The Effectiveness and Cost-Effectiveness of Hepatitis C Screening for Migrants in the EU/EEA: A Systematic Review

Chronic hepatitis C (HCV) is a public health priority in the European Union/European Economic Area (EU/EEA) and is a leading cause of chronic liver disease and liver cancer. Migrants account for a disproportionate number of HCV cases in the EU/EEA (mean 14% of cases and >50% of cases in some countries). We conducted two systematic reviews (SR) to estimate the effectiveness and cost-effectiveness of HCV screening for migrants living in the EU/EEA. We found that screening tests for HCV are highly sensitive and specific. Clinical trials report direct acting antiviral (DAA) therapies are well-tolerated in a wide range of populations and cure almost all cases (>95%) and lead to an 85% lower risk of developing hepatocellular carcinoma and an 80% lower risk of all-cause mortality. At 2015 costs, DAA based regimens were only moderately cost-effective and as a result less than 30% of people with HCV had been screened and less 5% of all HCV cases had been treated in the EU/EEA in 2015. Migrants face additional barriers in linkage to care and treatment due to several patient, practitioner, and health system barriers. Although decreasing HCV costs have made treatment more accessible in the EU/EEA, HCV elimination will only be possible in the region if health systems include and treat migrants for HCV.


Introduction
Chronic hepatitis C is an important public health problem in the EU/EEA, with an estimated 3.24 million persons having active hepatitis C virus (HCV) infection [1,2]. It is a leading cause of chronic liver disease and liver cancer in the EU/EEA due to undetected and untreated infections [3][4][5]. Since 2013, the landscape of HCV treatment has changed rapidly as pan-genotypic DAA HCV treatment regimens that cure most infections (>95% of cases) have become available, making HCV elimination possible [6][7][8][9]. In 2015 however, only 34% of HCV infected persons had been diagnosed and less than 5% of all HCV cases had been treated [2]. Identifying and treating all groups at risk for HCV in the EU/EEA will be essential to address the health and economic burden due to HCV in the EU/EEA and to reach WHO elimination goals by 2030 [3,5,[9][10][11].
HCV screening and control programs in the EU/EEA primarily focus on persons who inject drugs (PWID), as they are the largest and highest burden population [1]. Migrants from intermediate and high HCV prevalence countries (anti-HCV ≥ 2% and ≥5%, respectively) are an additional important and underappreciated group at increased HCV risk in the EU/EEA and often do not have identifiable HCV risk factors [12,13]. They are most likely to have been exposed to HCV in their countries of origin through receipt of contaminated blood products or unsafe injections or procedures, and have a prevalence of HCV that reflects that of their countries of origin [14,15]. The increased flow of migrants from intermediate and high HCV prevalence countries into the EU/EEA over the past few decades has resulted in a disproportionately high number of reported HCV cases (14%) occurring among migrants, who account for up to one half of all cases in some low HCV prevalence EU/EEA countries [12,13]. HCV diagnosis among migrants living in low incidence countries is delayed due to several patient, practitioner, and infrastructural barriers that may result in a higher burden of liver-associated complications compared to host populations [16,17]. We conducted a systematic review (SR) to estimate the effectiveness, resource use, costs, and cost-effectiveness of HCV screening programs for migrants in the EU/EEA.

Overall Approach and Key Questions
Using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach, the Campbell and Cochrane Collaboration Equity Methods Group and review team including clinicians, public health experts and researchers from across the EU/EEA, we conducted evidence syntheses. "Migrants", a focus of this review, included asylum seekers, refugees, undocumented migrants, and other foreign-born residents. A detailed description of the methods have been published and were registered in PROSPERO (CRD42016045798) [18].
We used the GRADE approach to rate the certainty of evidence starting with a simplified categorization of study types (i.e., meta-analyses, RCTs, and observational studies). The rating scheme allows for factors that may raise or lower the level of certainty. Factors that lower certainty of evidence include, risk of bias, inconsistency across studies, indirectness, and publication bias. Factors that increase certainty of evidence include large effect size and an observed dose-response effect. The final certainty ratings are reflective of the certainty in the estimated effect in the context of bias and limitations. Evidence was graded as high, moderate, low, or very low certainty, based on how likely further research would change the confidence in the estimate of effect. Low certainty and very low certainty do not mean absence of evidence for effectiveness, but rather signal highlights the need for more research to improve the precision of the estimate of effect.
This review followed the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) and Cochrane methodological approach [18]. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist for reporting the results of the systematic reviews (SR) [18,19]. The review team developed two overarching research questions (PICO: Population, Intervention, Comparison, Outcome) and a logic model ( Figure S1). The logic model showed key questions/concepts along the evidence chain along the screening effectiveness pathway [18,20,21]. The two overarching research questions (PICO) we sought to answer were: • What is the effectiveness of screening migrants arriving and living in the EU/EEA for HCV? • What is the cost, resource utilization, and cost-effectiveness for screening migrants for HCV?
The following key questions were identified along the screening effectiveness pathway. (1) What is the test accuracy and performance characteristics of screening tests for HCV? (2) What is the efficacy of new direct acting antiviral (DAA) treatments for HCV to decrease HCV associated morbidity and mortality? (3) What is the uptake of HCV of screening and treatment? (4) What is the cost-effectiveness of a screen-treat approach for HCV in the general population and the migrant population when treated with DAAs? [18].

Search Strategy and Selection Criteria
We conducted two searches, one for SRs and guidelines on the effectiveness and cost-effectiveness of HCV screening programs in migrants and a second search for SRs and primary studies on the resource use, costs, and cost-effectiveness of HCV screening programs in migrants. For the first search, Medline via OVID, EMBASE, CINAHL, Epistemonikos, and Cochrane CENTRAL were searched for publications between 1 January 2010 and 12 May 2016. A combination of key terms was used including "hepatitis C/HCV", "screening", "migrants", "costs", "cost-effectiveness" AND "guidelines", and "reviews". The search terms and the search strategy for Ovid Medline are included in the supplementary material (Table S1). We also searched grey literature websites for published guidelines and reports at CDC, ECDC, EASL (European Association for the Study of the Liver), and WHO. We applied no language restrictions to the search. In the second search, using terms of "hepatitis C/HCV", "screening", "costs", and "cost-effectiveness", Medline via Ovid, EMBASE, the NHS Economic Evaluation Database (NHS EED), Database of Abstracts of Reviews of Effects (DARE), and the Cost Effectiveness Analysis Tufts registry and Google scholar databases were searched for publications between 1 January 2000 and 31 May 2016. Reference lists of relevant reviews were also searched.

Study Selection and Quality Assessment
Two authors screened the titles and abstracts, assessed selected full-text articles for eligibility, and extracted data from included articles. Disagreements were resolved by consensus or by a third author. For the screening effectiveness search we included systematic reviews on the impact of HCV screening or antiviral therapies on the development of liver related morbidities such as cirrhosis, hepatocellular carcinoma, and the need for liver transplantation and all-cause or attributable mortality. For the cost-effectiveness search we included individual economic studies of screening strategies that included an arm of direct acting antiviral (DAA) therapies or studies of the cost-effectiveness of DAA therapies [8]. We only included studies published in full and in English or French. If more than one version of a SR was identified, the most recent was considered. Studies were excluded if they focused only on nongeneralizable subgroups (such as PWIDs) (Figures 1 and 2).   The methodologic quality of SRs was assessed by two authors using the AMSTAR tool (A Measurement Tool to Assess Systematic Reviews) [22]. The GRADE criteria were applied to assess the certainty of evidence in preselected outcome measures in the SRs (GRADE Tables S2-S5) [23]. For the second search which included individual studies the certainty of economic evidence in each study was assessed using the relevant items from the 1997 Drummond checklist [24].

Data Extraction and Synthesis
The following information was extracted from each study: study design, objectives, analyses, quality of the individual studies included in the systematic review, population examined, number of included studies, total number of participants included, intervention, outcome, and the results. For the economic studies we extracted the following data; economic methods used (e.g., microcosting study, within-trial cost-utility analysis, Markov model), description of the case base population, the intervention and comparator, absolute size and relative difference in resource use, cost-effectiveness results (e.g., incremental net benefit (INB) or incremental cost-effectiveness ratio (ICER)), and three specific questions for the GRADE Evidence to Decision table: the size of the resource requirements, the certainty of evidence around resource requirements, and whether the cost-effectiveness results The methodologic quality of SRs was assessed by two authors using the AMSTAR tool (A Measurement Tool to Assess Systematic Reviews) [22]. The GRADE criteria were applied to assess the certainty of evidence in preselected outcome measures in the SRs (GRADE Tables S2-S5) [23]. For the second search which included individual studies the certainty of economic evidence in each study was assessed using the relevant items from the 1997 Drummond checklist [24].

Data Extraction and Synthesis
The following information was extracted from each study: study design, objectives, analyses, quality of the individual studies included in the systematic review, population examined, number of included studies, total number of participants included, intervention, outcome, and the results. For the economic studies we extracted the following data; economic methods used (e.g., microcosting study, within-trial cost-utility analysis, Markov model), description of the case base population, the intervention and comparator, absolute size and relative difference in resource use, cost-effectiveness results (e.g., incremental net benefit (INB) or incremental cost-effectiveness ratio (ICER)), and three specific questions for the GRADE Evidence to Decision table: the size of the resource requirements, the certainty of evidence around resource requirements, and whether the cost-effectiveness results favored the intervention or comparison [25]. Key results were converted to 2015 Euros using the Cochrane web-based currency conversion tool: https://eppi.ioe.ac.uk/costconversion/default.aspx.

Search Results
In the search for the effectiveness of HCV screening we retrieved 1475 references and identified 10 additional records through other sources ( Figure 1). After duplicates were removed, 1459 references were screened by title and abstract. A total of 43 references were then selected for full text assessment. We did not identify any randomized controlled trials or SRs on the effectiveness of HCV screening in the general or migrant populations. We therefore included five SRs and one guideline that addressed the key questions along the screening evidence chain: the performance of HCV diagnostic tests (n = 2) [6,26], the impact of HCV treatment on preventing HCC and all-cause mortality (n = 3) [27][28][29], and the HCV care continuum (n = 1) ( Table 1) [30]. In the economic search, 682 articles were retrieved and an additional eight records identified through other sources ( Figure 2). After duplicate removal, 664 references were screened by title and abstracts. Of these, a total of 88 references underwent full text assessment and 14 individual studies of populations living in low HCV prevalence countries were included (

Performance of Diagnostic Tests
The performance of diagnostic testing for HCV has been recently summarized in the 2017 WHO Guidelines on Hepatitis B and C testing [6]. WHO estimates the sensitivity and specificity of 3rd generation HCV enzyme immunoassays (EIA) to be 98% and 99%, respectively [6]. Similarly, the sensitivity of the confirmatory test to detect virus nucleic acid (nucleic acid test; NATs) is estimated to be 96.2% (95% CI, 94.4-97.5) and the specificity to be 98.9% (98.3-99.3) [6]. There was no reported evidence suggesting that migrants or any other group would encounter lower performance rates. The accuracy of point of care testing, a strategy that potentially could increase screening uptake, was reviewed by Khuroo et al. who found that these tests performed well in low, middle, and high income countries [26]. The sensitivity (95.8% (93.9-97.1)) was slightly lower compared to EIAs but demonstrated comparable specificity (99.0% (98.5-99.3)) [26]. The GRADE certainty of the evidence in the Khuroo study was very low (S3 GRADE Table 1).

Impact of Therapy on Long-Term Outcomes
Interferon free DAAs are the recommended therapy for all HCV genotypes in the EU/EEA [8]. These regimens are well tolerated and cure > 95% of cases based on achieving sustained virologic response (SVR0, which is considered to be a reliable surrogate outcome for HCV cure) [7,8]. Within the search dates we did not identify any studies that examined the impact of DAA therapies on long-term HCV liver related outcomes or mortality, as these agents only became available in 2013. Three SRs that assessed the impact of older interferon (IFN) based HCV treatment on preventing liver related sequelae and all-cause mortality were included [27][28][29]. These three SRs included studies from Europe, North America, Australia, and Asia, and found that IFN-based HCV therapy significantly decreased rates of hepatocellular carcinoma (HCC), hepatic decompensation, and all-cause mortality in those on HCV treatment, and particularly in those who achieved SVR [27][28][29]. In a meta-analysis by Kimer et al., the risk of HCC was lower in those on antiviral HCV therapy (IFN or PEG-IFN alone or with ribavirin) compared to placebo or no intervention (RR = 0.53, 95% CI 0.34-0.81). This effect was much more pronounced among virological responders compared to nonresponders [27]. In the SR by Simmons et al., the adjusted hazard ratio (aHR) of all-cause mortality rate was lower in patients on treatment for chronic HCV after a median follow-up time of 5.4 years [28]. In those achieving SVR, compared with non-SVR, the aHR was 0.50 (95% CI: 0.37-0.67) in the general population, and 0.26 (95% CI: 0.18-0.37) in the cirrhotic group [28]. Finally, in the SR conducted by the Public Health Agency of Canada (PHAC), a reduction in hepatic mortality (60 fewer/1000, 95% CI: 59-62), HCC (18 fewer/1000, 95% CI: [17][18][19], hepatic decompensation or decompensated cirrhosis (46 fewer/1000, 95% CI: 46-47), and need for liver transplantation (4 fewer/1000, 95% CI: 4-5) among those treated with DAA to PEG-IFN was found [29]. The GRADE certainty of the evidence of the data in these three systematic reviews was very low to moderate (GRADE Tables S3-S5).

The HCV Care Continuum and Pathway
In a SR of studies of the HCV care continuum in the pre-DAA period among in the general population in the US from 2003-2013, only 50% of HCV cases were diagnosed and aware of their infection, 27% had HCV RNA confirmatory testing, 16% were prescribed HCV therapy, and 9% achieved SVR [30]. The results of this study were not stratified by risk group nor did they report on the barriers to uptake of the steps along the care cascade. A modeling study in Europe published after the search timeframe also demonstrated a weak HCV care continuum in 2015, consistent with the US SR by Yehia [2]. They found that in Europe in 2015, only 36% of HCV infected persons have been screened and, of those diagnosed, only 12.7% and 11.3% had been treated and cured, respectively [2]. They found that, to achieve WHO elimination targets, expansion of screening programs in the EU/EAA would be needed along with unrestricted access to treatment for all found to be infected.

Resource Use, Costs and Cost-Effectiveness
Screening for HCV in those treated with DAAs is cost-effective even at higher 2015 costs. A UK study evaluated the cost-effectiveness of screening and treating pregnant women attending antenatal clinics [44]. The incremental cost effectiveness ratio (ICER) for screening and treatment (PR) compared with no screening and no treatment was £2400 (€2745) per QALY gained. For screening and treating with DAAs compared with no screening and no treatment, the ICER was still cost-effective at £9139 (€10,455)/QALY gained. A Canadian study evaluated the cost-effectiveness of screening for HCV in different age groups and then treating with DAAs in the Canadian population where the HCV seroprevalence ranged from 0.3 to 0.8% [32]. The ICER for IFN-free DAAs vs. older therapies ranged from CAN$34,359 (€21,977) to CAN$44,034 (€28,166) per QALY gained [32]. In the US, Rein et al. found that screening followed by DAA therapy was moderately cost-effective with ICERs ranging from US$47,237 (€40,665) to US$72,169 (€62,128) per QALY gained [36]. The Drummond Criteria [24]: (i) Was a well-defined question posed in answerable form? (ii) Was a comprehensive description of the competing alternatives given (i.e., can you tell who did what to whom, where, and how often)? (iii) Was the effectiveness of the program or services established? (iv) Were all the important and relevant costs and consequences for each alternative identified? (v) Were costs and consequences measured accurately in appropriate physical units (e.g., hours of nursing time, number of physician visits, lost working days, and gained life years)? (vi) Were the cost and consequences valued credibly? (vii) Were costs and consequences adjusted for differential timing? (viii) Was an incremental analysis of costs and consequences of alternatives performed? (ix) Was allowance made for uncertainty in the estimates of costs and consequences? (x) Did the presentation and discussion of study results include all issues of concern to users?
Non-pan-genotypic DAA therapies have also been found to be moderately cost-effective, however they had a large budget impact at the 2015 cost of treatment [31]. Deuffic-Burban found that DAAs were moderately cost-effective for genotypes 1 and 4 regardless of fibrosis stage, ranging from €40,000 to €88,000 per quality-adjusted life year (QALY) gained ( Table 2). IFN-based regimens were estimated to be more cost-effective for genotypes 2 or 3 at €21,300 to €19,400 per QALY gained regardless of fibrosis stage [31]. DAAs were found to be moderately cost-effective for genotype 1 & 4 at a median threshold of €24,000/QALY gained and maximum upper limit of €80,000/QALY gained however, introducing these regimens on a wide scale would have a substantial budget impact of €3.5-7.2 billion on the French health care system [31,45]. Several US studies have evaluated the cost-effectiveness of DAA therapies compared to older PEG-INF-RBV therapies and found that DAA therapies were moderately cost-effective at a willingness-to-pay threshold of $50,000 US (€39,210), but varied significantly by HCV genotype, presence of liver fibrosis, and treatment history [33][34][35][36][37][38]. In the US, DAAs were moderately cost-effective for genotypes 1 & 4 whereas IFN-RBV was more cost-effective for genotypes 2 & 3. DAAs were more cost-effective in the presence of cirrhosis and treatment naïve patients. Providing DAAs to all eligible HCV patients would also have a huge budget impact, costing an additional US$65 (€56) billion over a 5-year period, whereas the resulting cost-offsets were estimated at only US$16 (€14) billion [34].

Discussion
The data in this review supports the effectiveness and cost-effectiveness of HCV screening in populations at risk for HCV infection, including migrants from intermediate and high HCV prevalence countries (anti-HCV ≥ 2% and ≥5%, respectively). Screening tests to detect the presence of HCV antibodies performed equally well in all populations. In the laboratory setting or at the point-of-care both are highly sensitive and specific [6,26]. DAA therapies are highly efficacious and well tolerated with >95% cure rates in a range of populations in different countries [7]. Achieving SVR with HCV therapy is associated with decreased risk and rate of liver disease progression, lower rates of HCC development, and improved survival [27][28][29]46]. At 2015 costs, DAA based regimens were only moderately cost-effective and as a result less than 30% of those with HCV had been screened and less than 5% of all HCV cases had been treated in the EU/EEA in 2015 [2,31]. Migrant populations in the EU/EEA face difficulties accessing care and treatment as a result of numerous barriers at the patient, provider, and health system level [16,47]. To reach HCV elimination goals in the EU/EEA dramatic scale up of HCV testing with diagnosis of all groups at HCV risk, including migrants, and linking those found to be positive to care and treatment will be required.
Migrants living in the EU/EEA bear a disproportionate burden of HCV [1,13]. They are older and more likely to have advanced liver disease and hepatocellular carcinoma compared to non-migrants at the time of HCV diagnosis [17,48,49]. This is likely due to missed or delayed diagnoses. In a survey in Finland, 63% of migrants found to be HCV positive had not been previously diagnosed [50]. Seventy percent of these HCV positive migrants had been living in Finland for more than five years. Similarly, in a population based study in Canada, it took a mean of 10 years after arrival for migrants to be diagnosed with HCV [48]. Another Canadian study found that it was cost-effective to screen immigrants for HCV followed by DAA treatment at an anti-HCV prevalence of 1.9%, which is the mean HCV seroprevalence of migrants living in the EU/EEA [12,13,51]. These data taken together suggest that early screening of migrants based on the HCV prevalence in the country of origin with linkage to care and treatment could prevent liver related sequelae in the migrant population and would be cost-effective. In an ECDC survey, 18 of 21 responding countries had national guidance on HCV testing; however, only six countries (29%) had guidance on testing migrants for HCV [52]. This highlights an important gap and an opportunity for health promotion among the migrant population.
Migrants face multiple barriers in accessing healthcare services resulting in gaps along all steps of the HCV care continuum [16,53]. Many groups of newly arriving migrants, including asylum seekers and undocumented migrants, lack entitlement to health care in the EU/EEA, thus preventing them from being diagnosed or receiving treatment [54]. Individual barriers include lack of knowledge and awareness of risk factors, fear, and stigma of blood-borne diseases, and socioeconomic, linguistic, and cultural barriers [47,[55][56][57]. Providers are frequently unaware that birth in an HCV endemic country is an important risk factor for HCV. Language and cultural discordance between patients and providers may lead to poor communication and low quality of care [16,53,58]. Lack of screening guidelines and programs and the lack of access to interpreters are important health system barriers. Barriers to effective screening and treatment programs may also include lack of political will to address migrant health issues or negative national attitudes toward migrants. Resulting policies may deny migrants the entitlement to health care or prevent the development of migrant-friendly heath care systems [58].
Evidence from primary studies of migrant populations suggests that HCV screening uptake and linkage to care could be improved by implementing decentralized community-based screening strategies, and cooperation with community-based organizations to overcome cultural and language barriers [59][60][61][62][63][64]. Furthermore, integrated point-of-care testing for HCV, HBV, and HIV increased testing uptake [59,65]. HCV screening and treatment programs for migrants in the EU/EEA will need to be tailored to their specific needs. In addition, it will be necessary to ensure universal access to health care in order to enhance uptake along the entire HCV care continuum.
The greatest barrier to scaling up HCV treatment in the EU/EEA has been the high cost of therapy, which has decreased dramatically in the past two years [66]. As a result, many countries in the EU/EEA have lifted restrictions and provide more widespread availability of HCV therapy in the region [67]. Dropping prices and simpler short treatment regimens have helped create an opportunity to eliminate HCV in the EU/EEA. Recent guidance from the WHO has highlighted the need to increase HCV case detection and linkage to HCV care [6]. This includes screening persons originating from countries with an intermediate (≥2%) and high (≥5%) HCV prevalence, which includes many of the migrants living in the EU/EEA. Consideration of existing prevention and control efforts and the capacity of existing systems must also be taken into account [68]. Knowledge of the HCV epidemiology in each EU/EEA Member State will be needed to identify those migrant groups at highest HCV risk, given that the top countries of origin of HCV infected migrants in each country varies (Table S6) [13]. Each country will need to assess their own capacity to increase HCV testing in at risk populations and to ensure that programs are in place that effectively link those with active HCV to care and provide HCV treatment.

Strengths and Limitations
The major strength of our study is that we used a systematic review process to identify relevant studies and the GRADE methodology to evaluate the certainty of the evidence. Our study had the following limitations. The findings were limited by the low to moderate certainty of the evidence of the included studies. Our study was also limited by the low number of studies reporting on linkage to care for migrants to EU/EEA. Finally, the included cost-effective studies modeled the 2015 cost of DDA therapies. We anticipate however, that HCV screening and treatment will be even more cost-effective given the dramatic decrease in cost of DAA therapy in the EU/EEA since 2015 [66,67].

Conclusions
In many EU/EEA countries migrants originating from intermediate and high HCV prevalence countries make up a large proportion of all HCV cases and have poorer liver related outcomes due to delayed diagnosis and treatment. This health disparity is due to the numerous barriers migrants face accessing HCV diagnosis, care, and treatment. Migrant focused programs will need to ensure entitlement to health services and will be most effective if they address linguistic and cultural barriers, are community-based, and integrated with screening for other diseases such as HIV and HBV. Although decreasing HCV costs have made treatment more accessible in the EU/EEA, HCV elimination will only be possible in the region if health systems include and treat migrants for HCV [66].
Supplementary Materials: The following are available online at http://www.mdpi.com/1660-4601/15/9/2013/ s1, Figure S1: Analytic Framework for HCV Screening in Migrants; Table S1. Effectiveness and Cost-effectiveness Search Strategy; Tables S2-S5. Study profile GRADE; Table S6. Chronic HCV burden in migrants: the 10 migrant groups from intermediate/high HCV prevalence countries with the highest number of HCV cases in host EU/EEA countries. The funding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.