later).24,25
Our model started with 1.2 million people infected with HIV in 2016.26 We used published estimates from the CDC’s Medical Monitoring Report and HIV Surveillance reports to obtain the distribution of the initial population by disease stage and treatment status (eAppendix Table 2).26,27 We modeled HIV incidence through 2 categories of HIV transmission: (1) between untreated HIV-infected patients and susceptible individuals and (2) between treated HIV-infected patients and susceptible individuals. A lower rate of transmission was used for the latter category, reflecting the lower transmission rates observed for treated versus untreated infected patients.28,29 eAppendix A presents the details of the model and parameters.
Treatment Regimens and Formulary Scenarios
Our model included 4 STRs: emtricitabine/rilpivirine/tenofovir disoproxil fumarate (FTC/RPV/TDF), elvitegravir/cobicistat/emtricitabine/tenofovir disoproxil fumarate (EVG/COBI/FTC/TDF), elvitegravir/ cobicistat/emtricitabine/tenofovir alafenamide (EVG/COBI/FTC/TAF), and abacavir/dolutegravir/lamivudine (ABC/DTG/3TC). A fifth regimen, darunavir/ritonavir (protease inhibitor [PI]) plus 2 or more nucleoside reverse transcriptase inhibitors (PI/2NRTIs), was included as a second-line therapy in the event of virologic failure in certain formularies.
To encapsulate a range of formulary treatment policies, we developed 5 formulary scenarios that differed in access to HIV treatment regimens (eAppendix B). Within each formulary, patients faced AEs (specifically, renal toxicity or bone fracture) associated with certain HIV medications. Patients who experienced AEs or virologic failure switched to a different treatment regimen, determined by the rules of the particular formulary. Different efficacy and AE rates were used depending on pre-existing clinical conditions for patients. Treatment discontinuation rates were defined as the number of patients randomized to treatment who discontinued because of death, pregnancy, or study withdrawal, but did not include patients who discontinued treatment due to an AE. All parameter estimates were derived from the published literature or clinical trials. Detailed schematics for each formulary scenario can be found in eAppendix B.
The first formulary scenario, “most restrictive,” distributed patients equally across the TDF-based regimens and only allowed access to the initial treatment, regardless of AEs. In the event of a virologic failure, however, patients were transitioned to a PI/2NRTIs regimen.
Under the second and third formulary scenarios, “step (renal)”and “step (any AEs),” respectively, patients were distributed equally across the TDF-based regimens and were transitioned to an alternative regimen (either ABC/DTG/3TC or EVG/COBI/FTC/TAF) after the occurrence of an AE. The step (renal) scenario only allowed patients to transition to EVG/COBI/FTC/TAF if they experienced a renal AE. Patients with a bone fracture event were moved to ABC/DTG/3TC. In contrast, the step (any AEs) scenario allowed for transitions to EVG/COBI/FTC/TAF if a patient experienced either of the AEs. Under both step scenarios, patients with virologic failure on their first-line regimen were moved to a second-line treatment of PI/2NRTIs.
The fourth scenario, “presorted” formulary, allowed pre-existing clinical conditions to determine the initial treatment regimen. Under this scenario, patients with an existing bone disease, osteopenia, started on ABC/DTG/3TC or EVG/COBI/FTC/TAF; patients with a disposition to a renal event started on EVG/COBI/FTC/TAF; and all other patients started on a TDF-based regimen. Switching regimens occurred in the case of an AE or treatment failure, as in the step formularies.
The last scenario, “open” formulary, was the least restrictive and reflected current market trends among a population facing minimal or no co-pays. Patients with osteopenia (35.1%) were started on ABC/DTG/3TC or EVG/COBI/FTC/TAF, whereas patients with a disposition to a renal AE (6.2%) were started on EVG/COBI/FTC/TAF. Additional patients were started on the TAF-based regimen such that this overall percentage matched the market distribution of treatment-naïve patients in the Ryan White program, a government program that provides cost-sharing assistance (and other services) to low-income people with HIV, as of the second quarter of 2016 (46%).30 All other patients were started on a TDF-based regimen. Patients switched to an alternative regimen after treatment failure or AE.
An estimated 3% to 8% of individuals have allele HLA-B*5701, which is associated with hypersensitivity reactions from the abacavir component in ABC/DTG/3TC.31-33 Based on those estimates, we assumed that 5.5% of the population had the allele. In the relevant formularies, we respected the hypersensitivity of those patients by using EVG/COBI/FTC/TAF, where ABC/DTG/3TC would have otherwise been chosen.
The costs of bone and renal AEs were calculated using data from the 2014 Medical Expenditure Panel Survey (eAppendix C).34 Costs included inpatient, outpatient, emergency department, prescription, and office-based costs for each medical condition.
Our model projected the annual number of renal AEs and bone fractures for the treated population, the cumulative value of treatment costs associated with these events over the next 10 years, and cumulative all-cause deaths. Costs are represented in 2016 US$ and discounted at an annual rate of 3%.35
We calibrated the model by comparing the percentage of patients with HIV treated by 2020, cumulative deaths from 2016 to 2025, and number of prevalent HIV cases in 2025 with published estimates from Shah et al36 by adjusting HIV infectivity and the rate of disease progression across treated and untreated stages.
All outcomes of interest were most favorable for patients in the open formulary, followed by the presorted formulary. By 2025, the number of renal and bone AEs, costs associated with those events, and cumulative all-cause deaths were considerably lower in the open formulary than in the other scenarios (Figure 1).
Our calibrated parameters produced estimates of HIV cases by 2025 and the percentage of individuals on treatment in 2020 that aligned with those reported in Shah et al (eAppendix D).36 Our model showed an increase in the HIV-infected population from 1.2 million in 2016 to approximately 1.53 million in 2025 under the 3 most restrictive formulary scenarios and 1.52 million in the open scenario. Those results fell within Shah et al’s CI of 1.24 million to 1.57 million. Across the formulary scenarios, our model produced a range of 515,000 to 532,000 cumulative deaths from 2016 to 2025, all of which fall within Shah et al’s CI of 364,000 to 578,000.
Bone and Renal Events
The open scenario led to 171,500 and 68,500 fewer cumulative bone and renal events among treated patients with HIV by 2025 compared with the average of the 3 more restrictive scenarios and the presorted formulary scenario, respectively (Figure 1 and Figure 2). Although the number of AEs was the greatest under the most restrictive formulary scenario, both step therapy formulary scenarios resulted in more AEs than the presorted and open formularies.
From 2016 to 2025, cumulative costs associated with both bone fractures and renal disease were $2.23 billion higher in the step (renal) formulary, $2.19 billion higher in the step (any AEs) formulary, and $648.7 million higher in the presorted formulary, relative to the open formulary (Figure 3). Although there were fewer renal AEs across all scenarios, treatment costs for renal events were much higher than for bone fractures, leading to greater overall total costs.
Inpatient costs incurred by a renal event or bone fracture were the largest component of the cumulative costs in 2016-2025, representing 52% and 48%, respectively (Table). Office-based provider costs were the second largest component, comprising 35% of renal AE costs and 19% of bone fracture AE costs. These 2 components alone represented $1.2 billion in cumulative costs in the open formulary, $1.7 billion in the presorted formulary, and slightly more than $3.0 billion in the step (any AEs) and step (renal) scenarios over 10 years (Table).
In more restrictive scenarios, considerably more people switched to a second-line treatment of PI/2NRTIs compared with the less restrictive formularies. In 2025, about 58,000 patients in the 3 most restrictive formularies were on a second-line regimen because of virologic failure. In contrast, an estimated 41,000 patients in the presorted formulary and 27,000 patients in the open formulary were on the second-line regimen due to virologic failure.
Compared with the open formulary, restricting access to HIV treatments led to 16,200 more deaths, on average, in the more restrictive formularies by 2025 (Figure 1). The presorted formulary resulted in about 7300 more cumulative deaths by 2025 than the open formulary.
Our findings illustrate that a scenario that matches the distribution of treatments observed in a treatment-naÏve HIV population that obtains a high degree of access to available treatments—Ryan White patients—significantly reduces AE treatment costs and mortality rates relative to more restrictive formulary designs. These results are driven by a greater initial utilization of therapies with better efficacy and AE profiles in our open design, along with sorting patients who are predisposed to certain AEs to more tolerable regimens. The open design outperforms the presorted design because even more patients are started on the newer TAF-based therapy, which has demonstrated lower rates of serious AEs and improved efficacy.37 Compared with the open formulary, we estimated that the average restrictive formulary design, which initiates fewer people on therapies with better efficacy and safety profiles and allows switching to those therapies only under particular circumstances, would resultin $2.28 billion in additional healthcare costs over 10 years and 3.5% more deaths of patients with HIV.
Although step therapy approaches aim to reduce the costs of treatment by prioritizing less-costly regimens, in practice, the number of people receiving non-STR therapies may contribute to future increases in healthcare costs.38 Our analysis, which incorporates a second-line multitablet regimen (MTR) therapy in the event of virologic failure on a first-line treatment, complements other research findings that STRs are associated with lower per patient healthcare and hospitalization costs compared with MTRs.39
Fewer cumulative all-cause deaths in the less restrictive formulary scenarios are attributable to several factors. First, more patients are receiving more efficacious regimens. Second, lower AE rates for those on the TAF-based therapy result in fewer patients switching regimens. Using values for a statistical life reported in the literature, the discounted value of the additional lives saved over 10 years in the open formulary versus the more restrictive formularies ranges from $58 billion to $188 billion, depending on the value of a statistical life used ($4.3 million, $9.2 million, or $13.8 million in 2016 US$).40 This implies that additional spending in that range is justified by a standard social value criterion of the benefits exceeding the costs (based only on lives saved, and excluding savings from reduced AEs).
Our findings demonstrate that policies that reduce patient access to HIV regimens with better outcomes in terms of efficacy, AE profiles, or adherence can have significant health and economic consequences. Some Medicaid plans have used preferred drug lists (PDLs), requiring a prescriber to receive prior authorization (PA) from the plan, and some health insurance Marketplace insurers have used restrictive formulary designs for HIV therapies.41-44 Our analysis suggests that the implications of such restrictions should be carefully considered. Effects on outcomes depend on the specific drugs chosen and the nature of the restrictions. With respect to Medicare, our analysis illustrates the possible benefits from the protected class status that CMS has always maintained for antiretroviral drugs, ruling out closed formularies, limiting utilization management strategies, and expediting formulary review.
Our study has several limitations. First, our model included some simplifying assumptions. For example, we assumed a uniform progression rate between stages 1 and 4 for treated patients that is intended to capture net progression, thereby embedding immunological recovery due to treatment. We did, however, calibrate our model so that key outputs were consistent with projections from a more complex published model. This approach may have resulted in an under- or overestimation of the projected burden of HIV. Second, although we varied transmission rates for untreated versus treated patients, we did not vary transmission rates by infection stage and instead used average rates based on the literature. This simplification might understate the outcome differences across the scenarios because transmission by treated patients would be relatively lower for formulary designs that generate better efficacy.
Third, the treatment efficacy parameters were based primarily on clinical trial data, which may not match real-world outcomes. Fourth, we focused only on fractures and renal AEs in our analysis. Thus, the cost differences across formularies would be greater if the frequencies of other events are, on balance, correlated with those included here. Certain cardiovascular events have been associated with ABC/DTG/3TC because of the abacavir component,45 but they were not included in our analysis because their occurrence was not explicitly analyzed in the clinical trials used in our model.
Fifth, the model did not incorporate effects of an aging population due to the limited follow-up of patients involved in the clinical trials used in this study. Also, the median age of patients in the trial data ranged from 33 to 38 years, while the median age of patients with HIV is above 45 years.19,27,46 Because the risk of osteopenia and osteoporosis in patients with HIV increases with age, as does the risk of fractures, it is reasonable to think that the differences among the formularies in terms of AE costs and rates will be greater than our modeling indicates and may increase as the population ages and patients spend more time on HIV regimens.47-49 Additionally, there may be varying quality-of-life effects experienced by an aging population that we did not capture.
Sixth, we acknowledge that, in practice, the definition of step therapy could vary across health plans. Our chosen formulary designs reflect a range of access restrictions. For example, a tiered co-payment design bears similarities to step designs: A preferred drug with a low co-pay is likely to be prescribed first, with movement to a higher-tier drug in the case of treatment failure or serious AE. There is also similarity between the use of PDLs or PA requirements and step therapy designs.
Finally, our characterization of an “open” formulary may be overly optimistic. Although the scenario matched the real-world utilization rates for therapies that have been observed in a treatment-naïve population facing low or no co-pays, we assumed that all patients with low kidney function or osteopenia were among those started on either TAF- or ABC-based regimens. Although this characterization captures the view that providers will sort patients to the most appropriate regimens, to the extent that they fail in this regard, outcomes will not be as favorable.
The findings from this study suggest that less restrictive formulary designs, which allow providers to start patients with HIV/AIDS on different regimens based on their proclivity to AEs (and result in more people using a TAF-based regimen, which is more effective and has a better AE profile), yield better outcomes and reduce AE treatment costs compared with more restrictive step therapy formularies. Although tiered co-pay designs and other utilization management strategies were not directly studied here, a similar impact on the sequencing of therapies for patients suggests that these practices would likely also result in suboptimal patient outcomes.Author Affiliations: Precision Health Economics (JB, CH, MK, LY, JC), LosAngeles, CA; Leonard D. Schaeffer Center for Health Policy and Economics, University of Southern California (JR), Los Angeles, CA.
Source of Funding: Funding for this study was provided by Gilead Sciences to Precision Health Economics.
Author Disclosures: Dr Baumgardner, Ms Huber, Dr Kabiri, Ms Yoon, and Ms Chou are employees of Precision Health Economics, which received financial support for this research from Gilead Sciences. Dr Romley is a consultant to Precision Health Economics.
Authorship Information: Concept and design (JB, CH, MK, LY, JR); acquisition of data (CH, LY); analysis and interpretation of data (JB, CH, MK, LY, JR); drafting of the manuscript (JB, CH, MK, LY, JC); critical revision of the manuscript for important intellectual content (JB, CH, JC, JR); statistical analysis (JB, MK); administrative, technical, or logistic support (CH, MK, LY, JC); and supervision (JB).
Address Correspondence to: James Baumgardner, PhD, Precision Health Economics, 11100 Santa Monica Blvd, Ste 500, Los Angeles, CA 90026. Email: firstname.lastname@example.org.REFERENCES
1. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV-1- Infected Adults and Adolescents. Washington, DC: HHS; 2016. aidsinfo.nih.gov/guidelines. Accessed July 14, 2015.
2. Bartlett JG. Ten years of HAART: foundation for the future. Paper presented at: 13th Conference on Retroviruses and Opportunistic Infections; February 5-8, 2006; Denver, CO. medscape.org/viewarticle/523119. Accessed August 7, 2017.
3. Aldir I, Horta A, Serrado M. Single-tablet regimens in HIV: does it really make a difference? Curr Med Res Opin. 2014;30(1):89-97. doi: 10.1185/03007995.2013.844685.
4. Bangsberg DR, Ragland K, Monk A, Deeks SG. A single tablet regimen is associated with higher adherence and viral suppression than multiple tablet regimens in HIV+ homeless and marginally housed people. AIDS. 2010;24(18):2835-2840. doi: 10.1097/QAD.0b013e328340a209.
5. Clumeck N, Molina JM, Henry K, et al; GS-236-0103 Study Team. A randomized, double-blind comparison of single-tablet regimen elvitegravir/cobicistat/emtricitabine/tenofovir DF vs ritonavir-boosted atazanavir plus emtricitabine/tenofovir DF for initial treatment of HIV-1 infection: analysis of week 144 results. J Acquir Immune Defic Syndr. 2014;65(3):e121-e124. doi: 10.1097/QAI.0000000000000089.
6. Dejesus E, Young B, Morales-Ramirez JO, et al; AI266073 Study Group. Simplification of antiretroviral therapy to a single-tablet regimen consisting of efavirenz, emtricitabine, and tenofovir disoproxil fumarate versus unmodified antiretroviral therapy in virologically suppressed HIV-1-infected patients. J Acquir Immune Defic Syndr. 2009;51(2):163-174. doi: 10.1097/QAI.0b013e3181a572cf.
7. What is the HIV care continuum? HIV.gov website. hiv.gov/federal-response/policies-issues/hiv-aids-care-continuum. Updated December 30, 2016. Accessed April 6, 2017.
8. HIV Surveillance Report, 2016. Atlanta, GA: CDC; 2016. cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-report-us.pdf. Accessed April 6, 2017.
9. Goldman DP, Joyce GF, Zheng Y. Prescription drug cost sharing: associations with medication and medical utilization and spending and health. JAMA. 2007;298(1):61-69. doi: 10.1001/jama.298.1.61.
10. HMO formulary. Kaiser Permanente website. healthy.kaiserpermanente.org/static/health/pdfs/formulary/mid/mid_hmo_formulary.pdf. Updated March 6, 2018. Accessed October 5, 2016.
11. Your 2016 prescription drug list. UnitedHealthcare website. myuhc.com/content/myuhc/Member/Assets/Pdfs/100-16910_FS_3T_Adv_PDL_116_6.pdf. Published 2016. Accessed October 5, 2016.
12. Zamani-Hank Y. The Affordable Care Act and the burden of high cost sharing and utilization management restrictions on access to HIV medications for people living with HIV/AIDS. Popul Health Manag. 2016;19(4):272- 278. doi: 10.1089/pop.2015.0076.
13. Schaecher KL. The importance of treatment adherence in HIV. Am J Manag Care. 2013;19(suppl 12):S231-S237. 14. Kesselheim AS, Huybrechts KF, Choudhry NK, et al. Prescription drug insurance coverage and patient health outcomes: a systematic review. Am J Public Health. 2015;105(2):e17-e30. doi: 10.2105/AJPH.2014.302240.
15. Bloch M, Tong WW, Hoy J, et al; TROP (Switch from Tenofovir to Raltegravir for Low Bone Density) Study Team. Switch from tenofovir to raltegravir increases low bone mineral density and decreases markers of bone turnover over 48 weeks. HIV Med. 2014;15(6):373-380. doi: 10.1111/hiv.12123.
16. d’Arminio Monforte A, Lepri AC, Rezza G, et al. Insights into the reasons for discontinuation of the first highly active antiretroviral therapy (HAART) regimen in a cohort of antiretroviral naïve patients. I.CO.N.A. Study Group. Italian Cohort of Antiretroviral-Naive Patients. AIDS. 2000;14(5):499-507.
17. O’Brien ME, Clark RA, Besch CL, Myers L, Kissinger P. Patterns and correlates of discontinuation of the initial HAART regimen in an urban outpatient cohort. J Acquir Immune Defic Syndr. 2003;34(4):407-414.
18. Sax PE, Wohl D, Yin MT, et al; GS-US-292-0104/0111 Study Team. Tenofovir alafenamide versus tenofovir disoproxil fumarate, coformulated with elvitegravir, cobicistat, and emtricitabine, for initial treatment of HIV-1 infection: two randomised, double-blind, phase 3, non-inferiority trials. Lancet. 2015;385(9987):2606-2615. doi: 10.1016/S0140-6736(15)60616-X.
19. Sax PE, Zolopa A, Brar I, et al. Tenofovir alafenamide vs. tenofovir disoproxil fumarate in single tablet regimens for initial HIV-1 therapy: a randomized phase 2 study. J Acquir Immune Defic Syndr. 2014;67(1):52-58. doi: 10.1097/QAI.0000000000000225.
20. Winston J, Chonchol M, Gallant J, et al. Discontinuation of tenofovir disoproxil fumarate for presumed renal adverse events in treatment-naive HIV-1 patients: meta-analysis of randomized clinical studies. HIV Clin Trials. 2014;15(6):231-245. doi: 10.1310/hct1506-231.
21. Woodward CL, Hall AM, Williams IG, et al. Tenofovir-associated renal and bone toxicity. HIV Med. 2009;10(8):482-487. doi: 10.1111/j.1468-1293.2009.00716.x.
22. Yuan Y, L’Italien G, Mukherjee J, Iloeje UH. Determinants of discontinuation of initial highly active antiretroviral therapy regimens in a US HIV-infected patient cohort. HIV Med. 2006;7(3):156-162. doi: 10.1111/j.1468-1293.2006.00355.x.
23. Cooper RD, Wiebe N, Smith N, Keiser P, Naicker S, Tonelli M. Systematic review and meta-analysis: renal safety of tenofovir disoproxil fumarate in HIV-infected patients. Clin Infect Dis. 2010;51(5):496-505. doi: 10.1086/655681.
24. Granich RM, Gilks CF, Dye C, De Cock KM, Williams BG. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: a mathematical model. Lancet. 2009;373(9657):48-57. doi: 10.1016/S0140-6736(08)61697-9.
25. Romley JA, Juday T, Solomon MD, Seekins D, Brookmeyer R, Goldman DP. Early HIV treatment led to life expectancy gains valued at $80 billion for people infected in 1996-2009. Health Aff (Millwood). 2014;33(3):370- 377. doi: 10.1377/hlthaff.2013.0623.
26. Behavioral and Clinical Characteristics of Persons Receiving Medical Care for HIV Infection—Medical Monitoring Project, United States, 2014 Cycle (June 2014- May 2015). Atlanta, GA: CDC; 2016. cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-hssr-mmp-2014.pdf. Accessed April 6, 2017.
27. CDC. Monitoring selected national HIV prevention and care objectives by using HIV surveillance data— United States and 6 dependent areas—2015. HIV AIDS Surveill Suppl Rep. 2017;22(2). cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-supplemental-report-vol-22-2.pdf. Accessed April 6, 2017.
28. Hall HI, Holtgrave DR, Tang T, Rhodes P. HIV transmission in the United States: considerations of viral load, risk behavior, and health disparities. AIDS Behav. 2013;17(5):1632-1636. doi: 10.1007/s10461-013-0426-z.
29. Skarbinski J, Rosenberg E, Paz-Bailey G, et al. Human immunodeficiency virus transmission at each step of the care continuum in the United States. JAMA Intern Med. 2015;175(4):588-596. doi: 10.1001/jamainternmed.2014.8180.
30. Data on file. Percentage of TAF share among antiretroviral-naive Ryan White patients in Q2 2016 [Ipsos HIV US Scope Q2 2016]. Ipsos; 2016.
31. Faruki H, Heine U, Brown T, Koester R, Lai-Goldman M. HLA-B*5701 clinical testing: early experience in the United States. Pharmacogenet Genomics. 2007;17(10):857-860. doi: 10.1097/FPC.0b013e328285da2e.
32. Mallal S, Phillips E, Carosi G, et al; PREDICT-1 Study Team. HLA-B*5701 screening for hypersensitivity to abacavir. N Engl J Med. 2008;358(6):568-579. doi: 10.1056/NEJMoa0706135.
33. Small C, Wohl D, Margolis DA, et al. Prevalence of HLA-B*5701 allele in HIV-infected subjects in North America and reductions in risk for development of abacavir associated hypersensitivity reaction. Paper presented at: 52nd International Conference on Antimicrobial Agents and Chemotherapy; September 9-12, 2012; San Francisco, CA. natap.org/2012/ICAAC/ICAAC_16.htm. Accessed July 24, 2017.
34. Medical Expenditure Panel Survey: household component event files. Agency for Healthcare Research and Quality website. meps.ahrq.gov/data_stats/download_data_files.jsp. Accessed October 1, 2015.
35. Russell LB, Gold MR, Siegel JE, Daniels N, Weinstein MC. The role of cost-effectiveness analysis in health and medicine. JAMA. 1996;276(14):1172-1177. doi: 10.1001/jama.1996.03540140060028.
36. Shah M, Perry A, Risher K, et al. Effect of the US National HIV/AIDS Strategy targets for improved HIV care engagement: a modelling study. Lancet HIV. 2016;3(3):e140-e146. doi: 10.1016/S2352-3018(16)00007-2.
37. Wang H, Lu X, Yang X, Xu N. The efficacy and safety of tenofovir alafenamide versus tenofovir disoproxil fumarate in antiretroviral regimens for HIV-1 therapy: meta-analysis. Medicine (Baltimore). 2016;95(41):e5146. doi: 10.1097/MD.0000000000005146.
38. Carlton RI, Bramley TJ, Nightengale B, Conner TM, Zacker C. Review of outcomes associated with formulary restrictions: focus on step therapy. Am J Pharm Benefits. 2010;2(1):50-58.
39. Cohen CJ, Meyers JL, Davis KL. Association between daily antiretroviral pill burden and treatment adher- ence, hospitalisation risk, and other healthcare utilisation and costs in a US Medicaid population with HIV. BMJ Open. 2013;3(8):e003028. doi: 10.1136/bmjopen-2013-003028.
40. Robinson LA, Hammitt JK. Valuing reductions in fatal illness risks: implications of recent research. Health Econ. 2016;25(8):1039-1052. doi: 10.1002/hec.3214.
41. Andrews M. 7 insurers alleged to use skimpy drug coverage to discourage HIV patients. Kaiser Health News website. khn.org/news/7-insurers-alleged-to-use-skimpy-drug-coverage-to-discourage-hiv-patients. Published October 18, 2016. Accessed April 3, 2017.
42. Preferred drug list: Illinois Medicaid. Illinois Department of Healthcare and Family Services website. illinois.gov/hfs/SiteCollectionDocuments/January2017PDL.pdf. Published January 1, 2017. Accessed March 30, 2017.
43. Preferred drug list. Mississippi Division of Medicaid website. medicaid.ms.gov/providers/pharmacy/preferred-drug-list. Accessed March 30, 2017.
44. Patient access to HIV drugs in exchange plans is limited compared to other sources of coverage [news release]. Washington, DC: Avalere Health; November 11, 2015. avalere.com/expertise/managed-care/insights/patient-access-to-hiv-drugs-in-exchange-plans-is-limited-compared-to-other. Accessed April 3, 2017.
45. Marcus JL, Neugebauer RS, Leyden WA, et al. Use of abacavir and risk of cardiovascular disease among HIV-infected individuals. J Acquir Immune Defic Syndr. 2016;71(4):413-419. doi: 10.1097/QAI.0000000000000881.
46. Sax PE, DeJesus E, Mills A, et al; GS-US-236-0102 Study Team. Co-formulated elvitegravir, cobicistat, emtricitabine, and tenofovir versus co-formulated efavirenz, emtricitabine, and tenofovir for initial treatment of HIV-1 infection: a randomised, double-blind, phase 3 trial, analysis of results after 48 weeks. Lancet. 2012;379(9835):2439-2448. doi: 10.1016/S0140-6736(12)60917-9.
47. Carr A, Grund B, Neuhaus J, et al; International Network for Strategic Initiatives in Global HIV Trials (INSIGHT) START Study Group. Prevalence of and risk factors for low bone mineral density in untreated HIV infection: a substudy of the INSIGHT Strategic Timing of AntiRetroviral Treatment (START) trial. HIV Med. 2015;16(suppl 1):137-146. doi: 10.1111/hiv.12242.
48. Hileman CO, Eckard AR, McComsey GA. Bone loss in HIV: a contemporary review. Curr Opin Endocrinol Diabetes Obes. 2015;22(6):446-451. doi: 10.1097/MED.0000000000000200.
49. Nasi M, De Biasi S, Gibellini L, et al. Ageing and inflammation in patients with HIV infection. Clin Exp Immunol. 2017;187(1):44-52. doi: 10.1111/cei.12814.