The Epidemiology of Hepatitis C Virus in the Fertile Crescent: Systematic Review and Meta-Analysis

Objective To characterize hepatitis C virus (HCV) epidemiology in countries of the Fertile Crescent region of the Middle East and North Africa (MENA), namely Iraq, Jordan, Lebanon, Palestine, and Syria. Methods We systematically reviewed and synthesized available records of HCV incidence and prevalence following PRISMA guidelines. Meta-analyses were implemented using a DerSimonian-Laird random effects model with inverse weighting to estimate the country-specific HCV prevalence among the various at risk population groups. Results We identified eight HCV incidence and 240 HCV prevalence measures in the Fertile Crescent. HCV sero-conversion risk among hemodialysis patients was 9.2% in Jordan and 40.3% in Iraq, and ranged between 0% and 3.5% among other populations in Iraq over different follow-up times. Our meta-analyses estimated HCV prevalence among the general population at 0.2% in Iraq (range: 0–7.2%; 95% CI: 0.1–0.3%), 0.3% in Jordan (range: 0–2.0%; 95% CI: 0.1–0.5%), 0.2% in Lebanon (range: 0–3.4%; 95% CI: 0.1–0.3%), 0.2% in Palestine (range: 0–9.0%; 95% CI: 0.2–0.3%), and 0.4% in Syria (range: 0.3–0.9%; 95% CI: 0.4–0.5%). Among populations at high risk, HCV prevalence was estimated at 19.5% in Iraq (range: 0–67.3%; 95% CI: 14.9–24.5%), 37.0% in Jordan (range: 21–59.5%; 95% CI: 29.3–45.0%), 14.5% in Lebanon (range: 0–52.8%; 95% CI: 5.6–26.5%), and 47.4% in Syria (range: 21.0–75.0%; 95% CI: 32.5–62.5%). Genotypes 4 and 1 appear to be the dominant circulating strains. Conclusions HCV prevalence in the population at large appears to be below 1%, lower than that in other MENA sub-regions, and tending towards the lower end of the global range. However, there is evidence for ongoing HCV transmission within medical facilities and among people who inject drugs (PWID). Migration dynamics appear to have played a role in determining the circulating genotypes. HCV prevention efforts should be targeted, and focus on infection control in clinical settings and harm reduction among PWID.


Introduction
Hepatitis C virus (HCV) is a main risk factor for liver cancer and cirrhosis and is estimated to affect 1-3% of the population in most countries globally [1,2]. The vast majority of HCV transmission is blood-borne and largely preventable [3,4]. Recent breakthroughs in HCV treatment research suggest promising prospects for treating and containing the infection, and raise hope for an improved quality of life among infected individuals [5].
In few countries of the Middle East and North Africa (MENA) region, namely Egypt and Pakistan, HCV is prevalent at high levels: 14.7% in Egypt [6,7] and 4.8% in Pakistan [8,9]. Meanwhile, infection levels remain poorly estimated for most MENA countries including those in the Fertile Crescent (FC). Geographically, for this study, we define the FC subregion of MENA to include Iraq, Jordan, Lebanon, Palestine, and Syria. Since Egypt was covered by a separate systematic review published recently [7], we did not include Egypt as part of our study. FC countries share historical, socio-cultural, and geo-political similarities, and their health care systems are often directly or indirectly interlinked or affected by decades of political turmoil and refugee waves within this territory.
Our study primarily aims at characterizing HCV epidemiology in these countries by 1) systematically reviewing and synthesizing all available published and unpublished records of HCV incidence and prevalence among the different population groups, and 2) estimating the population-level country-specific HCV prevalence by pooling available HCV prevalence measures among the general population in each country. A secondary outcome of our study is an analysis of HCV genotype diversity. This work is conducted under the umbrella of the MENA HCV Epidemiology Synthesis Project; an ongoing effort to characterize HCV epidemiology and inform key public health research, policy, and programming priorities in MENA [7,[10][11][12][13][14][15].

Data sources and search strategy
We systematically reviewed all HCV incidence (Table 1) and prevalence (Tables 2 and 3 and  S2 Table) data in FC following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [16]. The PRISMA checklist can be found in S1 Fig. The methodology was adapted from a recently published protocol for a systematic review of HCV incidence and prevalence measures for countries in the Horn of Africa subregion of MENA [10,17]. Our search criteria can be found in S2 Fig. Briefly, we surveyed PubMed and Embase (up to April 23, 2015), different regional databases (up to November 5-16, 2014), the abstract archives of the International AIDS Society conferences (up to November 16, 2014), and country-level and international organizations' reports available through the MENA HIV/AIDS Epidemiology Synthesis Project database (up to April 23, 2015) [18,19]. We used a broad search criteria with no language restrictions. PubMed and Embase were searched using text terms and MeSH/Emtree terms exploded to cover all subheadings. Our search was restricted to articles published after 1989, the year when HCV was first discovered [20,21].

Study selection
The search results were imported into a reference manager, Endnote, where duplicate publications were identified and excluded. The titles and abstracts of remaining records were screened for relevance, and full-texts of articles deemed relevant or potentially relevant were retrieved for further screening. The references of all articles and all reviews of literature were also checked to identify additional articles that could have been potentially missed. Any document reporting a measure of HCV incidence and/or prevalence based on primary data was considered for the review (Fig 1). Case reports, case series, editorials, letters to editors, commentaries, studies referring to hepatitis C as non-A non-B hepatitis, and records of HCV among military personnel who were stationed in this region for some time, but are from countries outside the region, were excluded from our review. A secondary independent screening for articles reporting HCV genotype information, irrespective of their inclusion of HCV incidence and/or prevalence measures, was also performed (S3 Fig). In this work, a "report" refers to a document containing outcome measures of interest, while a "study" refers to all details pertaining to a specific outcome measure. Accordingly, one report could contribute multiple studies, and multiple reports of the same study, after being identified as duplicates, are considered as one study.  Data extraction and data synthesis Data from relevant reports were extracted by HC. To check for consistency, 10% of the reports were double extracted by another co-author (KC). The list of extracted information can be found in S1 Table. The extracted data were then synthesized by country and by populations' risk of acquiring HCV (high, intermediate, and low risk). Our classification of risk was informed by HCV epidemiology literature [2,22,23] and a previously published classification [12]. In events where the level of exposure to HCV in certain populations was uncertain, and hence difficult to classify within the three pre-determined levels of risk, we followed a conservative approach that grouped such populations into a distinct risk category. We labeled this category as 'special clinical populations' since all of these populations were clinical populations. Details pertaining to the definitions and risk classification of identified populations can be found in S4 Fig. The quality of HCV incidence and prevalence measures identified through our review was determined by assessing the risk of bias (ROB) for each study using the Cochrane approach

Quantitative analysis
For each country, HCV antibody and RNA prevalence data in studies comprising at least 50 participants were stratified by population risk and summarized using reported ranges (Tables 2  and 3, and S2 and S3 Tables). Meta-analyses of HCV prevalence measures were conducted, for each country, for the various populations at risk whenever at least five HCV measures were available. Studies comprising a minimum of 25 participants were considered eligible for inclusion in the meta-analysis. HCV prevalence for the total sample was replaced with stratified measures whenever the sample size requirement was fulfilled at the level of each stratum. For each study, one final stratification was considered based on a pre-defined sequential order that prioritizes nationality followed by population, sex, year, region, and age. The variance of the prevalence measures was stabilized using the Freeman-Tukey type arcsine square-root transformation [25]. Estimates for HCV prevalence were weighed using the inverse variance method and then pooled using a DerSimonian-Laird random effects model [26]. This model assumes a normal distribution of true effect sizes across studies, and therefore accounts for the sampling variation as well as the heterogeneity in effect size [27]. Heterogeneity measures were also calculated (Table 4) [27,28].
For each country, we calculated the frequency of occurrence of each genotype relative to all identified genotypes (Table 5). Here, individuals testing positive for multiple genotypes contributed separately to the quantification of each of the identified genotypes. Shannon Diversity Index [29] was applied to the resulting distributions to assess the diversity of genotypes across

Search results
Fig 1 describes the process of study selection based on PRISMA guidelines [16]. We identified a total of 465 citations (92 through PubMed, 153 through Embase, and 220 through regional databases). A total of 166 records were identified as relevant or potentially relevant after removing duplicates and screening the titles and abstracts of remaining records. Out of these, 115 records were eligible for inclusion in the systematic review during the full-text screening stage. Twenty additional records were identified through screening the bibliographies of studies and reviews. Twenty one HCV prevalence measures from country-level data were also identified through the MENA HIV/AIDS Epidemiology Synthesis Project database [18,[32][33][34][35][36][37][38]. To sum, a total of 135 eligible reports and 21 HCV prevalence measures from county-level data *Q: the Cochran's Q statistic is a measure assessing the existence of heterogeneity in effect size.
**τ 2 : the estimated between-study variance in the double arcsine transformed proportions of the true effect sizes. The back-transformed τ 2 was not calculated as the methodology to do so is not currently available. † I 2 : a measure assessing the magnitude of between-study variation that is due to differences in effect size across studies rather than chance. ‡ Prediction interval: estimates the 95% interval in which the true effect size in a new HCV study will lie.
doi:10.1371/journal.pone.0135281.t004 were included in the systematic review. These yielded 8 HCV incidence/sero-conversion and 240 HCV prevalence measures. All 465 citations identified through our search underwent an independent secondary screening for HCV genotype studies (S3 Fig). A total of 163 records were identified as relevant or potentially relevant after removing duplicates and screening the titles and abstracts of remaining records. Out of these, 29 records were eligible for inclusion in this secondary systematic review during the full-text screening stage yielding 47 HCV genotype study measures.

HCV incidence overview
Out of all countries being analyzed (Iraq, Jordan, Lebanon, Palestine, and Syria), data on HCV incidence was only available for Iraq and Jordan. There were no studies reporting strictly HCV incidence rate per person-year. All eight identified incidence measures were based on risk of sero-conversion over a specified time frame (Table 1). One study conducted across all dialysis units in Jordan reported a sero-conversion risk of 9.2% over one year [39]. The rest of the studies were conducted in Iraq. In one study among hemodialysis patients, a sero-conversion risk of 40.3% over one year was reported [40]. Lower HCV sero-conversion risks ranging from 0-3.5% were observed among pediatric cancer patients [41][42][43]. A study among newborns to HCV infected mothers reported 100% clearance of maternal antibodies by six months of age [44].

HCV prevalence overview
We present here an overview of HCV prevalence studies identified in Iraq, Jordan, Lebanon, Palestine, and Syria. For the purpose of this analysis, we excluded HCV prevalence studies that were based on samples of less than 50 participants (37 studies). Accordingly, we report on a total of 203 HCV prevalence studies including 125 studies from Iraq, 15 from Jordan, 27 from Lebanon, 17 from Palestine, and 19 from Syria.
For the general population (Table 3), HCV prevalence varied between 0% and 5.1%, and ranged between 0% and 2.8% among blood donors, 0% and 5.1% among pregnant women, and was 0.4% in a national household-based survey [59] and overall less than 1% among other general population groups.
Jordan. Among populations at high risk (Table 2), high HCV prevalence levels ranging from 21% to 49.8% were observed among hemodialysis patients. A prevalence of 32.8% was also observed among thalassemic patients [68].
For the general population (Table 3), low HCV prevalence levels of less than 1% were observed among blood donors and people attending health centers (range: 0.1-0.9%). Among populations at intermediate risk and special clinical populations (S2 Table), low HCV prevalence of 0.6% was reported among health care workers [69] and high prevalence of 40.5% was reported among patients with hemolytic anemia [70].
Rather low HCV prevalence was observed among populations at intermediate risk ranging from 0% among female sex workers and men who have sex with men to 7.7% among HIV patients infected apparently via the sexual route (S2 Table). Among special clinical populations (S2 Table), HCV prevalence was 19.6% [85] among hepatocellular carcinoma patients and 0% [82,86] among patients with other malignancies.
Palestine. Among populations at high risk (Table 2), HCV prevalence was measured at 27.4% among hemodialysis patients in the West Bank [87], and at 17.9% among hemodialysis patients in the Gaza strip [88]. A prevalence of 45.2% was reported among PWID in East Jerusalem [89].
For the general population (Table 3), HCV prevalence ranged between 0.3% and 0.9% among blood donors. Among populations at intermediate risk (S2 Table), HCV prevalence ranged between 2% among female sex workers [97] and up to 3.8% among health care workers [98]. The only study among special clinical populations (S2 Table) was among acute viral hepatitis patients and reported a prevalence of 1.0% [99].
Overall HCV prevalence ranges and medians. For all countries, HCV prevalence ranged from 0-75% among populations at high risk with a median of 25.0%, from 0-9% among populations at low risk with a median of 0.4%, from 0-35.1% among populations at intermediate risk with a median of 1.2%, and from 0-71.9% among special clinical populations with a median of 3.4%.
There was significant evidence for heterogeneity in effect size in all meta-analyses, the pvalue was always <0.001 (Table 4). Most of the variation in all meta-analyses was attributed to variation in effect size across studies rather than chance (I 2 >60%; Table 4).

HCV RNA prevalence
Our search identified a total of 34 HCV RNA measures (S3 Table). Generally high HCV RNA prevalence was found among HCV antibody positive individuals at high risk across all countries ranging from 26.1% among hemodialysis patients to 88% among thalassemia patients. HCV RNA prevalence was also substantial among HCV antibody negative individuals at high risk ranging from 3.5% among hemodialysis patients to 16.7% among thalassemic children.
For the general population (S3 Table), the prevalence of HCV RNA among HCV antibody positive individuals in Iraq and Palestine hovered around 65% in different samples of blood donors [90,100,101], pregnant women [102], and outpatients [90], but was at 89.7% among blood donors in Jordan [103].

HCV genotypes
Information on HCV genotypes was available in 47 studies including a total of 1749 HCV RNA positive individuals (Table 5 and S6 Fig). In Iraq, the majority of participants were infected by a single strain, only 11.3% had multiple genotypes isolated from their plasma (S6A Fig) [45,48,101,102,[104][105][106][107][108][109][110]. The distribution of HCV genotypes showed a high prevalence for genotype 4 (62.5%) and genotype 1 (33.2%) followed by genotype 3 (3.8%) and genotype 2 (0.5%) ( Table 5). Subtypes 1a and 1b were isolated, respectively, from 33.2% and 66.8% of infected persons with genotype 1 (where subtype information was available). About two-third of the infections among populations at high risk and slightly more than half of those among general populations were attributed to genotype 4.
All HCV RNA positive individuals in Jordan were infected by a single strain (S6B Fig), with 57.7% carrying genotype 4 and 42.3% carrying genotype 1 (Table 5) [68,103,111,112]. Subtypes 1a and 1b were isolated, respectively, from 54.5% and 45.5% of infected persons with genotype 1. The majority of genotype 1 samples were among hemodialysis and thalassemia patients, while samples with genotype 4 were equally distributed between populations at high risk and the general population.
The largest possible value for the Shannon Diversity Index, assuming equal distribution of the seven known HCV genotypes [121], is 1. 95 [29]. Lebanon scored highest on this index (H = 1.29 out of 1.95 or 66.5%) indicating high diversity in the types and frequency of isolated genotypes ( Table 5). The next most diverse genotype distributions were found in Syria

Risk factors for HCV infection
Few studies (n = 5) investigated possible risk factors for HCV infection using multivariable logistic regression analyses. Common reported risk factors among hemodialysis patients included duration of dialysis and number of blood units transfused [54,94]. Among individuals presenting to laboratory units for diverse testing in Lebanon, hemodialysis was the main risk factor for HCV infection [84]. Meanwhile, maternal age and number of previous miscarriages were identified as the main risk factors for HCV infection among pregnant women in Iraq [102]. Among patients suspected of having acute viral hepatitis, blood transfusion, tattooing, and cupping were identified as risk factors for HCV infection in a study in Iraq [122].

Quality assessment of HCV incidence and prevalence measures
The results of the quality assessment conducted for the HCV incidence and prevalence measures are summarized in S4 Table. The study-specific quality assessment for each of these measures is found in S5-S8 Tables.
The majority of HCV incidence studies involved samples of less than 100 patients, and thus were classified as having low precision. All studies were based on convenience samples selected from clinical facilities and were characterized by a high response rate. The type and generation of the biological assays used to ascertain infection were specified in almost all studies, with more than half relying on the more recent, sensitive and specific 3 rd and 4 th generation ELISA tests.
The majority of HCV prevalence measures had high precision (81.3%). The ROB assessment was possible for 184 HCV prevalence measures, while 19 measures were judged to be of unknown quality as limited description of the samples was available and were excluded from further analyses. However, all studies with unknown ROB were based on testing among blood donors reported by ministries of health and blood banks. For the vast majority of measures (98.4%), HCV ascertainment was based on biological assays. Information about the generation of the used assay was missing for more than half of the studies. Among studies with information on the assay's generation, 85.2% reported the use of the more sensitive and specific 3 rd and 4 th generation assays. Most of the studies were drawn using convenience sampling. Half of the studies had low ROB in the response rate domain, while 3.8% had high ROB. Response rate information was missing for 46.2% of the measures.
Overall, HCV incidence and prevalence studies were of reasonable quality, but incidence studies were of lower precision. Nearly all studies had low ROB in at least one quality domain (99.0%), and 47.9% of all studies had low ROB in at least two of the three quality domains. Less than 3.6% of all studies had high ROB in two or three quality domains.

Discussion
We presented to our knowledge the first systematic review and synthesis of HCV incidence and prevalence across countries of the FC subregion of MENA. We also provided estimates for HCV prevalence at the national level and for various at risk populations. Our results suggest that HCV prevalence is at less than 1% among the general population across these countries (Table 4). Meanwhile, high prevalence levels were noted among hemodialysis, thalassemia and multi-transfused patients ( Table 2). Genotypes 4 and 1 seem to be the most common genotypes circulating in these countries (Table 5 and S6 Fig). Our estimates for the national HCV prevalence of less than 1% suggest that HCV levels in these countries are comparable to those in developed and other developing countries, but are towards the lower end of the global range [1,2,123]. These estimates also suggest that this subregion appears to have the lowest HCV prevalence of all MENA subregions [7,[11][12][13][14][15]]-substantially lower than in other MENA countries namely Egypt [6,7] and Pakistan [8,9].
A large fraction of HCV prevalence measures in the general population were among blood donors, and HCV prevalence in this normally healthy population may underestimate HCV prevalence in the population at large. In the USA, for example, HCV prevalence among blood donors [124] is substantially lower than that estimated in nationally-representative population-based surveys [125]. We conducted sensitivity analyses (not shown) to examine the impact of excluding blood donor data on our pooled estimates for HCV prevalence. For Iraq, HCV prevalence increased substantially from 0.2% to 0.8% with the exclusion of blood donors. Meanwhile, there were minor changes to the prevalence in Jordan (0.3% versus 0.2%) and Lebanon (0.2% versus 0.3%). Sensitivity analyses were not possible for Palestine and Syria due to the small number of studies among other general population groups (1 study in Palestine and 0 studies in Syria).
Overall, these results support our conclusion that HCV prevalence is likely to be below 1% among the population at large in each of the FC countries. It bears notice that our estimate for HCV prevalence in Iraq is substantially lower than that estimated recently using a different methodology in a global study (0.2% in our study versus 3.2% in Gower et al. study [22]). The difference in estimates arises from Gower et al.
[22] basing their estimate on a single study [102] while our pooled prevalence was based on 99 studies, a large fraction of which were identified through our expanded search of data in non-indexed journals.
Despite the lower prevalence among the population at large, our findings indicate on-going transmission in at least some medical facilities. Exposures to hemodialysis [54,84,94] or blood transfusion [54,122] were cited as risk factors for acquiring HCV. High HCV prevalence was generally observed among hemodialysis and thalassemia patients across countries (Table 2). High sero-conversion risks of up to 40% were found among hemodialysis patients in prospective studies (Table 1). Measurable HCV RNA prevalence among HCV antibody negative clinical populations were documented suggesting recent nosocomial HCV exposures (S3 Table). Higher HCV levels (compared to the general population) among health care workers, hospitalized populations, different clinical populations, diabetics, and dental providers were identified suggesting health care related HCV exposures (S2 Table). While some of the prevalent HCV exposures may date to times before the enforcement of more stringent blood screening and infection control protocols in these countries, the totality of the evidence supports the conclusion of ongoing transmission in some clinical settings and through medical procedures. These results emphasize the importance of reinforcing blood screening and infection control in health care facilities.
Other HCV risk factors identified in this review included exposure to infected needles and sharp objects during tattooing and cupping procedures [122]. The latter, also known as hijama, is a common practice by community healers in MENA [18,126]. There is also evidence elsewhere in MENA of a wider scope of community related HCV exposures as well as traditions of some professions, such as barbers, administering health care related procedures such as injections [7,18,127]. These findings highlight the relevance of controlling these exposures and insuring the safety of injections. This could be done, for example, by wide-scale replacement of current re-usable syringes with safety-engineered "smart" syringes [128,129].
The ongoing political and military conflicts in this subregion (Iraq, Palestine, and Syria) introduce more challenges and serious barriers to the implementation of stringent blood screening and infection control in medical facilities and in communities. Shortages of medical supplies and trained human capital in these countries are often exacerbated by intermittent electricity and water supplies making the provision of even basic medical care difficult [42, [130][131][132]. Implications of these emergencies on current HCV transmission are unknown.
Injecting drug use appears to be the dominant driver of HCV incidence in many developed countries [133][134][135]. The estimated population proportion of injecting drug use among the adult population in FC ranges between 0.07% in Syria and 0.35% in Palestine [136]. The population proportion of injecting drug use in MENA as a whole is estimated at 0.24%; in the intermediate range compared to global levels [136]. Our search identified only few studies assessing the status of HCV infection among PWID and other drug users [18,71,89]. These showed HCV prevalence varying between 21% in Syria and 52.8% in Lebanon (Table 2). Relatively low HCV prevalence has been observed among prisoners, a fraction of whom are often drug users, in Iraq (0.6%) [137] and in Lebanon (3.4%) [113]. HCV prevalence among PWID in FC seems thus somewhat lower than that observed in other MENA countries and globally [23,136]. With these overall limited data on HCV among PWID, it is difficult to judge the relative importance of injecting drug use as a driver of HCV incidence and prevalence in FC.
Migration dynamics within MENA may have played a role in HCV epidemiology in this subregion. This is specially so in relation to the epidemic in Egypt, the country with the largest HCV prevalence worldwide [6,7]. The epidemic in Egypt is almost exclusively dominated by genotype 4 (>90%) [138], which is also a highly prevalent genotype in FC (Table 5). Genotype 4 in MENA appears to have a larger frequency in countries geographically close to Egypt and/ or in countries that host or have hosted large migrant labor populations from Egypt [11,12]. Iraq in particular, but also Jordan and Lebanon, have hosted or are currently hosting millions of Egyptian migrant labor [139][140][141][142]. A large fraction of the labor market in Iraq at some point consisted of Egyptian migrant workers who formed the backbone of the major infrastructure and developmental projects and agricultural sector in this country-such as during the era of the Iraq-Iran war [139][140][141]. Movement of Egyptian labor, who often came from rural areas most affected by HCV [7], may have contributed to higher circulation of genotype 4 in host countries, but has had a limited impact on HCV prevalence in these countries. FC countries appear to have some of the lowest HCV prevalence levels globally.
The Gaza strip, which borders Egypt, seems to be the part of FC most influenced by HCV dynamics in Egypt. The Gaza strip was under Egyptian rule prior to the 1967 Israeli occupation and shares strong historical, cultural, and economic ties with Egypt. With the protracted Israeli siege of Gaza, Egypt has become the main gateway for Gaza residents to the world and the country that provides essential services to the people of Gaza. Gazans often seek medical treatment in Egypt for conditions that cannot be treated in this largely impoverished district. It is probably not surprising therefore that genotype 4 appears to be the dominant genotype in Gaza, and specially so among those with a history of travel, or of receiving medical care, including blood transfusions or surgical operations, in Egypt [118].
A substantial fraction of genotype 3 infections in FC were observed among PWID, affirming a global link between this genotype and injecting drug use [121,138,143]. Apart from Lebanon, very limited number of genotype 3 infections were identified (Table 5). One reason could be that PWID contribution to HCV incidence is not large in these countries, and/or that genotypes 4 and 1 are the dominant genotypes circulating among PWID. An alternative explanation, and possibly a more plausible one, is that existing studies of genotypes have not yet reached this largely hidden population whose contribution to HCV incidence and prevalence remains poorly understood.
Lebanon and Syria showed higher genotype diversity with genotype 5 only detected in these countries, and genotype 2 being more prevalent in Lebanon than in the rest of FC (Table 5). These findings are consistent with the large scale emigration rates and diaspora populations of these two countries globally-specifically so for Lebanon. Such population movements may provide avenues for genotypes that are prevalent in other parts of the world, say in West Africa and Latin America for genotype 2 [121,141] and South Africa for genotype 5 [121,138], to be introduced into circulation in some of these countries.
Our systematic review and meta-analyses are limited by the variability in quantity and quality of studies across countries. There were few studies that assessed HCV incidence and some countries did not have sufficient number of prevalence measures in some of the at risk populations to warrant conduct of meta-analyses. With the convenience sampling in most studies, and the country-specific variability in frequency, specific population and geographical location of available data, the estimates may not be representative of the true HCV population prevalence. Data was also limited for some of the specific populations that play an important role in HCV epidemiology, most notably PWID. HCV RNA measures were limited in quantity and were often based on relatively small samples.
The number of genotyped infections was not large in all countries and available genotype data may not have captured the diversity and true frequency of genotypes across the different at risk populations in each country. For example, there were some differences, though minor, between the genotype frequencies identified here compared to those of two recent studies of global genotype distribution [22,121]. The main source of differences appears to be the inclusion of more studies on HCV genotypes in our review, largely as a consequence of including regional and national databases in our search of non-indexed published literature (S3 and S6 Figs).
Our meta-analyses highlighted heterogeneity in effect size among included studies. This is not unexpected considering the differences in prevalence measures in terms of the specific population studied, sampling methodology and participant recruitment, age-group representation in the sample, year of study, location and assay used [12]. However, due to the relatively small number of outcome measures for each country, we were unable to conduct a meaningful multivariable meta-regression analysis to identify sources of heterogeneity in effect size.

Conclusions
HCV prevalence in the population at large in FC appears to be below 1%, and is tending towards the lower end of the global range. HCV prevalence among high risk populations, such as hemodialysis and thalassemia patients, is considerable and the evidence indicates ongoing HCV transmission in at least some medical facilities. Genotypes 4 and 1 appear to be the most prevalent, and genotype distributions suggest a role for recent migrations in the circulation of different genotypes across national borders. With the limitations in available data, infection levels among PWID and the contribution of this population to HCV incidence and prevalence remain uncertain. The implications of the ongoing military conflicts and ensuing emergencies in FC on current HCV transmission are unknown.
Further research is needed for a better understanding of the epidemiology of this infection in this MENA subregion. Elucidation of the role of key populations in HCV transmission, such as PWID, is needed. Conduct of nationally-representative population-based surveys to estimate HCV prevalence, delineate spatial variability in prevalence, identify modes of exposure, and assess HCV knowledge and attitudes, as has been done recently in Egypt [6,[144][145][146][147][148], will enhance our knowledge of the epidemiology of this infection in this part of the world.
Our findings inform planning of health service provision, articulation of HCV policy guidelines, and implementation of effective HCV programming to reduce HCV transmission and decrease the burden of its associated diseases. Since HCV exposures in FC are linked to specific settings, HCV prevention efforts should be targeted and focus on infection control in clinical settings and harm reduction among PWID. Adoption of the new World Health Organization guidelines for the use of safety-engineered syringes [128,129] can also minimize exposure to HCV and other blood-borne pathogens.  Table. Lists of variables extracted from relevant reports with hepatitis C virus (HCV) incidence and/or prevalence information or with HCV genotype information.