- Email: [email protected]
Value of Comprehensive HCV Treatment among Vulnerable, High-Risk Populations
VALUE IN HEALTH ] (]]]]) ]]]–]]]
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jval
Value of Comprehensive HCV Treatment among Vulnerable, High-Risk Populations Gigi A. Moreno, PhD1, Alice Wang, MA2, Yuri Sánchez González, PhD2, Oliver Díaz Espinosa, PhD1,*, Diana K. Vania, MSc1, Brian R. Edlin, MD3, Ronald Brookmeyer, PhD4 1 Precision Health Economics, Los Angeles, CA, USA; 2AbbVie, Inc., Mettawa, IL, USA; 3Weill Cornell Medical College, Cornell University, New York City, NY, USA; 4Fielding School of Public Health, University of California Los Angeles, Los Angeles, CA, USA
AB STR A CT
Objectives: The objective of this study was to explore the trade-offs society and payers make when expanding treatment access to patients with chronic hepatitis C virus (HCV) infection in early stages of disease as well as to vulnerable, high-risk populations, such as people who inject drugs (PWID) and HIV-infected men who have sex with men (MSM-HIV). Methods: A discrete time Markov model simulated HCV progression and treatment over 20 years. Population cohorts were defined by behaviors that influence the risk of HCV exposure: PWID, MSM-HIV, an overlap cohort of individuals who are both PWID and MSM-HIV, and all other adults. Six different treatment scenarios were modeled, with varying degrees of access to treatment at different fibrosis stages and to different risk cohorts. Benefits were measured as quality-adjusted lifeyears and a $150,000/quality-adjusted life-year valuation was used to assess social benefits. Results: Compared with limiting treatment to METAVIR fibrosis stages F3 or F4 and excluding PWID, expanding treatment to patients in all fibrosis stages and including PWID reduces
cumulative new infections by 55% over a 20-year horizon and reduces the prevalence of HCV by 93%. We find that treating all HCV-infected individuals is cost saving and net social benefits are over $500 billion greater compared with limiting treatment. Including PWID in treatment access saves 12,900 to 41,200 lives. Conclusions: Increased access to treatment brings substantial value to society and over the long-term reduces costs for payers, as the benefits accrued from long-term reduction in prevalent and incident cases, mortality, and medical costs outweigh the cost of treatment. Keywords: hepatitis C, people who inject drugs, men who have sex with men, Markov model.
Introduction
disease progression, which will reduce morbidity, mortality, and health care expenditures related to cirrhosis, decompensated liver disease, hepatocellular carcinoma, and liver transplants [9]. The most significant mode of transmission for HCV infection in the United States is through injection drug use [10]; approximately 54% to 77% of new HCV diagnoses are among people who inject drugs (PWID) [11,12]. An estimated 3.5 million people have injected drugs in the United States during their lifetime [4], with the prevalence of HCV infection in this population estimated at 73% (range 70%–77%) [13]. Because of the high probability of contracting HCV infection through needle-sharing [14], treating PWID infected with HCV, particularly in early stages of the disease, may reduce transmission. Although the probability of HCV transmission through sexual contact is low, it is believed that HIV co-infection may aid the
Hepatitis C virus (HCV) infection is a transmissible viral infection that is usually asymptomatic in the early stages of the disease and may lead to serious liver complications including decompensated cirrhosis and hepatocellular carcinoma, which are associated with a significant mortality and health care and cost burden [1–3]. It is estimated that at least 3 to 5 million people are infected with HCV in the United States [4,5]; however, approximately half are unaware of their infection due to the asymptomatic nature of early stages of the disease [6–8]. Comprehensive HCV treatment may have important societal benefits beyond the immediate benefit to the patient. With each successfully treated patient, the pool of individuals able to transmit the disease decreases. In addition, early treatment for HCV has the potential to hinder
Copyright & 2017, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Conflict of interest: G. A. Moreno and D. K. Vania were employees of Precision Health Economics, LLC, at the time the manuscript was prepared. A. Wang and Y. S. Gonzalez are employees of AbbVie Inc. and may own AbbVie stock or stock options. O. D. Espinosa is an employee of Precision Health Economics, LLC, which received payment from AbbVie for conducting research analysis. B. R. Edlin and R. Brookmeyer are consultants to Precision Health Economics, LLC. * Address correspondence to: Oliver Díaz Espinosa, Precision Health Economics, 11100 Santa Monica Boulevard, Suite 500, Los Angeles, CA 90025. E-mail: [email protected]. 1098-3015$36.00 – see front matter Copyright & 2017, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). http://dx.doi.org/10.1016/j.jval.2017.01.015
2
VALUE IN HEALTH ] (]]]]) ]]]–]]]
infectivity of HCV [15]. The incidence rate of HCV infection among men who have sex with men (MSM) who are infected with human immunodeficiency virus (MSM-HIV) is four times greater than that among HIV-uninfected MSM [16]. Furthermore, it has been estimated that approximately 25% of HIV-infected individuals are coinfected with HCV [17]. The occurrence of overlapping high-risk behaviors (e.g., MSM-HIV who also inject drugs) may further increase the risk of HCV infection. Taylor et al. [18] observe that MSM-HIV who report injection drug use are more than six times as likely to acquire HCV infection compared with those who do not engage in injection drug use [18]. Thus, MSM-HIV are a significant source of HCV transmission given the interaction with PWID and that HIV infection may increase infectivity of HCV. Despite the demonstrated efficacy of new interferon-free direct-acting antiviral agents, access to these treatments has been limited in the United States to those in advanced fibrosis stages or with HIV co-infection [19,20]. In addition, many payers have policies that deny treatment to HCV-infected PWID. The objective of this study was to explore the trade-offs society and payers make when access is expanded to patients in early stages of disease as well as to vulnerable populations such as PWID and MSM-HIV, taking into account transmission between these groups. Herein, we describe the benefits and costs to society under these expanded treatment access scenarios and consider their effects on the total number of HCV infections in the United States. We also consider how these trade-offs are affected when treatment access for PWID is combined with disease prevention programs, such as needle-syringe programs (NSPs).
Methods We developed a discrete time Markov model using the R programming language to simulate HCV progression and treatment over a time horizon of 20 years [21]. In each yearly cycle, the population transitions through each disease state at assumed probabilities (transition probabilities are reported in Supplemental Materials found at 10.1016/j.jval.2017.01.015). The structure of the model is described in detail in Van Nuys et al. [22] and in Appendix Figure 1 in Supplemental Materials found at http://dx. doi.org/10.1016/j.jval.2017.01.015.
the simulation. Because we assumed no ongoing infections among OA, the size of this cohort shrinks during the simulation. Mortality rates for the risk cohorts are computed as described in Linthicum et al. [23]. Within each risk cohort, we modeled the three most common HCV genotypes in the United States (genotypes 1, 2, and 3) [26]. We assumed that individuals may be infected with only one genotype at a time, but if cured and reinfected, they may re-enter the model with any of the three genotypes. Specifically, the cohort cured of HCV in early fibrosis stages (F0-F2) returns to the susceptible uninfected population. Those cured in later stages (F3 and higher) may progress to additional liver damage, albeit at a slower rate than the infected population. If re-infected, the cohort cured in later fibrosis stages re-enters the infected population at the highest fibrosis stage attained by that cohort (see Appendix Figure 1).
Model Parameters HCV treatment costs, medical expenditures (nontreatment medical costs), quality-of-life weights, and mortality rates were collected from estimates in the published literature. All cost parameters were inflated to 2015 U.S. dollars, and future costs and quality-adjusted life-years (QALYs) were discounted at an annual 3% rate. Treatment costs vary by genotype and fibrosis stage, where the late-stage genotype 2 cohort (F3 and higher) and the entire genotype 3 cohort experience higher treatment costs than genotype 1 because of differing treatment durations, consistent with a previous study [22]. Medical expenditures and quality-of-life weights vary by disease state. We assumed that PWID have the same medical expenditures as the OA cohort and the OV cohort has the same medical expenditures as the MSMHIV cohort. Treatment efficacy, treatment costs, and medical expenditures do not vary over time, with the exception of treatment costs, which are discounted 79% after 16 years to account for patent expiration and future market competition [22]. We conducted sensitivity analyses to test the effects of several key assumptions on our results. We tested the sensitivity of results by varying the assumed values of six key parameters: cost of treatment (⫾15%); percent insured (extreme 20%–90%); quality-of-life weights (⫾5%), rate of diagnosis (extreme 25%– 100%); starting population (⫾5%), and sustained virologic response rate (⫾15%).
Study Population To simulate real-world transmission of HCV, the initial population was stratified into risk cohorts, defined by behaviors that influence the risk of HCV exposure and subsequent infection. Previous studies have modeled mutually exclusive risk cohorts [22,23], potentially overlooking the additive risk of HCV infection among individuals who are both PWID and MSM-HIV [24]. Building upon previous studies, three high-risk groups and a low-risk, catchall group of all other adults transitioned through the Markov model. The risk cohorts include 1) individuals who self-identify as PWID but not MSM-HIV; 2) HIV-infected individuals who self-identify as MSM but not PWID; 3) the overlap (“OV”) of individuals who are both PWID and MSM-HIV; and 4) all other adults (“OA”) who were born before 1992 when systematic testing of the blood supply for HCV was introduced. We assumed there was no ongoing HCV transmission in the OA cohort [25]. Appendix Table 2 in Supplemental Materials found at http://dx.doi.org/10.1016/j.jval.2017.01.015 presents the distribution of the starting population by risk cohort. In the model, an individual belongs to one risk cohort for the duration of the simulation. As described in Supplemental Materials, our model allows interactions between the PWID and MSMHIV groups through contact with the OV group. We also assumed that the high-risk cohorts experienced ongoing entry and exit (mortality) such that their cohort sizes remained constant over
Policy Scenarios We simulated six treatment access scenarios: a Status Quo scenario and five alternative scenarios with varying degrees of expanded access to treatment. All scenarios are evaluated against the Status Quo (baseline) scenario, a stylized representation of treatment access policies prevailing in 2015 among the largest U.S. payers. In the Status Quo scenario, we assume that only those who do not inject drugs have access to treatment if their disease has advanced to METAVIR fibrosis stage F3 or F4 and if they are insured [27–32]. The METAVIR scoring system quantifies the degree of liver fibrosis in patients with liver diseases such as HCV [33]. In the first alternative scenario, we assume a partial expansion of treatment to patients in fibrosis stages F2 to F4, continuing to exclude PWID from treatment (“Partial Expansion”). In the second alternative scenario, treatment access is expanded to patients in all fibrosis stages (F0-F4; “Treat All”), excluding access to PWID. In the remaining three alternative scenarios, we expand the Status Quo, Partial Expansion, and Treat All scenarios to include PWID. A final analysis assessed the effects of NSPs as a possible strategy to reduce the transmission of HCV among PWID. These programs offer access to clean drug injection equipment for PWID and have been proven to reduce the spread of HIV [34,35]. Their evidence in reducing the spread of HCV is less clear, but HCV
VALUE IN HEALTH ] (]]]]) ]]]–]]]
incidence rates among PWID have fallen by an order of magnitude after the implementation of HIV prevention interventions, including NSPs [36]. To assess the possible impact of expanded NSPs combined with expanded access to treatment for the PWID population, we increased costs in our analysis to include the annual cost of the programs, estimated to be $1694 per user [37,38]. On the basis of current estimates of PWID who access or are willing to access an NSP, we assumed that 61% of the PWID population in our model accessed an NSP [39,40]. All other model assumptions remained the same. Only alternative treatment scenarios that include PWID were considered in the NSP analysis.
Model Outputs We compute epidemiological and clinical outcomes, including HCV prevalence, incidence, deaths averted, and cases of liver disease. We also estimate economic outputs, including cost savings to payers, net social benefits, and incremental costeffectiveness ratios (ICERs) for each alternative scenario relative to the Status Quo over a 20-year horizon. In each time period, net benefits are calculated as the difference between the benefits (QALYs valued at $150,000/QALY [41]) and total costs (treatment, disability, and medical expenditures). Unless otherwise noted, results for each alternative policy are given as changes relative to the Status Quo scenario.
Results HCV Prevalence Figure 1A depicts the simulated U.S. population infected with HCV starting from the most recent National Health and Nutrition Examination Survey (NHANES) estimates in 2012 [4]; under the Status Quo, the number of people with HCV infection decreases by 50% after approximately 11 years. A partial expansion in treatment to include patients in fibrosis stages F2 to F4 decreases the time it takes to reduce prevalence by 50% to 7 years. Further
(A)
3
expanding treatment to PWID in fibrosis stages F2 to F4 reduces the prevalence by half in 6 years. A policy of HCV treatment access for all fibrosis stages and PWID reduces HCV infections by 50% in 2 years.
HCV Incidence Figure 1B depicts the simulated number of new HCV infections per year in the U.S. population; under all treatment scenarios, the incidence of new HCV infections in the United States will decrease. In Partial Expansion and Treat All scenarios that exclude PWID from treatment, the number of new infections is 16,700 and 14,800 per year, respectively, at the end of the 20-year simulation. Including PWID in treatment access significantly reduces infections: in the Partial Expansion and Treat All scenarios including PWID, new infections decrease to 10,700 and 1,900 per year, respectively. This represents a 36% and 87% decrease, respectively, in the total number of new infections compared with scenarios that exclude PWID.
Deaths and Liver Disease Averted The benefits of expanding access to treat earlier fibrosis stages and PWID are also observed in the number of averted deaths and averted cases of liver disease relative to the Status Quo. Table 1 and Figure 2 show that expanding access to earlier fibrosis stages saves approximately 5,800 to 15,100 additional lives relative to Status Quo. Including PWID in treatment access saves an additional 12,900 to 41,200 lives over the Status Quo. Reductions in the number of new infections and cases of decompensated cirrhosis and hepatocellular carcinoma are also greatest when PWID are included in treatment access.
Cost Savings to Payers Over time, the savings of expanding treatment access outweigh the medical and treatment costs. Figure 3A reports the cumulative costs under the five alternative policy scenarios relative to the Status Quo. Under all scenarios, there is a large initial
(B)
Fig. 1 – (A) HCV prevalence in the United States under different policy scenarios. (B) Annual HCV incidence in the United States under different policy scenarios. Notes. Point at which HCV-infected population decreased by 50% indicated by black diamonds. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. HCV, hepatitis C virus; PWID, people who inject drugs.
4
VALUE IN HEALTH ] (]]]]) ]]]–]]]
Table 1 – Differences in new infections, cases of decompensated cirrhosis, hepatocellular carcinoma, and number of deaths among hepatitis C virus–infected population over 20-y horizon relative to Status Quo with and without drug use restrictions. Outcomes
Number of new infections Cases of DC
Cases of HCC
Number of deaths
Measure
Total Difference % difference Total Difference % difference Total Difference % difference Total Difference % difference
Status Quo
497,149
1,018,844
530,187
462,821
Policies with drug use restrictions (excluding PWID)
Policies without drug use restrictions (including PWID)
Partial Expansion
Treat All
Status Quo including PWID
Partial Expansion including PWID
Treat All including PWID
475,773 21,376 4.3% 980,531 38,313 3.8% 478,585 51,602 9.7% 457,028 5,793 1.3%
435,161 61,988 12.5% 974,177 44,667 4.4% 470,444 59,743 11.3% 447,748 15,073 3.3%
470,169 26,980 5.4% 956,901 61,943 6.1% 523,662 6,525 1.2% 449,876 12,945 2.8%
389,479 107,670 21.7% 898,165 120,679 11.8% 452,525 77,662 14.6% 438,811 24,010 5.2%
224,360 272,789 54.9% 886,179 132,665 13.0% 439,205 90,982 17.2% 421,589 41,232 8.9%
Notes. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. DC, decompensated cirrhosis; HCC, hepatocellular carcinoma; PWID, people who inject drugs.
investment; the largest investment is in the scenario that expands treatment both to earlier fibrosis stages and to PWID. The break-even point—the time it takes for net benefits from alternative scenarios to equal and surpass net benefits under Status Quo—is reached sooner in scenarios that treat earlier fibrosis stages but restrict access to PWID. Status Quo and Partial Expansion including PWID scenarios do not reach the break-even point by 20 years, indicating they are more costly than the Status Quo and have negative savings for payers. Although expanding
treatment to all fibrosis stages has greater initial costs compared with other scenarios, the cost savings at the end of the simulation are positive in Treat All scenarios, both excluding and including PWID.
Net Social Benefits In Figure 3B, we explore the value of expanding access to HCV treatment from society’s perspective. Under all scenarios, the cumulative net benefits are negative in the first few years because of the increased number of people treated. In Partial Expansion and Treat All scenarios, excluding and including PWID, net benefits exceed the Status Quo within 4 to 6 years. A break-even analysis illustrates that scenarios that include PWID reach the break-even point 1 to 2 years later than scenarios that exclude PWID because of increased costs of treating more patients with HCV infection. The incremental net benefits of Partial Expansion and Treat All scenarios including PWID are equal to and surpass the net benefits of the same scenarios excluding PWID after approximately 10 years; the total cumulative net benefits relative to Status Quo are 15% to 17% higher after 20 years when treating PWID compared with not treating them. Our model predicts that the incremental net benefits of expanding treatment increase more rapidly under Treat All, suggesting that this policy maximizes the societal benefit of reduced infectivity. This benefit is more pronounced when PWID are eligible for treatment.
Cumulative Benefits and Costs
Fig. 2 – Cumulative deaths averted among hepatitis C virus– infected population relative to Status Quo over a 20-year horizon. Notes. Deaths averted are computed as the cumulative deaths in each alternative scenario minus the cumulative deaths in the Status Quo scenario over 20 years. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. PWID, people who inject drugs.
Figure 4A summarizes the overall impacts of increasingly inclusive treatment access scenarios, excluding PWID. There are higher treatment costs in more inclusive scenarios, but these are offset by greater savings in short-term disability, medical expenditures, QALY gains, and hence total net social benefits. Treatment costs are greater in scenarios that include PWID (Fig. 4B), but so are the benefits. The highest net social benefits are seen in the Treat All including PWID scenario amounting to approximately $500 billion over 20 years.
VALUE IN HEALTH ] (]]]]) ]]]–]]]
(A)
5
(B)
Fig. 3 – (A) Cumulative treatment and medical cost savings relative to Status Quo. (B) Cumulative net benefits to society relative to Status Quo. Notes. Cumulative net benefits are relative to Status Quo scenario. Dollar values discounted at 3% annual rate. Break-even points indicated by black diamonds. QALYs valued at $150,000/QALY. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. PWID, people who inject drugs; QALY, quality-adjusted life-year.
Incremental Cost-Effectiveness Table 2 reports ICERs for each alternative policy scenario relative to the Status Quo under 5-, 10-, 15-, and 20-year horizons. Although costs are greater in more inclusive scenarios, QALYs are also higher, resulting in lower ICERs for policy scenarios that expand treatment to earlier fibrosis stages and to PWID. After 5 years, Treat All including PWID is the only scenario with an ICER below $50,000/QALY. After 20 years, all expanded treatment scenarios are cost-effective, and Partial Expansion, Treat All, and Treat All including PWID are cost saving compared with the Status Quo.
Needle-Syringe Programs We find that when NSPs are added to different treatment scenarios that include PWID, the clinical benefits are superior to the same scenarios without NSPs. HCV incidence declines at a faster rate when NSPs are added. In the first 10 years, the total number of new infections among all cohorts is 10% lower with the combined NSP and Partial Expansion scenario and 26% lower with NSPs under Treat All, relative to the same scenarios without NSP. Relative to no NSPs, the total number of lives saved within the first 10 years is 16% greater in all scenarios compared with no NSP; the number of lives saved increases with expanded treatment access (range 7,600-14,000 lives). We also find that when treatment access is expanded to all fibrosis stages, total cost savings are positive after 19 years. Including NSPs results in higher initial costs, but yields greater net benefits at the end of the 20-year simulation in Partial Expansion ($218 billion) and Treat All ($527 billion) scenarios, compared with the same scenarios with no NSP. Expanding treatment access to earlier fibrosis stages and PWID and providing access to NSPs to minimize the spread of HCV have a dramatic effect on the
number of infections, deaths averted, and net benefits; the greatest impact can be seen in the Treat All scenario when NSPs are added.
Sensitivities Appendix Figure 2A,B in Supplemental Materials found at http:// dx.doi.org/10.1016/j.jval.2017.01.015 report the net benefits for Treat All scenarios (excluding and including PWID) relative to Status Quo under deterministic parameter variations. The model is most sensitive to variations in the percent of people insured and rate of diagnosis. All sensitivity analyses yield positive net social benefits, where the Treat All scenario results range from $213 billion to $503 billion (excluding PWID) and $246 billion to $625 billion (including PWID) at the end of the 20-year simulation. Appendix Figure 3 in Supplemental Materials found at http://dx. doi.org/10.1016/j.jval.2017.01.015 demonstrates the effect that varying the value of a QALY (dollar/QALY) has on net benefits relative to Status Quo. Qualitatively, there are no major differences in the outcome. Even at the lowest level of $50,000/QALY, the treatment of all fibrosis stages (excluding and including PWID) generates net social benefits of $158 billion to $170 billion after a 20-year horizon, respectively.
Conclusions Limiting HCV treatment to later stages of disease leads to increased disease transmission and more people progressing to advanced liver disease, increasing future costs to payers. This article extends the literature by explicitly modeling the trade-offs society makes when high-risk vulnerable populations, particularly PWID, are excluded from HCV treatment access. An innovation of our model is that it allows for interaction and
6
VALUE IN HEALTH ] (]]]]) ]]]–]]]
(A)
(B)
Fig. 4 – (A) Cumulative benefits and costs to society over a 20-year horizon relative to Status Quo; treatment access policies excluding people who inject drugs. (B) Cumulative benefits and costs to society over a 20-year horizon relative to Status Quo; treatment access policies including people who inject drugs. Notes. Cumulative savings are relative to Status Quo scenario and are discounted at 3% annual rate. QALYs valued at $150,000/QALY. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. PWID, people who inject drugs; QALY, quality-adjusted life-year. transmission between high-risk groups, where previous studies have kept these groups mutually exclusive, ignoring the increased infectivity from such interactions [22,23].
We find that expanding treatment to earlier fibrosis stages and including PWID provides significant benefit to society due to reduced infectivity and transmission, manifested in greater
Table 2 – Incremental cost-effectiveness ratios of alternative policies. Policy scenario
After 5 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY) After 10 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY) After 15 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY) After 20 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY)
Polices with drug use restrictions (excluding PWID)
Policies without drug use restrictions (including PWID)
Partial Expansion
Treat All
Partial Expansion
Treat All
995,026 40,408 0.26 158,157
1,904,710 21,109 0.41 51,569
1,283,028 31,337 0.22 140,466
2,509,265 16,023 0.33 48,222
1,253,478 18,291 0.52 34,897
2,005,964 23,202 0.88 26,315
1,688,413 30,823 0.47 65,594
2,833,079 33,783 0.72 47,164
1,393,861 3,215 0.70 4,585
2,013,839 60 1.20 50
1,922,099 16,306 0.63 25,757
2,927,548 11,976 0.98 12,182
1,470,043 6,331 0.81 Cost saving
2,017,048 12,369 1.38 Cost saving
2,063,787 5,864 0.73 8,013
2,965,534 1,626 1.15 Cost saving
Notes. ICER computed as the change in total costs divided by the change in QALYs relative to Status Quo. Costs include direct costs (HCV treatment costs and other medical expenditures). Changes are discounted at a 3% discount rate. Cost values are reported in 2015$. “Cost saving” indicates a reduction in costs and an increase in QALYs relative to Status Quo. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-year; pp, per person; PWID, people who inject drugs.
VALUE IN HEALTH ] (]]]]) ]]]–]]]
quality of life and reduced medical expenditures and mortality. This finding is consistent with existing literature, which also states that treating HCV-infected individuals at earlier fibrosis stages, or treating all patients with HCV infection, is more costeffective compared with the Status Quo [42,43]. The literature also supports our finding that the costs for treating HCV at earlier fibrosis stages are offset by averted costs of care in later stages of disease [42]. Our model predicts that more inclusive scenarios lead to faster disease eradication and the rate of decline in prevalence and incidence is faster when treatment access includes PWID. Treating PWID far exceeds the effects of expanding access to lower fibrosis stages on the prevalence and incidence of HCV infection and the cases of decompensated liver disease and death. Although substantial barriers to treatment exist for PWID, evidence shows that HCV treatment is safe and effective for PWID [44]. Findings suggest that PWID exhibit comparable HCV treatment outcomes compared with groups with other means of transmission [45,46]. In addition, PWID who successfully achieve sustained virologic response are more likely to lead a healthier lifestyle compared with before treatment [47]. Including PWID in access to HCV treatment can also have significant public health benefits. Friedman et al. [48] found that higher metropolitan HIV prevalence among PWID and MSM is positively related to AIDS incidence and AIDS-related mortality among heterosexual non-PWID. Other studies have found that HIV may be transmitted from PWID and MSM to heterosexuals through people who use illicit drugs by routes other than injection [48–51]. This supports our inclusion of the Overlap group and intercohort transmission functions. Friedman et al’s findings suggest that a combination of harm reduction programs and HIV testing for PWID is linked to fewer new AIDS cases in heterosexuals [48]. These findings have important implications for the current study. Higher prevalence of HCV infection among PWID means higher prevalence, costs, and disease burden among the community as a whole. Furthermore, guidelines from international organizations, including the World Health Organization, the American Association for the Study of Liver Diseases, the Infection Diseases Society of America, and the International Network for Hepatitis in Substance Users, have all recommended including PWID in HCV treatment access [52–54], to both reduce disease in these persons and prevent onward transmission to others. Excluding PWID from HCV treatment is shortsighted and a major impediment to eradicating HCV in the U.S. population [55]. Because the majority of incident HCV cases occur among PWID, focusing HCV prevention and eradication efforts on this high-risk population is necessary and strongly supported by the literature [56–58]. As is evidenced by our findings, as well as the existing literature [59], the benefits accrued from long-term avoidance of complications and increased medical costs far outweigh the cost of direct-acting antiviral agents. We considered the impacts of expanding HCV treatment to PWID when other strategies to reduce the transmission of HCV among PWID are also implemented. Specifically, we considered alternatives that extend access to treatment to PWID and also include an NSP to reduce the spread of HCV. We find that simply adding an NSP to the Status Quo reduced infections and saved lives but did not yield positive net social benefits; expanding treatment to earlier stages had more of an impact on cost savings and net benefits to society. In this article, we have shown that expanding HCV treatment to all stages of the disease and PWID may lead to significant social benefits. However, in our model, treatment and medical cost savings do not break-even relative to the Status Quo for 13 years. This suggests a misalignment between social benefits of an inclusive policy and the short-term costs faced by payers. To
7
realize the benefits of inclusive HCV treatment, public policies may be required to realign these incentives. As in any modeling-based analysis, our approach has some limitations and our findings should be viewed in the context of these limitations. Markov models are designed to capture cohortlevel effects and therefore cannot forecast individual disease processes and outcomes. We rely on parameter estimates from the literature, from different studies and populations. Although we made an effort to take parameters from sources with similar underlying populations, variability across the underlying populations is inevitable. We mitigate this by reporting changes relative to the Status Quo and conducting sensitivity analyses around key model parameters. We rely on NHANES data for estimates of the starting population. Although NHANES provides reliable populationlevel estimates, subpopulation estimates are less reliable due to small sample sizes. In addition, because of NHANES’ selfreporting design, it is possible that stigmatized behaviors, such as sexual activities and injection drug use, are underreported, which would affect subpopulation estimates. However, NHANESbased estimates are similar to other estimates reported in the literature. Finally, NHANES also does not capture homeless and incarcerated populations, both of which have high HCV prevalence and limited treatment access. We also rely on parameter estimates from the literature for the United States as a whole and recognize that there may be important regional differences within the United States. In future research, this model may be adapted to focus on regions within the United States that are particularly affected by injection drug use and HCV. A key limitation of our model is that it does not account for growth in opioid use over time. Although we assume a constant population within the groups for model tractability, increased opioid use, and thus increased infectivity due to needle-sharing, has created a spike in the population size for the PWID and OV groups, which our model does not take into account. However, our results are likely a lower estimate on the costs of treating PWID and on the benefits of reducing infectivity among this population. Although we evaluate access policies within a U.S. framework, our results may inform policy outside the United States, even in countries where cost-effectiveness thresholds are lower (e.g., £20,000-30,000/QALY in the United Kingdom). In particular, our analysis indicates that a 5-year horizon is sufficient for the treatment of all fibrosis stages including PWID to be costeffective at a $50,000/QALY threshold. Moreover, this strategy is cost-effective at the $15,000/QALY threshold after 15 years and cost saving after 20 years. In conclusion, increased access to HCV treatment reduces the prevalence and incidence of HCV infection, as well as the associated mortality and downstream medical costs. There is significant gain in societal value by expanding treatment to early stages of liver fibrosis among HCV-infected patients. Furthermore, society also gains significantly from expanding treatment to PWID through decreased infections, which reduce the cost of treatment and other medical expenditures, improve quality of life, and reduce mortality. The value of expanded treatment access is being recognized in the United States, with major private payers, Medicaid, and Medicare beginning to provide access to patients in earlier stages of liver fibrosis. However, many payers continue to impose access restrictions to PWID who are actively injecting drugs. As our findings suggest, the potential benefits to society of treating this high-risk population far exceed the costs of providing treatment. Source of financial support: The design, analysis, and financial support of this study were provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the study.
8
VALUE IN HEALTH ] (]]]]) ]]]–]]]
Supplemental Materials Supplemental material accompanying this article can be found in the online version as a hyperlink at http://dx.doi.org/10.1016/j. jval.2017.01.015 or, if a hard copy of article, at www.valueinhealth journal.com/issues (select volume, issue, and article).
R EF E R EN CE S
[1] Armstrong GL, Wasley A, Simard EP, et al. The prevalence of hepatitis C virus infection in the United States, 1999 through 2002. Ann Intern Med 2006;144:705–14. [2] Biggins SW, Bambha KM, Terrault NA, et al. Projected future increase in aging hepatitis C virus-infected liver transplant candidates: a potential effect of hepatocellular carcinoma. Liver Transplant 2012;18:1471–8. [3] Centers for Disease Control and Prevention. Hepatitis C information for health professionals. 2015. http://www.cdc.gov/hepatitis/HCV/. [Accessed January 20, 2015]. [4] National Health and Nutrition Examination Survey Data [Waves 20032004 through 2011-2012]. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2010. Available from: https://www.cdc.gov/nchs/nhanes/. [Accessed November 1, 2015]. [5] Edlin BR, Eckhardt BJ, Shu MA, et al. Toward a more accurate estimate of the prevalence of hepatitis C in the United States. Hepatology 2015;62:1353–63. [6] Yehia BR, Schranz AJ, Umscheid CA, Lo Re V III. The treatment cascade for chronic hepatitis C virus infection in the United States: a systematic review and meta-analysis. PloS One 2014;9:e101554. [7] Denniston MM, Klevens RM, McQuillan GM, Jiles RB. Awareness of infection, knowledge of hepatitis C, and medical follow‐up among individuals testing positive for hepatitis C: National Health and Nutrition Examination Survey 2001–2008. Hepatology 2012;55:1652–61. [8] Smith BD, Morgan RL, Beckett GA, et al. Recommendations for the identification of chronic hepatitis C virus infection among persons born during 1945-1965. MMWR Recomm Rep 2012;61:1–32. [9] Younossi ZM, Singer ME, Mir HM, et al. Impact of interferon free regimens on clinical and cost outcomes for chronic hepatitis C genotype 1 patients. J Hepatol 2014;60:530–7. [10] Williams IT, Bell BP, Kuhnert W, Alter MJ. Incidence and transmission patterns of acute hepatitis C in the United States, 1982-2006. Arch Intern Med 2011;171:242–8. [11] Wasley A, Gallagher KM, Grytdal S. Surveillance for Acute Viral Hepatitis, United States, 2006. Department of Health and Human Services, Centers for Disease Control and Prevention. March 21, 2008; 57, No. SS-2. [12] Suryaprasad AG, White JZ, Xu F, et al. Emerging epidemic of hepatitis C virus infections among young nonurban persons who inject drugs in the United States, 2006-2012. Clin Infect Dis 2014;59:1411–9. [13] Nelson PK, Mathers BM, Cowie B, et al. Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews. Lancet (London, England) 2011;378:571–83. [14] Crofts N, Aitken CK, Kaldor JM. The force of numbers: why hepatitis C is spreading among Australian injecting drug users while HIV is not. Med J Australia 1999;170:220–1. [15] Bradshaw D, Matthews G, Danta M. Sexually transmitted hepatitis C infection: the new epidemic in MSM? Curr Opinion Infect Dis 2013;26:66–72. [16] Yaphe S, Bozinoff N, Kyle R, et al. Incidence of acute hepatitis C virus infection among men who have sex with men with and without HIV infection: a systematic review. Sexually Transmitted Infect 2012;88:558–64. [17] Centers for Disease Control and Prevention. HIV and viral hepatitis. 2014. Available from: http://www.cdc.gov/hiv/pdf/library_factsheets_ hiv_and_viral_hepatitis.pdf. [September 1, 2016]. [18] Taylor LE, Holubar M, Wu K, et al. Incident hepatitis C virus infection among US HIV-infected men enrolled in clinical trials. Clin Infect Dis 2011;52:812–8. [19] Canary LA, Klevens RM, Holmberg SD. Limited access to new hepatitis C virus treatment under state Medicaid programs. Ann Intern Med 2015;163:226–8. [20] Anthem Blue Cross and Blue Shield. Medication quantity limit Sovaldi. 2015. https://www.anthem.com/provider/noapplication/f0/s0/t0/ pw_e210963.pdf. [Accessed October 7, 2015]. [21] Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993;13:322–38.
[22] Van Nuys K, Brookmeyer R, Chou JW, et al. Broad hepatitis C treatment scenarios return substantial health gains, but capacity is a concern. Health Aff 2015;34:1666–74. [23] Linthicum MT, Sanchez Gonzales Y, Mulligan K, et al. Value of expanding HCV screening and treatment policies in the United States. Am J Managed Care 2016;22:SP227–35. [24] Witt MD, Seaberg EC, Darilay A, et al. Incident hepatitis C virus infection in men who have sex with men: a prospective cohort analysis, 1984–2011. Clin Infect Dis 2013;57:77–84. [25] Smith BD, Morgan RL, Beckett GA, et al. Recommendations for the identification of chronic hepatitis C virus infection among persons born during 1945-1965. MMWR Recomm Rep 2012;61:1–32. [26] Manos MM, Shvachko VA, Murphy RC, et al. Distribution of hepatitis C virus genotypes in a diverse US integrated health care population. J Med Virol 2012;84:1744–50. [27] UnitedHealthcare. UnitedHealthcare Pharmacy, Clinical Pharmacy Programs, Prior Authorization/Medical Necessity Harvoni™ (ledipasvir/ sofosbuvir). 2014. https://www.unitedhealthcareonline.com/ ccmcontent/ProviderII/UHC/en-US/Assets/ProviderStaticFiles/ ProviderStaticFilesPdf/ToolsandResources/PharmacyResources/ Sovaldi_Medical_Necessity.pdf. [Accessed October 7, 2015]. [28] Aetna. Pharmacy clinical policy bulletins, Aetna non-Medicare prescription drug plan: hepatitis C. 2015. http://www.aetna.com/ products/rxnonmedicare/data/2014/GI/hepatitis_c.html. [Accessed October 7, 2015]. [29] Anthem. Viekira Pak (ombitasvir/paritaprevir/ritonavir þ dasabuvir). 2015. https://www.anthem.com/provider/noapplication/f0/s0/t0/ pw_e210963.pdf?na=pharminfo. [Accessed October 7, 2015]. [30] Health Net. Prior authorization protocol: SOLVALDI™ (sofosbuvir). 2015. https://www.healthnet.com/static/general/unprotected/html/national/ pa_guidelines/sovaldi_natl.html. [Accessed October 7, 2015]. [31] Canary LA, Klevens RM, Holmberg SD. Limited access to new hepatitis C virus treatment under state Medicaid programs. Ann Intern Med 2015;163:226–8. [32] Barua S, Greenwald R, Grebely J, et al. Restrictions for Medicaid reimbursement of sofosbuvir for the treatment of hepatitis C virus infection in the United States. Ann Intern Med 2015;163:215–23. [33] Ghany MG, Strader DB, Thomas DL, Seeff LB. Diagnosis, management, and treatment of hepatitis C: an update. Hepatology 2009;49:1335–74. [34] Mathers BM, Degenhardt L, Ali H, et al. HIV prevention, treatment, and care services for people who inject drugs: a systematic review of global, regional, and national coverage. Lancet (London, England) 2010;375:1014–28. [35] Wodak A, Cooney A. Do needle syringe programs reduce HIV infection among injecting drug users: a comprehensive review of the international evidence. Substance Use Misuse 2006;41:777–813. [36] Edlin BR. Perspective: test and treat this silent killer. Nature 2011;474:S18–9. [37] Health Canada. Vancouver’s INSITE service and other supervised injection sites: what has been learned from research? Health Canada, March 31, 2008. Available from: http://www.hc-sc.gc.ca/ahc-asc/pubs/ _sites-lieux/insite/index-eng.php. [Accessed September 1, 2016]. [38] Honeycutt AA, Harris JL, Khavjou O, et al. The costs and impacts of testing for hepatitis C virus antibody in public STD clinics. Public Health Rep 2007;122(Suppl 2):55–62. [39] Green TC, Hankins CA, Palmer D, et al. My place, your place, or a safer place: the intention among Montreal injecting drug users to use supervised injecting facilities. Can J Public Health 2004;95(2):110–4. [40] Wood E, Tyndall MW, Montaner JS, Kerr T. Summary of findings from the evaluation of a pilot medically supervised safer injecting facility. Can Med Assoc J 2006;175:1399–404. [41] Hirth RA, Chernew ME, Miller E, et al. Willingness to pay for a qualityadjusted life year in search of a standard. Med Decis Making 2000;20:332–42. [42] Chahal HS, Marseille EA, Tice JA, et al. Cost-effectiveness of early treatment of hepatitis C virus genotype 1 by stage of liver fibrosis in a US treatment-naive population. JAMA Intern Med 2016;176:65–73. [43] Visconti AJ, Doyle JS, Weir A, et al. Assessing the cost‐effectiveness of treating chronic hepatitis C virus in people who inject drugs in Australia. J Gastroenterol Hepatol 2013;28:707–16. [44] Edlin BR, Kresina TF, Raymond DB, et al. Overcoming barriers to prevention, care, and treatment of hepatitis C in illicit drug users. Clin Infect Dis 2005;40:S276–85. [45] Aspinall E, Corson S, Doyle J, et al. Peginterferon and ribavirin treatment for chronic hepatitis C virus in people who inject drugs: a systematic review and meta-analysis. Clin Infect Dis 2013;57:80–9. [46] Dimova RB, Zeremski M, Jacobson IM, et al. Determinants of hepatitis C virus treatment completion and efficacy in drug users assessed by meta-analysis. Clin Infect Dis 2013;56:806–16. [47] Innes HA, McDonald SA, Dillon JF, et al. Toward a more complete understanding of the association between a hepatitis C sustained viral response and cause‐specific outcomes. Hepatology 2015;62:355–64. [48] Friedman SR, West BS, Tempalski B, et al. Do metropolitan HIV epidemic histories and programs for people who inject drugs and men
VALUE IN HEALTH ] (]]]]) ]]]–]]]
[49]
[50]
[51] [52]
[53]
who have sex with men predict AIDS incidence and mortality among heterosexuals? Ann Epidemiol 2014;24:304–11. Neaigus A, Miller M, Gyarmathy VA, Friedman SR. HIV heterosexual sexual risk from injecting drug users among HIV-seronegative noninjecting heroin users. Substance Use Misuse 2011;46:208–17. Strathdee SA, Stockman JK. Epidemiology of HIV among injecting and non-injecting drug users: current trends and implications for interventions. Curr HIV/AIDS Rep 2010;7:99–106. Volz E, Frost SD, Rothenberg R, Meyers LA. Epidemiological bridging by injection drug use drives an early HIV epidemic. Epidemics 2010;2:155–64. AASLD/IDSA. HCV guidance: recommendations for testing, managing, and treating HCV. 2016. http://www.hcvguidelines.org/. [Accessed June 15, 2016]. World Health Organization. Guidelines for the Screening Care and Treatment of Persons With Hepatitis C Infection. Geneva, Switzerland: World Health Organization, 2014.
9
[54] Robaeys G, Grebely J, Mauss S, et al. Recommendations for the management of hepatitis C virus infection among people who inject drugs. Clin Infect Dis 2013;57:S129–37. [55] Leask J, Dillon J. Review article: treatment as prevention–targeting people who inject drugs as a pathway towards hepatitis C eradication. Aliment Pharmacol Therapeut 2016;44(2):145–56. [56] Hagan H, Pouget ER, Des Jarlais DC. A systematic review and metaanalysis of interventions to prevent hepatitis C virus infection in people who inject drugs. J Infect Dis 2011;204:74–83. [57] Hellard M, Doyle JS, Sacks‐Davis R, et al. Eradication of hepatitis C infection: the importance of targeting people who inject drugs. Hepatology 2014;59:366–9. [58] Hellard M, McBryde E, Davis RS, et al. Hepatitis C transmission and treatment as prevention–the role of the injecting network. Int J Drug Policy 2015;26:958–62. [59] Hagan LM, Wolpe PR, Schinazi RF. Treatment as prevention and cure towards global eradication of hepatitis C virus. Trends Microbiol 2013;21:625–33.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jval
Value of Comprehensive HCV Treatment among Vulnerable, High-Risk Populations Gigi A. Moreno, PhD1, Alice Wang, MA2, Yuri Sánchez González, PhD2, Oliver Díaz Espinosa, PhD1,*, Diana K. Vania, MSc1, Brian R. Edlin, MD3, Ronald Brookmeyer, PhD4 1 Precision Health Economics, Los Angeles, CA, USA; 2AbbVie, Inc., Mettawa, IL, USA; 3Weill Cornell Medical College, Cornell University, New York City, NY, USA; 4Fielding School of Public Health, University of California Los Angeles, Los Angeles, CA, USA
AB STR A CT
Objectives: The objective of this study was to explore the trade-offs society and payers make when expanding treatment access to patients with chronic hepatitis C virus (HCV) infection in early stages of disease as well as to vulnerable, high-risk populations, such as people who inject drugs (PWID) and HIV-infected men who have sex with men (MSM-HIV). Methods: A discrete time Markov model simulated HCV progression and treatment over 20 years. Population cohorts were defined by behaviors that influence the risk of HCV exposure: PWID, MSM-HIV, an overlap cohort of individuals who are both PWID and MSM-HIV, and all other adults. Six different treatment scenarios were modeled, with varying degrees of access to treatment at different fibrosis stages and to different risk cohorts. Benefits were measured as quality-adjusted lifeyears and a $150,000/quality-adjusted life-year valuation was used to assess social benefits. Results: Compared with limiting treatment to METAVIR fibrosis stages F3 or F4 and excluding PWID, expanding treatment to patients in all fibrosis stages and including PWID reduces
cumulative new infections by 55% over a 20-year horizon and reduces the prevalence of HCV by 93%. We find that treating all HCV-infected individuals is cost saving and net social benefits are over $500 billion greater compared with limiting treatment. Including PWID in treatment access saves 12,900 to 41,200 lives. Conclusions: Increased access to treatment brings substantial value to society and over the long-term reduces costs for payers, as the benefits accrued from long-term reduction in prevalent and incident cases, mortality, and medical costs outweigh the cost of treatment. Keywords: hepatitis C, people who inject drugs, men who have sex with men, Markov model.
Introduction
disease progression, which will reduce morbidity, mortality, and health care expenditures related to cirrhosis, decompensated liver disease, hepatocellular carcinoma, and liver transplants [9]. The most significant mode of transmission for HCV infection in the United States is through injection drug use [10]; approximately 54% to 77% of new HCV diagnoses are among people who inject drugs (PWID) [11,12]. An estimated 3.5 million people have injected drugs in the United States during their lifetime [4], with the prevalence of HCV infection in this population estimated at 73% (range 70%–77%) [13]. Because of the high probability of contracting HCV infection through needle-sharing [14], treating PWID infected with HCV, particularly in early stages of the disease, may reduce transmission. Although the probability of HCV transmission through sexual contact is low, it is believed that HIV co-infection may aid the
Hepatitis C virus (HCV) infection is a transmissible viral infection that is usually asymptomatic in the early stages of the disease and may lead to serious liver complications including decompensated cirrhosis and hepatocellular carcinoma, which are associated with a significant mortality and health care and cost burden [1–3]. It is estimated that at least 3 to 5 million people are infected with HCV in the United States [4,5]; however, approximately half are unaware of their infection due to the asymptomatic nature of early stages of the disease [6–8]. Comprehensive HCV treatment may have important societal benefits beyond the immediate benefit to the patient. With each successfully treated patient, the pool of individuals able to transmit the disease decreases. In addition, early treatment for HCV has the potential to hinder
Copyright & 2017, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Conflict of interest: G. A. Moreno and D. K. Vania were employees of Precision Health Economics, LLC, at the time the manuscript was prepared. A. Wang and Y. S. Gonzalez are employees of AbbVie Inc. and may own AbbVie stock or stock options. O. D. Espinosa is an employee of Precision Health Economics, LLC, which received payment from AbbVie for conducting research analysis. B. R. Edlin and R. Brookmeyer are consultants to Precision Health Economics, LLC. * Address correspondence to: Oliver Díaz Espinosa, Precision Health Economics, 11100 Santa Monica Boulevard, Suite 500, Los Angeles, CA 90025. E-mail: [email protected]. 1098-3015$36.00 – see front matter Copyright & 2017, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). http://dx.doi.org/10.1016/j.jval.2017.01.015
2
VALUE IN HEALTH ] (]]]]) ]]]–]]]
infectivity of HCV [15]. The incidence rate of HCV infection among men who have sex with men (MSM) who are infected with human immunodeficiency virus (MSM-HIV) is four times greater than that among HIV-uninfected MSM [16]. Furthermore, it has been estimated that approximately 25% of HIV-infected individuals are coinfected with HCV [17]. The occurrence of overlapping high-risk behaviors (e.g., MSM-HIV who also inject drugs) may further increase the risk of HCV infection. Taylor et al. [18] observe that MSM-HIV who report injection drug use are more than six times as likely to acquire HCV infection compared with those who do not engage in injection drug use [18]. Thus, MSM-HIV are a significant source of HCV transmission given the interaction with PWID and that HIV infection may increase infectivity of HCV. Despite the demonstrated efficacy of new interferon-free direct-acting antiviral agents, access to these treatments has been limited in the United States to those in advanced fibrosis stages or with HIV co-infection [19,20]. In addition, many payers have policies that deny treatment to HCV-infected PWID. The objective of this study was to explore the trade-offs society and payers make when access is expanded to patients in early stages of disease as well as to vulnerable populations such as PWID and MSM-HIV, taking into account transmission between these groups. Herein, we describe the benefits and costs to society under these expanded treatment access scenarios and consider their effects on the total number of HCV infections in the United States. We also consider how these trade-offs are affected when treatment access for PWID is combined with disease prevention programs, such as needle-syringe programs (NSPs).
Methods We developed a discrete time Markov model using the R programming language to simulate HCV progression and treatment over a time horizon of 20 years [21]. In each yearly cycle, the population transitions through each disease state at assumed probabilities (transition probabilities are reported in Supplemental Materials found at 10.1016/j.jval.2017.01.015). The structure of the model is described in detail in Van Nuys et al. [22] and in Appendix Figure 1 in Supplemental Materials found at http://dx. doi.org/10.1016/j.jval.2017.01.015.
the simulation. Because we assumed no ongoing infections among OA, the size of this cohort shrinks during the simulation. Mortality rates for the risk cohorts are computed as described in Linthicum et al. [23]. Within each risk cohort, we modeled the three most common HCV genotypes in the United States (genotypes 1, 2, and 3) [26]. We assumed that individuals may be infected with only one genotype at a time, but if cured and reinfected, they may re-enter the model with any of the three genotypes. Specifically, the cohort cured of HCV in early fibrosis stages (F0-F2) returns to the susceptible uninfected population. Those cured in later stages (F3 and higher) may progress to additional liver damage, albeit at a slower rate than the infected population. If re-infected, the cohort cured in later fibrosis stages re-enters the infected population at the highest fibrosis stage attained by that cohort (see Appendix Figure 1).
Model Parameters HCV treatment costs, medical expenditures (nontreatment medical costs), quality-of-life weights, and mortality rates were collected from estimates in the published literature. All cost parameters were inflated to 2015 U.S. dollars, and future costs and quality-adjusted life-years (QALYs) were discounted at an annual 3% rate. Treatment costs vary by genotype and fibrosis stage, where the late-stage genotype 2 cohort (F3 and higher) and the entire genotype 3 cohort experience higher treatment costs than genotype 1 because of differing treatment durations, consistent with a previous study [22]. Medical expenditures and quality-of-life weights vary by disease state. We assumed that PWID have the same medical expenditures as the OA cohort and the OV cohort has the same medical expenditures as the MSMHIV cohort. Treatment efficacy, treatment costs, and medical expenditures do not vary over time, with the exception of treatment costs, which are discounted 79% after 16 years to account for patent expiration and future market competition [22]. We conducted sensitivity analyses to test the effects of several key assumptions on our results. We tested the sensitivity of results by varying the assumed values of six key parameters: cost of treatment (⫾15%); percent insured (extreme 20%–90%); quality-of-life weights (⫾5%), rate of diagnosis (extreme 25%– 100%); starting population (⫾5%), and sustained virologic response rate (⫾15%).
Study Population To simulate real-world transmission of HCV, the initial population was stratified into risk cohorts, defined by behaviors that influence the risk of HCV exposure and subsequent infection. Previous studies have modeled mutually exclusive risk cohorts [22,23], potentially overlooking the additive risk of HCV infection among individuals who are both PWID and MSM-HIV [24]. Building upon previous studies, three high-risk groups and a low-risk, catchall group of all other adults transitioned through the Markov model. The risk cohorts include 1) individuals who self-identify as PWID but not MSM-HIV; 2) HIV-infected individuals who self-identify as MSM but not PWID; 3) the overlap (“OV”) of individuals who are both PWID and MSM-HIV; and 4) all other adults (“OA”) who were born before 1992 when systematic testing of the blood supply for HCV was introduced. We assumed there was no ongoing HCV transmission in the OA cohort [25]. Appendix Table 2 in Supplemental Materials found at http://dx.doi.org/10.1016/j.jval.2017.01.015 presents the distribution of the starting population by risk cohort. In the model, an individual belongs to one risk cohort for the duration of the simulation. As described in Supplemental Materials, our model allows interactions between the PWID and MSMHIV groups through contact with the OV group. We also assumed that the high-risk cohorts experienced ongoing entry and exit (mortality) such that their cohort sizes remained constant over
Policy Scenarios We simulated six treatment access scenarios: a Status Quo scenario and five alternative scenarios with varying degrees of expanded access to treatment. All scenarios are evaluated against the Status Quo (baseline) scenario, a stylized representation of treatment access policies prevailing in 2015 among the largest U.S. payers. In the Status Quo scenario, we assume that only those who do not inject drugs have access to treatment if their disease has advanced to METAVIR fibrosis stage F3 or F4 and if they are insured [27–32]. The METAVIR scoring system quantifies the degree of liver fibrosis in patients with liver diseases such as HCV [33]. In the first alternative scenario, we assume a partial expansion of treatment to patients in fibrosis stages F2 to F4, continuing to exclude PWID from treatment (“Partial Expansion”). In the second alternative scenario, treatment access is expanded to patients in all fibrosis stages (F0-F4; “Treat All”), excluding access to PWID. In the remaining three alternative scenarios, we expand the Status Quo, Partial Expansion, and Treat All scenarios to include PWID. A final analysis assessed the effects of NSPs as a possible strategy to reduce the transmission of HCV among PWID. These programs offer access to clean drug injection equipment for PWID and have been proven to reduce the spread of HIV [34,35]. Their evidence in reducing the spread of HCV is less clear, but HCV
VALUE IN HEALTH ] (]]]]) ]]]–]]]
incidence rates among PWID have fallen by an order of magnitude after the implementation of HIV prevention interventions, including NSPs [36]. To assess the possible impact of expanded NSPs combined with expanded access to treatment for the PWID population, we increased costs in our analysis to include the annual cost of the programs, estimated to be $1694 per user [37,38]. On the basis of current estimates of PWID who access or are willing to access an NSP, we assumed that 61% of the PWID population in our model accessed an NSP [39,40]. All other model assumptions remained the same. Only alternative treatment scenarios that include PWID were considered in the NSP analysis.
Model Outputs We compute epidemiological and clinical outcomes, including HCV prevalence, incidence, deaths averted, and cases of liver disease. We also estimate economic outputs, including cost savings to payers, net social benefits, and incremental costeffectiveness ratios (ICERs) for each alternative scenario relative to the Status Quo over a 20-year horizon. In each time period, net benefits are calculated as the difference between the benefits (QALYs valued at $150,000/QALY [41]) and total costs (treatment, disability, and medical expenditures). Unless otherwise noted, results for each alternative policy are given as changes relative to the Status Quo scenario.
Results HCV Prevalence Figure 1A depicts the simulated U.S. population infected with HCV starting from the most recent National Health and Nutrition Examination Survey (NHANES) estimates in 2012 [4]; under the Status Quo, the number of people with HCV infection decreases by 50% after approximately 11 years. A partial expansion in treatment to include patients in fibrosis stages F2 to F4 decreases the time it takes to reduce prevalence by 50% to 7 years. Further
(A)
3
expanding treatment to PWID in fibrosis stages F2 to F4 reduces the prevalence by half in 6 years. A policy of HCV treatment access for all fibrosis stages and PWID reduces HCV infections by 50% in 2 years.
HCV Incidence Figure 1B depicts the simulated number of new HCV infections per year in the U.S. population; under all treatment scenarios, the incidence of new HCV infections in the United States will decrease. In Partial Expansion and Treat All scenarios that exclude PWID from treatment, the number of new infections is 16,700 and 14,800 per year, respectively, at the end of the 20-year simulation. Including PWID in treatment access significantly reduces infections: in the Partial Expansion and Treat All scenarios including PWID, new infections decrease to 10,700 and 1,900 per year, respectively. This represents a 36% and 87% decrease, respectively, in the total number of new infections compared with scenarios that exclude PWID.
Deaths and Liver Disease Averted The benefits of expanding access to treat earlier fibrosis stages and PWID are also observed in the number of averted deaths and averted cases of liver disease relative to the Status Quo. Table 1 and Figure 2 show that expanding access to earlier fibrosis stages saves approximately 5,800 to 15,100 additional lives relative to Status Quo. Including PWID in treatment access saves an additional 12,900 to 41,200 lives over the Status Quo. Reductions in the number of new infections and cases of decompensated cirrhosis and hepatocellular carcinoma are also greatest when PWID are included in treatment access.
Cost Savings to Payers Over time, the savings of expanding treatment access outweigh the medical and treatment costs. Figure 3A reports the cumulative costs under the five alternative policy scenarios relative to the Status Quo. Under all scenarios, there is a large initial
(B)
Fig. 1 – (A) HCV prevalence in the United States under different policy scenarios. (B) Annual HCV incidence in the United States under different policy scenarios. Notes. Point at which HCV-infected population decreased by 50% indicated by black diamonds. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. HCV, hepatitis C virus; PWID, people who inject drugs.
4
VALUE IN HEALTH ] (]]]]) ]]]–]]]
Table 1 – Differences in new infections, cases of decompensated cirrhosis, hepatocellular carcinoma, and number of deaths among hepatitis C virus–infected population over 20-y horizon relative to Status Quo with and without drug use restrictions. Outcomes
Number of new infections Cases of DC
Cases of HCC
Number of deaths
Measure
Total Difference % difference Total Difference % difference Total Difference % difference Total Difference % difference
Status Quo
497,149
1,018,844
530,187
462,821
Policies with drug use restrictions (excluding PWID)
Policies without drug use restrictions (including PWID)
Partial Expansion
Treat All
Status Quo including PWID
Partial Expansion including PWID
Treat All including PWID
475,773 21,376 4.3% 980,531 38,313 3.8% 478,585 51,602 9.7% 457,028 5,793 1.3%
435,161 61,988 12.5% 974,177 44,667 4.4% 470,444 59,743 11.3% 447,748 15,073 3.3%
470,169 26,980 5.4% 956,901 61,943 6.1% 523,662 6,525 1.2% 449,876 12,945 2.8%
389,479 107,670 21.7% 898,165 120,679 11.8% 452,525 77,662 14.6% 438,811 24,010 5.2%
224,360 272,789 54.9% 886,179 132,665 13.0% 439,205 90,982 17.2% 421,589 41,232 8.9%
Notes. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. DC, decompensated cirrhosis; HCC, hepatocellular carcinoma; PWID, people who inject drugs.
investment; the largest investment is in the scenario that expands treatment both to earlier fibrosis stages and to PWID. The break-even point—the time it takes for net benefits from alternative scenarios to equal and surpass net benefits under Status Quo—is reached sooner in scenarios that treat earlier fibrosis stages but restrict access to PWID. Status Quo and Partial Expansion including PWID scenarios do not reach the break-even point by 20 years, indicating they are more costly than the Status Quo and have negative savings for payers. Although expanding
treatment to all fibrosis stages has greater initial costs compared with other scenarios, the cost savings at the end of the simulation are positive in Treat All scenarios, both excluding and including PWID.
Net Social Benefits In Figure 3B, we explore the value of expanding access to HCV treatment from society’s perspective. Under all scenarios, the cumulative net benefits are negative in the first few years because of the increased number of people treated. In Partial Expansion and Treat All scenarios, excluding and including PWID, net benefits exceed the Status Quo within 4 to 6 years. A break-even analysis illustrates that scenarios that include PWID reach the break-even point 1 to 2 years later than scenarios that exclude PWID because of increased costs of treating more patients with HCV infection. The incremental net benefits of Partial Expansion and Treat All scenarios including PWID are equal to and surpass the net benefits of the same scenarios excluding PWID after approximately 10 years; the total cumulative net benefits relative to Status Quo are 15% to 17% higher after 20 years when treating PWID compared with not treating them. Our model predicts that the incremental net benefits of expanding treatment increase more rapidly under Treat All, suggesting that this policy maximizes the societal benefit of reduced infectivity. This benefit is more pronounced when PWID are eligible for treatment.
Cumulative Benefits and Costs
Fig. 2 – Cumulative deaths averted among hepatitis C virus– infected population relative to Status Quo over a 20-year horizon. Notes. Deaths averted are computed as the cumulative deaths in each alternative scenario minus the cumulative deaths in the Status Quo scenario over 20 years. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. PWID, people who inject drugs.
Figure 4A summarizes the overall impacts of increasingly inclusive treatment access scenarios, excluding PWID. There are higher treatment costs in more inclusive scenarios, but these are offset by greater savings in short-term disability, medical expenditures, QALY gains, and hence total net social benefits. Treatment costs are greater in scenarios that include PWID (Fig. 4B), but so are the benefits. The highest net social benefits are seen in the Treat All including PWID scenario amounting to approximately $500 billion over 20 years.
VALUE IN HEALTH ] (]]]]) ]]]–]]]
(A)
5
(B)
Fig. 3 – (A) Cumulative treatment and medical cost savings relative to Status Quo. (B) Cumulative net benefits to society relative to Status Quo. Notes. Cumulative net benefits are relative to Status Quo scenario. Dollar values discounted at 3% annual rate. Break-even points indicated by black diamonds. QALYs valued at $150,000/QALY. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. PWID, people who inject drugs; QALY, quality-adjusted life-year.
Incremental Cost-Effectiveness Table 2 reports ICERs for each alternative policy scenario relative to the Status Quo under 5-, 10-, 15-, and 20-year horizons. Although costs are greater in more inclusive scenarios, QALYs are also higher, resulting in lower ICERs for policy scenarios that expand treatment to earlier fibrosis stages and to PWID. After 5 years, Treat All including PWID is the only scenario with an ICER below $50,000/QALY. After 20 years, all expanded treatment scenarios are cost-effective, and Partial Expansion, Treat All, and Treat All including PWID are cost saving compared with the Status Quo.
Needle-Syringe Programs We find that when NSPs are added to different treatment scenarios that include PWID, the clinical benefits are superior to the same scenarios without NSPs. HCV incidence declines at a faster rate when NSPs are added. In the first 10 years, the total number of new infections among all cohorts is 10% lower with the combined NSP and Partial Expansion scenario and 26% lower with NSPs under Treat All, relative to the same scenarios without NSP. Relative to no NSPs, the total number of lives saved within the first 10 years is 16% greater in all scenarios compared with no NSP; the number of lives saved increases with expanded treatment access (range 7,600-14,000 lives). We also find that when treatment access is expanded to all fibrosis stages, total cost savings are positive after 19 years. Including NSPs results in higher initial costs, but yields greater net benefits at the end of the 20-year simulation in Partial Expansion ($218 billion) and Treat All ($527 billion) scenarios, compared with the same scenarios with no NSP. Expanding treatment access to earlier fibrosis stages and PWID and providing access to NSPs to minimize the spread of HCV have a dramatic effect on the
number of infections, deaths averted, and net benefits; the greatest impact can be seen in the Treat All scenario when NSPs are added.
Sensitivities Appendix Figure 2A,B in Supplemental Materials found at http:// dx.doi.org/10.1016/j.jval.2017.01.015 report the net benefits for Treat All scenarios (excluding and including PWID) relative to Status Quo under deterministic parameter variations. The model is most sensitive to variations in the percent of people insured and rate of diagnosis. All sensitivity analyses yield positive net social benefits, where the Treat All scenario results range from $213 billion to $503 billion (excluding PWID) and $246 billion to $625 billion (including PWID) at the end of the 20-year simulation. Appendix Figure 3 in Supplemental Materials found at http://dx. doi.org/10.1016/j.jval.2017.01.015 demonstrates the effect that varying the value of a QALY (dollar/QALY) has on net benefits relative to Status Quo. Qualitatively, there are no major differences in the outcome. Even at the lowest level of $50,000/QALY, the treatment of all fibrosis stages (excluding and including PWID) generates net social benefits of $158 billion to $170 billion after a 20-year horizon, respectively.
Conclusions Limiting HCV treatment to later stages of disease leads to increased disease transmission and more people progressing to advanced liver disease, increasing future costs to payers. This article extends the literature by explicitly modeling the trade-offs society makes when high-risk vulnerable populations, particularly PWID, are excluded from HCV treatment access. An innovation of our model is that it allows for interaction and
6
VALUE IN HEALTH ] (]]]]) ]]]–]]]
(A)
(B)
Fig. 4 – (A) Cumulative benefits and costs to society over a 20-year horizon relative to Status Quo; treatment access policies excluding people who inject drugs. (B) Cumulative benefits and costs to society over a 20-year horizon relative to Status Quo; treatment access policies including people who inject drugs. Notes. Cumulative savings are relative to Status Quo scenario and are discounted at 3% annual rate. QALYs valued at $150,000/QALY. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. PWID, people who inject drugs; QALY, quality-adjusted life-year. transmission between high-risk groups, where previous studies have kept these groups mutually exclusive, ignoring the increased infectivity from such interactions [22,23].
We find that expanding treatment to earlier fibrosis stages and including PWID provides significant benefit to society due to reduced infectivity and transmission, manifested in greater
Table 2 – Incremental cost-effectiveness ratios of alternative policies. Policy scenario
After 5 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY) After 10 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY) After 15 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY) After 20 y Treated population Δ Costs ($) pp Δ QALYs pp ICER ($/QALY)
Polices with drug use restrictions (excluding PWID)
Policies without drug use restrictions (including PWID)
Partial Expansion
Treat All
Partial Expansion
Treat All
995,026 40,408 0.26 158,157
1,904,710 21,109 0.41 51,569
1,283,028 31,337 0.22 140,466
2,509,265 16,023 0.33 48,222
1,253,478 18,291 0.52 34,897
2,005,964 23,202 0.88 26,315
1,688,413 30,823 0.47 65,594
2,833,079 33,783 0.72 47,164
1,393,861 3,215 0.70 4,585
2,013,839 60 1.20 50
1,922,099 16,306 0.63 25,757
2,927,548 11,976 0.98 12,182
1,470,043 6,331 0.81 Cost saving
2,017,048 12,369 1.38 Cost saving
2,063,787 5,864 0.73 8,013
2,965,534 1,626 1.15 Cost saving
Notes. ICER computed as the change in total costs divided by the change in QALYs relative to Status Quo. Costs include direct costs (HCV treatment costs and other medical expenditures). Changes are discounted at a 3% discount rate. Cost values are reported in 2015$. “Cost saving” indicates a reduction in costs and an increase in QALYs relative to Status Quo. Status Quo, treat F3–F4; Partial Expansion, treat F2–F4; Treat All, treat F0–F4. ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-year; pp, per person; PWID, people who inject drugs.
VALUE IN HEALTH ] (]]]]) ]]]–]]]
quality of life and reduced medical expenditures and mortality. This finding is consistent with existing literature, which also states that treating HCV-infected individuals at earlier fibrosis stages, or treating all patients with HCV infection, is more costeffective compared with the Status Quo [42,43]. The literature also supports our finding that the costs for treating HCV at earlier fibrosis stages are offset by averted costs of care in later stages of disease [42]. Our model predicts that more inclusive scenarios lead to faster disease eradication and the rate of decline in prevalence and incidence is faster when treatment access includes PWID. Treating PWID far exceeds the effects of expanding access to lower fibrosis stages on the prevalence and incidence of HCV infection and the cases of decompensated liver disease and death. Although substantial barriers to treatment exist for PWID, evidence shows that HCV treatment is safe and effective for PWID [44]. Findings suggest that PWID exhibit comparable HCV treatment outcomes compared with groups with other means of transmission [45,46]. In addition, PWID who successfully achieve sustained virologic response are more likely to lead a healthier lifestyle compared with before treatment [47]. Including PWID in access to HCV treatment can also have significant public health benefits. Friedman et al. [48] found that higher metropolitan HIV prevalence among PWID and MSM is positively related to AIDS incidence and AIDS-related mortality among heterosexual non-PWID. Other studies have found that HIV may be transmitted from PWID and MSM to heterosexuals through people who use illicit drugs by routes other than injection [48–51]. This supports our inclusion of the Overlap group and intercohort transmission functions. Friedman et al’s findings suggest that a combination of harm reduction programs and HIV testing for PWID is linked to fewer new AIDS cases in heterosexuals [48]. These findings have important implications for the current study. Higher prevalence of HCV infection among PWID means higher prevalence, costs, and disease burden among the community as a whole. Furthermore, guidelines from international organizations, including the World Health Organization, the American Association for the Study of Liver Diseases, the Infection Diseases Society of America, and the International Network for Hepatitis in Substance Users, have all recommended including PWID in HCV treatment access [52–54], to both reduce disease in these persons and prevent onward transmission to others. Excluding PWID from HCV treatment is shortsighted and a major impediment to eradicating HCV in the U.S. population [55]. Because the majority of incident HCV cases occur among PWID, focusing HCV prevention and eradication efforts on this high-risk population is necessary and strongly supported by the literature [56–58]. As is evidenced by our findings, as well as the existing literature [59], the benefits accrued from long-term avoidance of complications and increased medical costs far outweigh the cost of direct-acting antiviral agents. We considered the impacts of expanding HCV treatment to PWID when other strategies to reduce the transmission of HCV among PWID are also implemented. Specifically, we considered alternatives that extend access to treatment to PWID and also include an NSP to reduce the spread of HCV. We find that simply adding an NSP to the Status Quo reduced infections and saved lives but did not yield positive net social benefits; expanding treatment to earlier stages had more of an impact on cost savings and net benefits to society. In this article, we have shown that expanding HCV treatment to all stages of the disease and PWID may lead to significant social benefits. However, in our model, treatment and medical cost savings do not break-even relative to the Status Quo for 13 years. This suggests a misalignment between social benefits of an inclusive policy and the short-term costs faced by payers. To
7
realize the benefits of inclusive HCV treatment, public policies may be required to realign these incentives. As in any modeling-based analysis, our approach has some limitations and our findings should be viewed in the context of these limitations. Markov models are designed to capture cohortlevel effects and therefore cannot forecast individual disease processes and outcomes. We rely on parameter estimates from the literature, from different studies and populations. Although we made an effort to take parameters from sources with similar underlying populations, variability across the underlying populations is inevitable. We mitigate this by reporting changes relative to the Status Quo and conducting sensitivity analyses around key model parameters. We rely on NHANES data for estimates of the starting population. Although NHANES provides reliable populationlevel estimates, subpopulation estimates are less reliable due to small sample sizes. In addition, because of NHANES’ selfreporting design, it is possible that stigmatized behaviors, such as sexual activities and injection drug use, are underreported, which would affect subpopulation estimates. However, NHANESbased estimates are similar to other estimates reported in the literature. Finally, NHANES also does not capture homeless and incarcerated populations, both of which have high HCV prevalence and limited treatment access. We also rely on parameter estimates from the literature for the United States as a whole and recognize that there may be important regional differences within the United States. In future research, this model may be adapted to focus on regions within the United States that are particularly affected by injection drug use and HCV. A key limitation of our model is that it does not account for growth in opioid use over time. Although we assume a constant population within the groups for model tractability, increased opioid use, and thus increased infectivity due to needle-sharing, has created a spike in the population size for the PWID and OV groups, which our model does not take into account. However, our results are likely a lower estimate on the costs of treating PWID and on the benefits of reducing infectivity among this population. Although we evaluate access policies within a U.S. framework, our results may inform policy outside the United States, even in countries where cost-effectiveness thresholds are lower (e.g., £20,000-30,000/QALY in the United Kingdom). In particular, our analysis indicates that a 5-year horizon is sufficient for the treatment of all fibrosis stages including PWID to be costeffective at a $50,000/QALY threshold. Moreover, this strategy is cost-effective at the $15,000/QALY threshold after 15 years and cost saving after 20 years. In conclusion, increased access to HCV treatment reduces the prevalence and incidence of HCV infection, as well as the associated mortality and downstream medical costs. There is significant gain in societal value by expanding treatment to early stages of liver fibrosis among HCV-infected patients. Furthermore, society also gains significantly from expanding treatment to PWID through decreased infections, which reduce the cost of treatment and other medical expenditures, improve quality of life, and reduce mortality. The value of expanded treatment access is being recognized in the United States, with major private payers, Medicaid, and Medicare beginning to provide access to patients in earlier stages of liver fibrosis. However, many payers continue to impose access restrictions to PWID who are actively injecting drugs. As our findings suggest, the potential benefits to society of treating this high-risk population far exceed the costs of providing treatment. Source of financial support: The design, analysis, and financial support of this study were provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the study.
8
VALUE IN HEALTH ] (]]]]) ]]]–]]]
Supplemental Materials Supplemental material accompanying this article can be found in the online version as a hyperlink at http://dx.doi.org/10.1016/j. jval.2017.01.015 or, if a hard copy of article, at www.valueinhealth journal.com/issues (select volume, issue, and article).
R EF E R EN CE S
[1] Armstrong GL, Wasley A, Simard EP, et al. The prevalence of hepatitis C virus infection in the United States, 1999 through 2002. Ann Intern Med 2006;144:705–14. [2] Biggins SW, Bambha KM, Terrault NA, et al. Projected future increase in aging hepatitis C virus-infected liver transplant candidates: a potential effect of hepatocellular carcinoma. Liver Transplant 2012;18:1471–8. [3] Centers for Disease Control and Prevention. Hepatitis C information for health professionals. 2015. http://www.cdc.gov/hepatitis/HCV/. [Accessed January 20, 2015]. [4] National Health and Nutrition Examination Survey Data [Waves 20032004 through 2011-2012]. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2010. Available from: https://www.cdc.gov/nchs/nhanes/. [Accessed November 1, 2015]. [5] Edlin BR, Eckhardt BJ, Shu MA, et al. Toward a more accurate estimate of the prevalence of hepatitis C in the United States. Hepatology 2015;62:1353–63. [6] Yehia BR, Schranz AJ, Umscheid CA, Lo Re V III. The treatment cascade for chronic hepatitis C virus infection in the United States: a systematic review and meta-analysis. PloS One 2014;9:e101554. [7] Denniston MM, Klevens RM, McQuillan GM, Jiles RB. Awareness of infection, knowledge of hepatitis C, and medical follow‐up among individuals testing positive for hepatitis C: National Health and Nutrition Examination Survey 2001–2008. Hepatology 2012;55:1652–61. [8] Smith BD, Morgan RL, Beckett GA, et al. Recommendations for the identification of chronic hepatitis C virus infection among persons born during 1945-1965. MMWR Recomm Rep 2012;61:1–32. [9] Younossi ZM, Singer ME, Mir HM, et al. Impact of interferon free regimens on clinical and cost outcomes for chronic hepatitis C genotype 1 patients. J Hepatol 2014;60:530–7. [10] Williams IT, Bell BP, Kuhnert W, Alter MJ. Incidence and transmission patterns of acute hepatitis C in the United States, 1982-2006. Arch Intern Med 2011;171:242–8. [11] Wasley A, Gallagher KM, Grytdal S. Surveillance for Acute Viral Hepatitis, United States, 2006. Department of Health and Human Services, Centers for Disease Control and Prevention. March 21, 2008; 57, No. SS-2. [12] Suryaprasad AG, White JZ, Xu F, et al. Emerging epidemic of hepatitis C virus infections among young nonurban persons who inject drugs in the United States, 2006-2012. Clin Infect Dis 2014;59:1411–9. [13] Nelson PK, Mathers BM, Cowie B, et al. Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews. Lancet (London, England) 2011;378:571–83. [14] Crofts N, Aitken CK, Kaldor JM. The force of numbers: why hepatitis C is spreading among Australian injecting drug users while HIV is not. Med J Australia 1999;170:220–1. [15] Bradshaw D, Matthews G, Danta M. Sexually transmitted hepatitis C infection: the new epidemic in MSM? Curr Opinion Infect Dis 2013;26:66–72. [16] Yaphe S, Bozinoff N, Kyle R, et al. Incidence of acute hepatitis C virus infection among men who have sex with men with and without HIV infection: a systematic review. Sexually Transmitted Infect 2012;88:558–64. [17] Centers for Disease Control and Prevention. HIV and viral hepatitis. 2014. Available from: http://www.cdc.gov/hiv/pdf/library_factsheets_ hiv_and_viral_hepatitis.pdf. [September 1, 2016]. [18] Taylor LE, Holubar M, Wu K, et al. Incident hepatitis C virus infection among US HIV-infected men enrolled in clinical trials. Clin Infect Dis 2011;52:812–8. [19] Canary LA, Klevens RM, Holmberg SD. Limited access to new hepatitis C virus treatment under state Medicaid programs. Ann Intern Med 2015;163:226–8. [20] Anthem Blue Cross and Blue Shield. Medication quantity limit Sovaldi. 2015. https://www.anthem.com/provider/noapplication/f0/s0/t0/ pw_e210963.pdf. [Accessed October 7, 2015]. [21] Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993;13:322–38.
[22] Van Nuys K, Brookmeyer R, Chou JW, et al. Broad hepatitis C treatment scenarios return substantial health gains, but capacity is a concern. Health Aff 2015;34:1666–74. [23] Linthicum MT, Sanchez Gonzales Y, Mulligan K, et al. Value of expanding HCV screening and treatment policies in the United States. Am J Managed Care 2016;22:SP227–35. [24] Witt MD, Seaberg EC, Darilay A, et al. Incident hepatitis C virus infection in men who have sex with men: a prospective cohort analysis, 1984–2011. Clin Infect Dis 2013;57:77–84. [25] Smith BD, Morgan RL, Beckett GA, et al. Recommendations for the identification of chronic hepatitis C virus infection among persons born during 1945-1965. MMWR Recomm Rep 2012;61:1–32. [26] Manos MM, Shvachko VA, Murphy RC, et al. Distribution of hepatitis C virus genotypes in a diverse US integrated health care population. J Med Virol 2012;84:1744–50. [27] UnitedHealthcare. UnitedHealthcare Pharmacy, Clinical Pharmacy Programs, Prior Authorization/Medical Necessity Harvoni™ (ledipasvir/ sofosbuvir). 2014. https://www.unitedhealthcareonline.com/ ccmcontent/ProviderII/UHC/en-US/Assets/ProviderStaticFiles/ ProviderStaticFilesPdf/ToolsandResources/PharmacyResources/ Sovaldi_Medical_Necessity.pdf. [Accessed October 7, 2015]. [28] Aetna. Pharmacy clinical policy bulletins, Aetna non-Medicare prescription drug plan: hepatitis C. 2015. http://www.aetna.com/ products/rxnonmedicare/data/2014/GI/hepatitis_c.html. [Accessed October 7, 2015]. [29] Anthem. Viekira Pak (ombitasvir/paritaprevir/ritonavir þ dasabuvir). 2015. https://www.anthem.com/provider/noapplication/f0/s0/t0/ pw_e210963.pdf?na=pharminfo. [Accessed October 7, 2015]. [30] Health Net. Prior authorization protocol: SOLVALDI™ (sofosbuvir). 2015. https://www.healthnet.com/static/general/unprotected/html/national/ pa_guidelines/sovaldi_natl.html. [Accessed October 7, 2015]. [31] Canary LA, Klevens RM, Holmberg SD. Limited access to new hepatitis C virus treatment under state Medicaid programs. Ann Intern Med 2015;163:226–8. [32] Barua S, Greenwald R, Grebely J, et al. Restrictions for Medicaid reimbursement of sofosbuvir for the treatment of hepatitis C virus infection in the United States. Ann Intern Med 2015;163:215–23. [33] Ghany MG, Strader DB, Thomas DL, Seeff LB. Diagnosis, management, and treatment of hepatitis C: an update. Hepatology 2009;49:1335–74. [34] Mathers BM, Degenhardt L, Ali H, et al. HIV prevention, treatment, and care services for people who inject drugs: a systematic review of global, regional, and national coverage. Lancet (London, England) 2010;375:1014–28. [35] Wodak A, Cooney A. Do needle syringe programs reduce HIV infection among injecting drug users: a comprehensive review of the international evidence. Substance Use Misuse 2006;41:777–813. [36] Edlin BR. Perspective: test and treat this silent killer. Nature 2011;474:S18–9. [37] Health Canada. Vancouver’s INSITE service and other supervised injection sites: what has been learned from research? Health Canada, March 31, 2008. Available from: http://www.hc-sc.gc.ca/ahc-asc/pubs/ _sites-lieux/insite/index-eng.php. [Accessed September 1, 2016]. [38] Honeycutt AA, Harris JL, Khavjou O, et al. The costs and impacts of testing for hepatitis C virus antibody in public STD clinics. Public Health Rep 2007;122(Suppl 2):55–62. [39] Green TC, Hankins CA, Palmer D, et al. My place, your place, or a safer place: the intention among Montreal injecting drug users to use supervised injecting facilities. Can J Public Health 2004;95(2):110–4. [40] Wood E, Tyndall MW, Montaner JS, Kerr T. Summary of findings from the evaluation of a pilot medically supervised safer injecting facility. Can Med Assoc J 2006;175:1399–404. [41] Hirth RA, Chernew ME, Miller E, et al. Willingness to pay for a qualityadjusted life year in search of a standard. Med Decis Making 2000;20:332–42. [42] Chahal HS, Marseille EA, Tice JA, et al. Cost-effectiveness of early treatment of hepatitis C virus genotype 1 by stage of liver fibrosis in a US treatment-naive population. JAMA Intern Med 2016;176:65–73. [43] Visconti AJ, Doyle JS, Weir A, et al. Assessing the cost‐effectiveness of treating chronic hepatitis C virus in people who inject drugs in Australia. J Gastroenterol Hepatol 2013;28:707–16. [44] Edlin BR, Kresina TF, Raymond DB, et al. Overcoming barriers to prevention, care, and treatment of hepatitis C in illicit drug users. Clin Infect Dis 2005;40:S276–85. [45] Aspinall E, Corson S, Doyle J, et al. Peginterferon and ribavirin treatment for chronic hepatitis C virus in people who inject drugs: a systematic review and meta-analysis. Clin Infect Dis 2013;57:80–9. [46] Dimova RB, Zeremski M, Jacobson IM, et al. Determinants of hepatitis C virus treatment completion and efficacy in drug users assessed by meta-analysis. Clin Infect Dis 2013;56:806–16. [47] Innes HA, McDonald SA, Dillon JF, et al. Toward a more complete understanding of the association between a hepatitis C sustained viral response and cause‐specific outcomes. Hepatology 2015;62:355–64. [48] Friedman SR, West BS, Tempalski B, et al. Do metropolitan HIV epidemic histories and programs for people who inject drugs and men
VALUE IN HEALTH ] (]]]]) ]]]–]]]
[49]
[50]
[51] [52]
[53]
who have sex with men predict AIDS incidence and mortality among heterosexuals? Ann Epidemiol 2014;24:304–11. Neaigus A, Miller M, Gyarmathy VA, Friedman SR. HIV heterosexual sexual risk from injecting drug users among HIV-seronegative noninjecting heroin users. Substance Use Misuse 2011;46:208–17. Strathdee SA, Stockman JK. Epidemiology of HIV among injecting and non-injecting drug users: current trends and implications for interventions. Curr HIV/AIDS Rep 2010;7:99–106. Volz E, Frost SD, Rothenberg R, Meyers LA. Epidemiological bridging by injection drug use drives an early HIV epidemic. Epidemics 2010;2:155–64. AASLD/IDSA. HCV guidance: recommendations for testing, managing, and treating HCV. 2016. http://www.hcvguidelines.org/. [Accessed June 15, 2016]. World Health Organization. Guidelines for the Screening Care and Treatment of Persons With Hepatitis C Infection. Geneva, Switzerland: World Health Organization, 2014.
9
[54] Robaeys G, Grebely J, Mauss S, et al. Recommendations for the management of hepatitis C virus infection among people who inject drugs. Clin Infect Dis 2013;57:S129–37. [55] Leask J, Dillon J. Review article: treatment as prevention–targeting people who inject drugs as a pathway towards hepatitis C eradication. Aliment Pharmacol Therapeut 2016;44(2):145–56. [56] Hagan H, Pouget ER, Des Jarlais DC. A systematic review and metaanalysis of interventions to prevent hepatitis C virus infection in people who inject drugs. J Infect Dis 2011;204:74–83. [57] Hellard M, Doyle JS, Sacks‐Davis R, et al. Eradication of hepatitis C infection: the importance of targeting people who inject drugs. Hepatology 2014;59:366–9. [58] Hellard M, McBryde E, Davis RS, et al. Hepatitis C transmission and treatment as prevention–the role of the injecting network. Int J Drug Policy 2015;26:958–62. [59] Hagan LM, Wolpe PR, Schinazi RF. Treatment as prevention and cure towards global eradication of hepatitis C virus. Trends Microbiol 2013;21:625–33.