Cost‐effectiveness of integrated treatment for hepatitis C virus (HCV) among people who inject drugs in Norway: An economic evaluation of the INTRO‐HCV trial

Abstract Background and aims The INTRO‐HCV randomized controlled trial conducted in Norway over 2017–2019 found that integrated treatment, compared with standard‐of‐care hospital treatment, for hepatitis C virus (HCV) with direct‐acting antivirals (DAAs) improved treatment outcomes among people who inject drugs (PWID). We evaluated cost‐effectiveness of the INTRO‐HCV intervention. Design A Markov health state transition model of HCV disease progression and treatment with cost‐effectiveness analysis from the health‐provider perspective. Primary cost, utility, and health outcome data were derived from the trial. Costs and health benefits (quality‐adjusted life‐years, QALYs) were tracked over 50 years. Probabilistic and univariate sensitivity analyses investigated DAA price reductions and variations in HCV treatment and disease care cost assumptions, using costs from different countries (Norway, United Kingdom, United States, France, Australia). Setting and participants PWID attending community‐based drug treatment centers for people with opioid dependence in Norway. Measurements Incremental cost‐effectiveness ratio (ICER) in terms of cost per QALY gained, compared against a conventional (€70 000/QALY) willingness‐to‐pay threshold for Norway and lower (€20 000/QALY) threshold common among high‐income countries. Findings Integrated treatment resulted in an ICER of €13 300/QALY gained, with 99% and 71% probability of being cost‐effective against conventional and lower willingness‐to‐pay thresholds, respectively. A 30% lower DAA price reduced the ICER to €6 900/QALY gained, with 91% probability of being cost‐effective at the lower willingness‐to‐pay threshold. A 60% and 90% lower DAA price had 36% and >99% probability of being cost‐saving, respectively. Sensitivity analyses suggest integrated treatment was cost‐effective at the lower willingness‐to‐pay threshold (>60% probability) across different assumptions on HCV treatment and disease care costs with 30% DAA price reduction, and became cost‐saving with 60%–90% price reductions. Conclusions Integrated hepatitis C virus treatment for people who inject drugs in community settings is likely cost‐effective compared with standard‐of‐care referral pathways in Norway and may be cost‐saving in settings with particular characteristics.


INTRODUCTION
Injecting drug use (IDU) is a key driver of hepatitis C virus (HCV) transmission [1,2].Despite the development of highly efficacious directacting antiviral (DAA) treatments for HCV, people who inject drugs (PWID) frequently have low levels of diagnosis and treatment because of stigma and lack of access to services [3].Implementing HCV treatment through efficient and integrated delivery platforms can help overcome these barriers by increasing accessibility to DAAs among PWID, which is urgently required to progress toward achieving World Health Organization (WHO) targets for eliminating HCV globally [4].However, despite considerable potential for cost-savings, until recently evidence supporting integrated care for HCV among PWID has been limited [5].
The integrated treatment of hepatitis C (INTRO-HCV) study [6] was a multi-center, randomized controlled trial that enrolled 298 HCV-infected PWID attending one of eight multidisciplinary opioid agonist therapy (OAT) clinics or two community care centers (CCC) in Norway over 2017 to 2019.The purpose of the trial was to evaluate the impact on HCV treatment outcomes of integrated HCV treatment within decentralized OAT clinics and CCC compared to standard-of-care treatment occurring in hospital outpatient clinics.
Overall, the INTRO-HCV trial demonstrated that the integrated pathway led to superior treatment outcomes in terms of a higher proportion of patients initiating treatment (integrated, 98% vs standardof-care, 77%), a faster time to treatment initiation (hazard ratio of 2.2), and a higher proportion achieving sustained virological response (SVR) (effective cure) following treatment (integrated, 95%; standardof-care, 88% among those completed treated and polymerase chain reaction [PCR]-tested post-treatment) [7].
In this study, we use modelling combined with the INTRO-HCV trial findings [7] to estimate the long-term costs, health outcomes and cost-effectiveness of the integrated HCV treatment delivery approach versus the standard-of-care treatment referral to hospital outpatient clinics for PWID in community settings.

Study description
The INTRO-HCV trial recruited PWID from OAT clinics or CCC in Bergen and Stavanger (Western Norway) who were diagnosed with detectable HCV RNA and eligible for HCV treatment.Randomization occurred after recruitment.DAA regimens were used in both treatment pathways, with the primary trial outcomes being treatment uptake and SVR assessed at 12 weeks after treatment completion.

Model description
We developed a closed cohort Markov model of HCV disease progression and treatment among HCV-diagnosed individuals (Figure 1), based on the INTRO-HCV cohort.The model was stratified by HCVassociated disease progression (Figure 1a), with two mild fibrosis states (meta-analysis of histological data in viral hepatitis [METAVIR] stages F0-F1), two moderate fibrosis states (METAVIR stages F2-F3), compensated cirrhosis (CC) (METAVIR stage F4), decompensated cirrhosis (DC), hepatocellular carcinoma (HCC), liver transplantation (LT) and post-liver transplantation (PLT) states.Advanced liver disease states (DC, HCC, LT and PLT) were associated with HCV-related mortality.The model was further stratified by HCV infection status, treatment uptake and SVR outcome (Figure 1b).HCV treatment was offered to patients in health states F0 to F4 as in the INTRO-HCV trial, which resulted in either SVR with a possibility of reinfection, or therapeutic failure, whereby patients remain infected and may be retreated.Following treatment success, liver disease progression was assumed to halt during mild and moderate fibrosis (F0-F3) stages, and to reduce for more severe hepatic complications (F4/CC, DC, HCC).
Age-dependent mortality was also included.

Model parameterization
Parameters relating to HCV treatment were obtained from the INTRO-HCV trial [7] (Table 1).In the first year, the proportion of diagnosed individuals that were treated assumed the first-year trial outcomes (integrated, 93.9%; standard-of-care, 72.0%).Following the first year, the treatment rates for the model were derived to give the total treated proportion in each arm of the trial (integrated, 98.0%; standard-of-care, 77.3%), yielding fitted second-year treatment rates of 66.7% for the integrated pathway and 19.1% for the standardof-care pathway (Figure S1), with no treatment occurring thereafter.Individuals, who have failed treatment, have been reinfected, or did not initiate HCV treatment would remain in the infected compartment and would be eligible for second-year treatment.SVR rates (integrated, 94.6%; standard-of-care, 87.5%) also came from the trial.
Uncertainty was associated with all trial-derived parameters (Table 1).
A low level of reinfection was assumed following successful treatment based on a recent study among PWID in Norway (3.7% [95% CI = 1.6%-7.2%] of SVR reinfected annually) [12].Further details are in the Supporting Information.
Health state transition probabilities came from the literature [13][14][15] (Table S1) and adjusted for the INTRO-HCV cohort's higher proportion of HCV genotype 3 (60%), which is associated with faster disease progression [16] (Supporting Information).Individuals with chronic HCV infection were assumed to progress through the disease stages [17].
Background death probabilities were derived from Norwegian life tables for 2017 [8], assuming the trial cohort's average age at the start of the modelled cohort (43 years) and adjusting the mortality rate by the standardized mortality ratio for PWID in Western Europe [9] (17.5; 95% CI = 14.6-20.4)if currently engaged in IDU, assuming an average injection duration of 14 years [10,11] (+/−10 years).Background death probabilities were adjusted for sex to reflect the malefemale ratio in the INTRO-HCV trial and changed as the cohort aged.

Estimation of costs
All costs associated with HCV treatment were collected directly from the INTRO-HCV trial using a healthcare perspective (Table 2, Table S2).T A B L E 2 Health utilities and costs.See further details in the Supporting Information.

Derivation of health utilities
Health-related quality of life health (HRQoL) measures were also collected from the INTRO-HCV trial [7] using the EuroQoL EQ-5D-5L questionnaire [22] among patients with chronic HCV infection before treatment initiation and at 1 year following treatment success.These were stratified by fibrosis stage and converted into health utilities using a United Kingdom (UK) value set in the absence of a Norwegian value set (Table 2).Further details in Supporting Information.Remaining health utilities for post-cirrhotic disease and current IDU status were derived from existing literature values (Table 2).Health utilities were sampled from their uncertainty distributions.

COST-EFFECTIVE INTEGRATED HCV TREATMENT
The impact of treatment for each pathway was estimated in terms of total quality-adjusted life-years (QALYs) gained by applying the health utilities to their respective HCV disease progression states, with possible differences in health utilities depending on current injecting status (i.e.currently injecting or ceased injecting) and accumulating them annually over the time horizon, as described in previous studies [23][24][25].Secondary outcomes were the differences in treatment uptake and proportion of incident liver disease complications prevented (F4/CC, DC and HCC and liver-related deaths) between the integrated versus standard-of-care treatment pathways.
Further details in Supporting Information.
The incremental cost-effectiveness ratio (ICER) was estimated as the ratio of the difference in total costs between the integrated and standard-of-care treatment pathways divided by the corresponding difference in QALYs [26], which represented the incremental costs of integrated treatment associated per QALY gained from the health provider's perspective over a lifetime time horizon.All costs and QALYs were discounted at 4% per annum [21,27].The analysis did not follow a preregistered analysis plan, but used similar methods to our previous studies [24,28,29].This study has been reported as per the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) guidelines (Checklist S1).

Sensitivity analyses
To account for uncertainty in model parameters, we conducted a probabilistic sensitivity analysis by simultaneously sampling all model parameters 1000 times from their respective statistical distributions (Table 1) and simulating the model for each sampled parameter set.

Because DAA medication costs as provided by the Norwegian Health
Services are confidential, we evaluated the ICER for different reductions in the DAA list price (30%/60%/90%).To determine the probability of the integrated pathway being cost-effective compared to the standard-of-care pathway, the estimated ICERs were assessed against a conventional (€70 000/QALY) willingness-to-pay (WTP) threshold suggested by the Norwegian Directorate of Health [30] and a lower (€20 000/QALY) threshold commonly considered by other highincome countries [31].The probability of integrated treatment being cost-saving was also assessed.
We performed further univariate sensitivity analyses to investigate how the ICER changed by varying key parameters in the model.(none or double the baseline rate) and currency conversion (lower or higher rates or using market exchange rates).Further details are in Table S3.
We then performed additional sensitivity analyses to investigate the cost-effectiveness of the integrated treatment pathway under different assumptions on the costs of service delivery and care, using cost data from other country settings.Specifically, we considered settings with (scenario S1) similar treatment costs, but lower healthcare management costs for late-stage HCV-related disease (using costs from United Kingdom); (scenario S2) higher treatment costs (using costs from United States [US]); (scenario S3) lower treatment costs, but higher healthcare costs (using costs from France); and (scenario S4) lower costs of both treatment and healthcare management (using costs from Australia).For these sensitivity analyses, we assumed the same trial effects for the integrated and standard-of-care treatment pathways, but used the corresponding cost assumptions from published cost-effectiveness studies for other settings (Table 3, Table S4).
Actual negotiated DAA prices are confidential, but are likely to be lower than the list price so, for these analyses, we considered DAA medications at list price and also 30%/60%/90% price reductions.
Meanwhile, annual healthcare costs for managing HCV-related disease were adjusted by producer price indices (PPI) to the year 2021.
The median ICER and probability of being cost-effective or costsaving for integrated treatment were evaluated for each country example, with all costs converted to 2021 Euros (€) using PPP for cross-country comparison.
Model results have been presented as the median and 95% uncertainty interval (UI) of 1000 model simulations.All model analyses were performed in MATLAB (version 2021a).
a Non-DAA-related HCV treatment costs and annual healthcare costs for chronic HCV disease stages in each country were obtained from published country-level studies and adjusted for inflation using PPI to 2021 from cost year in which the original study occurred, if known.Further details are in Table S4.
b Assumed to be combined with one-time LT costs because annual costs were not available from published studies.2).

Cost of HCV treatment pathways
The standard-of-care pathway was estimated to cost €31 093 (95% UI = 24 314-38 756) per person diagnosed and eligible for HCV treatment over 50-years, with two thirds of the costs due to HCV treatment (68.4%) and the remainder due to healthcare management (31.6%)(Table S5).recently being adopted into routine healthcare practice among PWID in community settings in Norway [39].Our results are potentially generalizable to other settings with similar OAT provision infrastructures for PWID and suggest ways that an integrated approach to delivering HCV treatment can be implemented in these settings.
The integrated treatment pathway involved HCV treatment being delivered at OAT clinics or CCC by on-site multidisciplinary teams, including clinical specialists in addiction medicine, psychologists providing mental health services, nurses, social workers and peer counsellors.Patients had frequent follow-up appointments (usually several times per week) for substance use disorders or other medical conditions, with follow-up related to HCV treatment being integrated into these appointments.The standard-of-care treatment pathway involved patients being referred for clinical assessment, HCV treatment and 4 to 5 follow-up consultations at a standard medical outpatient hospital clinic within 1 to 25 km travelling distance, with transportation costs being incurred by the patient.Appointments unrelated to HCV were scheduled separately.
Abbreviations: DAA, direct-acting antivirals; DC, decompensated cirrhosis; €, Euros; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HR, human resources; INTRO-HCV, integrated treatment of hepatitis C virus; IQR, interquartile range; LT, liver transplant; NA, not applicable; PLT, post-liver transplant; PPP, purchasing power parities; PWID, people who inject drugs; SVR, sustained virological response; UK, United Kingdom.aTotal treatment unit cost per patient and all annual healthcare management costs have a +/−20% uncertainty associated with them and are each sampled from their respective uniform distributions.
These included assumptions on total treatment unit cost (half or double the baseline cost), reinfection incidence (2/100 person-years or 12.1/100 person-years), second-year treatment rate (integrated pathway same as standard-of-care), treatment duration (none or continued treatment for 10 years), average age of cohort (20 years younger or 10 years older), initial distribution of disease states (less or more severe disease states compared to baseline), injecting duration (5 or 30 years), healthcare management costs (none or double healthcare costs), whether SVR differed by pathway (no difference compared to higher SVR rate for integrated pathway at baseline), health utilities (UK-based literature values compared to trial values) [14, 15], time horizon (25 years compared to 50 years at baseline), discounting Analysis of theHRQoL data from the INTRO-HCV trial revealed lower median utilities for increasing fibrosis scores at treatment initiation, T A B L E 3 Mean costs used in sensitivity analyses estimating the cost-effectiveness of integrated versus standard-of-care treatment in settings with different conditions, using selected countries as case examples.

F
I G U R E 4 Univariate sensitivity analyses on the incremental cost-effectiveness ratio (ICER) of the integrated treatment pathway compared to the standard-of-care treatment pathway, assuming list costs for direct-acting antivirals (DAA) treatments (€47 800 per 12-week treatment).Costs and quality-adjusted life-years (QALYs) are discounted at a rate of 4.0% per annum.Time horizon is 50 years.The conventional (dashed line) and lower (dotted line) willingness-to-pay (WTP) thresholds are shown.Results are for 1000 model simulations.
Base case parameters and annual transition probabilities used in the model.
ICERs for the integrated treatment pathway compared to the standard treatment pathway, at baseline and for specific reductions in DAA price (30%/60%/90%).
T A B L E 4