Harm reduction program and hepatitis C prevalence in people who inject drugs (PWID) in Iran: an updated systematic review and cumulative meta-analysis

Background Prevalence of hepatitis C virus (HCV) infection among people who inject drugs (PWID) in Iran is high. Since 2005, the Iranian government has implemented a harm reduction program to control HCV. We aimed to describe the prevalence of HCV antibody (Ab) in Iranian PWID before and after the implementation of harm reduction with cumulative meta-analysis. Methods Following PRISMA guidelines, we conducted a systematic review and meta-analysis of studies published on the seroprevalence of HCV among PWID. We systematically reviewed the literature to identify eligible studies up to December 2018 in international and national databases. Pooled prevalence and 95% confidence intervals were calculated using Der Simonian and Laird method, taking into account conceptual heterogeneity. Subgroup analyses were performed by harm reduction implementation and studies’ characteristics to assess the sources of heterogeneity. We used Cochran–Armitage test for the linear trend of the prevalence of HCV Ab among PWID. Results We reviewed 5966 papers and reports and extracted data from 62 eligible records. The pooled HCV Ab prevalence among PWID in Iran was 46.5% (95% confidence interval [95% CI] 41.1–52.0%). Overall, the Cochran–Armitage test for trend indicated a significant decreasing trend of HCV Ab prevalence (P = 0.04). The cumulative meta-analysis showed a slight decline in the prevalence of HCV Ab between the years 2005 and 2018. Conclusions The HCV Ab prevalence among PWID in Iran is high, with a considerable geographical variation. The prevalence of HCV Ab among PWID in Iran slightly decreased after 2005 which could be, at least to some extent, related to the implementation of extensive harm reduction programs in the country.


Introduction
Hepatitis C virus (HCV) infection is the main global health concern. The global prevalence of viraemic HCV infection is estimated to be 1% in 2015, corresponding to 71.1 million (62.5-79.4) viraemic infections [1]. Evidence shows the incidence of three to four million people each year, and millions of people with chronic infection are at risk of developing cirrhosis and liver cancer [2]. HCV is transmitted widely through injection of the drug; however, unsafe medical injections were the predominant route of transmission in developing countries [3]. No effective HCV vaccine is yet available. Therefore, public health interventions are the only means of preventing HCV, especially in high-risk populations. People who Open Access *Correspondence: alavian@thc.ir 1 Baqiyatallah Research Center for Gastroenterology and Liver Diseases (BRCGL), Baqiyatallah University of Medical Sciences, Tehran, Iran Full list of author information is available at the end of the article inject drugs (PWID) are particularly at risk for contracting HCV and other blood-borne diseases, including those caused by human immunodeficiency virus (HIV) and hepatitis B virus (HBV) [4][5][6]. Opioid substances are associated with the highest drug-related burden [7,8].
Injecting drug use has also expanded in the world in the last decades, and it is estimated that 15.6 million people inject drugs worldwide [9]. According to a Harm Reduction International report in 2013 and other sources, it is estimated that there are 170,000 to 230,000 PWID in Iran [10][11][12][13]. Opium has traditionally been used in Iran, but rapid changes in drug use patterns have occurred in recent decades, leading to an increased number of PWID [14,15]. Iran has one of the highest numbers of people who use drugs (PWUD); many of them have a history of drug injection [14][15][16]. In Iran, a history of drug use, especially drug injection, is the main risk for transmission of HCV [17,18]. World Health Organization has set a goal for the elimination of HCV by 2030 which needs continuous harm reduction for containment of HCV epidemics accompanied by "test and treat" initiatives in the target populations such as PWID (HCV micro-elimination) in Iran [19,20]. More specifically, the prevalence of HCV antibody (Ab) among PWID in Iran has been reported 40-75%, depending on the study location [21]. A systematic review of PWID between 2000 and 2016 showed that the prevalence of HCV Ab in PWID in Iran was 42% (95% CI 33-52%) [22].
In Iran, during the end of 1990s and early 2000s, drug injection was common and general knowledge of PWUD on blood-borne infections was low [23,24]. In 2002, Iran implemented a harm reduction program including opioid substitution treatment (OST), needle and syringe programs (NSP), outreach, and prison-based programs, primarily in response to the HIV epidemics among PWID [25]. These programs included education, OST by methadone and buprenorphine, and ensuring access to sterile syringes and needles, as well as condoms [25]. The need for a harm reduction program originated in a survey of 900 street drug users which found that the HIV prevalence was 25% among PWID [26]. These programs have been scaled up at the national level of Iran and assigned as an official policy since 2005 [25,27]. The Iranian Ministry of Health reported that in 2007, 654 centers provided methadone maintenance therapy (MMT) and buprenorphine maintenance therapy (BMT) to 20,000 and 2,000 opioid dependents, respectively. In prisons, 11,000 prisoners were receiving MMT [28]. According to the report of the Ministry of Health, in the year 2011, thousands of methadone clinics, mainly in the private sector, were providing OST and about 500,000 PWUD were receiving OST. Moreover, tens of thousands of prisoners were receiving MMT [29]. About the needle and syringe program, in 2007, a total of 120 drop-in-centers (DICs) and 150 outreach teams distributed 1,400,000 needles and syringes [28]. In mid-2011, hundreds of sites were providing needles and syringes, as well as condoms, and in 1 year, millions of syringes were distributed [30][31][32]. Although there is sufficient evidence on the effectiveness of NSP and OST in the reduction in self-reported injecting risky behavior [32][33][34], evidence regarding the effectiveness of these programs in the containment of HCV epidemics is inadequate in Iran [32,33].
Previously, a systematic review had been conducted on the HCV prevalence among PWID in Iran, and the included studies were conducted from 2000 through 2016 [22]. The current systematic review was conducted to: (1) provide an updated estimate of the HCV Ab prevalence among PWID in Iran; (2) estimate the HCV Ab prevalence in study settings and geographical locations; (3) analyze the trend of HCV Ab prevalence over time; and 4) compare the HCV Ab prevalence before and after the implementation of harm reduction program in Iran.

Methods
An extensive and comprehensive search was done in the international (PubMed, ISI Web of Science, Scopus, Embase), regional (IMEMR), and national (Barakat Knowledge Network System) databases in December 2018. To access studies not yet published by that date, expert authors in this field were contacted. Moreover, the reference lists of the studies included in the final analysis as well as previous systematic reviews conducted in Iran were reviewed [22,35]. In this review, PWID was defined as people who have injected illicit drugs at least once during the past 12 months. The methods used were following the PRISMA [36] and GATHER guideline [37] (Checklist presented in Appendix 1). The protocol was registered in PROSPERO (CRD42018104303).

Search strategy
A broad search was conducted using the MeSH terms and text words (and their combinations and truncated synonyms) of geographical location (i.e., country and province names) and HCV in electronic data sources including PubMed, EMBASE, Education Resource Information Center, MEDLINE, MEDLINE in process, Psy-cINFO, Scopus, and Web of Science, Iranmedex, Google Scholar, Iranian Data Bank of Hepatitis Research, Scientific Information Database (SID), Magiran, and the Iran Blood Transfusion Journal. In our search strategy, we did not apply any limitations in the time of publication and language. Iranian databases were searched using the related keywords and the Persian equivalents of HCV, considering all possible combinations. We reviewed the titles and abstracts to select potentially relevant papers.
If there was doubt about the suitability of the paper based on the abstract, the full text was reviewed. We manually searched the references and relevant articles for inclusion. We also looked at the electronic abstract list of congresses conducted in Iran and also at the electronic database of students' thesis through universities' electronic libraries and Web sites. Furthermore, we searched and identified studies that were not captured by our database by reviewing the previously published metaanalyses and the reference lists in retrieved articles [22,35,38]. The search strategy is presented in Appendix 2.

Study selection
All records identified through our search were imported into an EndNote library where duplicate publications were identified and excluded (Fig. 1). Similar to our Fig. 1 Flowchart of the search and selection process previous systematic reviews [39], the remaining unique reports underwent two stages of screening, performed by AR and HSH. The titles and abstracts were first screened, and those deemed relevant or potentially relevant underwent further screening, in which the full texts were retrieved and assessed for eligibility, based on our inclusion and exclusion criteria. Eligible reports were included in this study, while the remaining ineligible reports were excluded for reasons indicated in Fig. 1. The references of all full-text articles and literature reviews were also screened for further potentially relevant reports.

Eligibility criteria
The objective of screening was to identify the scientific documents with epidemiological evidence on the prevalence and/or incidence of HCV Ab among PWID. The following five eligibility criteria were used in the selection of relevant studies: document type, study design, disease area, geographical setting, and population.
Concerning the document type, all scientific documents reporting original data (i.e., gathered directly by conducting surveys and laboratory tests on specimens), in the form of a peer-reviewed manuscript, progress report, abstract, technical report, or substantive scientific commentary, were included. Documents not reporting epidemiological data (e.g., legal cases, legislation) and those not reporting original data (e.g., data simulation), as well as documents lacking scientific and methodological details needed for the assessment of the validity of findings (e.g., media reports), were excluded. Documents that were self-described as "systematic reviews" or "metaanalyses" (scientific documents such as peer-reviewed manuscripts or abstracts summarizing results from a group of epidemiological studies) were retained for hand searching of the references. All study designs reporting data on the prevalence and incidence of disease, including cross-sectional studies, cohort studies, population case-control studies, and even experimental studies (e.g., randomized controlled trials) in which a biological survey was conducted to test for HCV before the introduction of the intervention (baseline) were included. Case reports, case series, and qualitative studies were excluded. For the disease area, studies with biological evidence on infection with HCV were included. Self-reported data on the diseases were excluded. The geographical scope was limited to studies conducted within Iran; the studies conducted among Iranian populations residing outside of Iran were not included. For the study population, studies conducted in human subjects who self-identified as current or former drug users were included. Studies on populations who were in short-term mandatory drug treatment and rehabilitation detention centers were included. Studies in non-human subjects, blood donors, dialysis patients, pregnant women, families or sexual partners of HCV/HBV patients, and populations with other chronic diseases (e.g., studies reporting HCV among people with liver cancer) were excluded. No limits were set on study implementation or publication date.

Data extraction, quality assessment, and risk of bias
Two authors (AR and HSH) reviewed the retrieved studies independently, and the following information was extracted: name of the first author, year of publication, date of study, location of study, total sample size, the prevalence of HCV Ab, recruitment setting, and recruitment method. The third author (SMA) was determined as the arbiter to resolve any disagreements.
Two independent investigators (AR and HSH) assessed the quality of the studies using the Critical Appraisal Skills Programme (CASP) tools for observational studies [40] and a critical appraisal tool for prevalence studies [41]. We used the following criteria for critical appraisal: (1) sampling method; (2) recruitment setting; (3) presenting the type of test; (4) presenting the definition of PWID; and (5) refusal rate. The documents were evaluated in terms of the above-mentioned criteria. Discrepancies were resolved by discussion or through consultation with the third investigator (SMA).

Data analysis
The extracted data were entered into the Excel software. Then, Stata 16.0 was used for analysis. Pooled prevalence and 95% confidence intervals were calculated using Der Simonian and Laird method, taking into account conceptual heterogeneity, and I 2 , Tau 2 , and X 2 were applied to assess heterogeneity among studies. The "metaprop" command was used to calculate the pooled prevalence of HCV Ab and the prevalence in different subgroups, by available geographical locations and study settings. The pooled prevalence of HCV Ab before 2005 (before scaleup of the harm reduction program at the national level) and during and after 2005 (after scale-up of the harm reduction program at the national level) was presented in a forest plot, and the heterogeneity of studies conducted in each period was estimated. The Q test was applied to assess heterogeneity between periods. Furthermore, the presence and the effect of publication bias were investigated using a combination of the visual inspection of funnel plots and Begg's and Egger's tests. Moreover, a trim-and-fill analysis was performed to assess the stability of overall prevalence when the results suggested obvious publication bias. The prevalence of HCV Ab for each province was also calculated and depicted on the map, using the GIS software.
To assess differences in the accumulation of evidence for HCV Ab prevalence in PWID in Iran, cumulative meta-analyses were conducted. The cumulative metaanalysis provides cumulative pooled estimates and 95% CIs. As studies are successively added, the overall prevalence and 95% CIs are recalculated providing evidence of the evolution of HCV Ab prevalence over time. To assess the sequential contributions of studies and evaluate changes in HCV Ab prevalence over time, studies were added alphabetically by years of implementation to a random-effects model using the "metacum" user-written command in Stata version 16.0. Besides, the sequential contributions of studies were evaluated before 2005 and during and after 2005 for HCV Ab prevalence over time by cumulative meta-analysis.
For investigating the trend of HCV Ab prevalence over time, the pooled prevalence every 3 years was estimated and presented in a line graph because the number of studies conducted in each year was few. The Cochran-Armitage test for linear trend was used by Winpepi software to test the variation in the prevalence of HCV Ab.

Study screening and characteristics
A total of 7550 documents were found through searching databases, and 300 were retrieved through other review articles. After removing the duplicates, the titles and abstracts of 5966 documents were screened, of which 5683 documents were excluded because of not meeting the study criteria. If the inclusion criteria were not clear from the abstract, the full text was assessed. In this stage, 221 documents were excluded for different reasons. Finally, data were extracted and subjected to a metaanalysis from the remaining 62 documents (60 cross-sectional studies and 2 case-control studies) ( Fig. 1) [15,21,.
The prevalence of HCV Ab was assessed in 27,033 PWID in 62 studies. The smallest and largest sample size was 34 and 3,284 PWID, respectively. Twenty-three studies were conducted in DICs, and 16 were done in hospitals and healthcare centers. Table 1 presents the characteristics of the studies. There is a differential distribution of the studies across time by the study recruitment setting (Appendix 3).

Risk of bias in studies
The risk of bias and the quality of the study were assessed for the 62 included studies. Most of the studies did not have optimal condition. The highest risk of bias was related to the recruitment setting and sampling method that was seen in most of the studies (Appendix 4).

Cumulative meta-analyses
A cumulative meta-analysis was conducted to reflect the dynamic trend of results and evaluate the influence of individual study on the overall results. Figure 5 shows a forest plot for the cumulative meta-analysis for the trend of HCV Ab prevalence in PWID. A high prevalence (53%) of HCV Ab was first observed in 2000 and remained unchanged or had paradoxical changes after 24 more studies published between 2000 to 2005, and thereafter, the prevalence of HCV Ab between 2005 to 2018 slowly declined (Fig. 5). The cumulative meta-analysis was also presented by the study setting in Appendix 7. The results showed that with an increasing number of studies in later years the prevalence of HCV Ab is somewhat reduced or unchanged in all settings. In the prison setting, the trend of HCV Ab prevalence was increasing by 2002, and then   The trend of HCV Ab prevalence among PWID was presented by the recruitment setting in Appendix 9. The trend of HCV Ab prevalence among PWID in the prison setting did not decrease significantly (Cochran-Armitage test for linear trend = 1.85, P value = 0.17). However, trend analysis results indicated a statistically significant decrease in HCV Ab prevalence among PWID in the hospitals and healthcare centers (Cochran-Armitage test for linear trend = 38.72, P value < 0.001) and DIC settings (Cochran-Armitage test for linear trend = 5.93, P value = 0.01) (Appendix 9).

Publication bias
When we plotted the prevalence estimates against their standard errors, there was no publication bias (Fig. 7). Furthermore, the results were confirmed with Begg and Mazumdar's tau 0.22 (P value = 0.83) and Egger's regression intercept 1.43 (P value = 0.15). Trim-and-fill method for calibration of publication bias was performed. However, the missing study was not identified by the trimand-fill method.

Meta-regression analyses
The results of the univariate meta-regression analyses for investigation into the sources of heterogeneity showed that only the recruitment setting had a significant effect on HCV Ab prevalence: date of study (b = − 0.01, 95% CI − 0.02, 0.001, P value = 0.08), study type as crosssectional versus case-control study (b = − 0.13, 95% CI . Univariate meta-regression showed that the 8.81% of between-study variance is explained by date of study, 0.22% by study design, 0.0% by sampling method, and 21.88% by recruitment setting.

Discussion
PWID is the main population affected by HCV infection worldwide [102]. The present systematic review and meta-analysis was conducted on 62 observational studies, involving 27,033 PWID. This study examined the prevalence of HCV Ab before and after implementation of the harm reduction program in PWID in Iran. To our knowledge, this is the first systematic review and metaanalysis to examine the prevalence of HCV Ab among PWID before and after implementation of the harm reduction program. In the period since the last major systematic reviews [22,38], there has been an increase in the amount of evidence documenting drug injection and the prevalence of HCV Ab in PWID in Iran. This updated estimation showed that approximately one in two PWID in Iran are tested positive for HCV Ab, with substantial regional-and provincial-level variations in the prevalence of this blood-borne infection. The results showed a high prevalence of HCV Ab among PWID in Iran, which is consistent with a worldwide estimate in this population [103] and other systematic reviews [22,38], and still lower than the prevalence reported in certain other countries (e.g., Bulgaria, Estonia, and South Africa) [104,105] but higher than in studies conducted in Kuwait (12.3%) [106], Kingdom of Saudi Arabia (42.7%) [107], and Brazil (35.6%) [108]. These differences in the prevalence can be attributed to differences in the health systems, screening methods, and the type of high-risk behaviors of individuals [109,110]. In particular, in the developing countries, the harm reduction programs such as syringe distribution are not fully implemented [111]. Moreover, in the countries with high coverage of HCV care, cases are diagnosed and treated earlier and the transmission of HCV is prevented in the community, which can also be the reason for the inconsistent findings among different studies [112,113]. The pooled prevalence of HCV Ab was 49% before 2005 and 45% during and after 2005. Evaluation of the trend showed a slightly decreasing trend in the prevalence of HCV Ab in PWID over time. Besides, with the evaluation of the cumulative meta-analysis of HCV Ab prevalence, it seems there was a slightly declined or stable prevalence of HCV Ab among PWID in Iran. The study of Nematollahi et al. [22] showed a decreasing trend in the prevalence of hepatitis C among high-risk groups (β = − 0.021) in 2000-2015. These studies confirmed the findings of the current systematic review, which showed a slight decline in the prevalence of HCV Ab among PWID since 2005.
Globally, studies have shown that implementing NSP and maintenance treatment with methadone reduces high-risk behaviors [30][31][32]. According to the actions associated with harm reduction in Iran after 2005, this study indicated that the prevalence of HCV Ab among Fig. 2 Pooled HCV Ab prevalence in PWID in Iran PWID in Iran did not increase. Although the difference in the prevalence before 2005 and after 2005 was not significant, there seemed to be lower prevalence estimates in later years. This study examined the prevalence of HCV Ab which is defined as the number of affected persons with HCV Ab in the population at a specific time [114]. Moreover, when there is a gradual decreasing or stable prevalence of HCV Ab, it is partly attributable to a decrease in the incidence of HCV infection. Thus, the stable or slightly declining prevalence of HCV Ab after 2005 which indicates that new cases of the disease are less reported after 2005 can be partly caused by implementation of the harm reduction program. However, it should be noted that the decrease or stability in the prevalence of HCV Ab may also be influenced by other factors such as the high mortality of PWID with HCV infection, exclusion from being PWID by stopping drug injection and study design differences. Furthermore, another issue that may justify the stable or decreasing prevalence of HCV Ab is that over the past two decades in Iran, drug use patterns changed from traditional drug use, mainly heroin and opium to the recreational use of amphetamine-type stimulants, especially crystal methamphetamine [115,116]. Several Iranian studies showed that the prevalence of blood-borne infections (including HCV) among people with methamphetamine use is less than those with traditional drug use [39,[117][118][119]. This is because drug injection or syringe sharing is more in traditional drug users than methamphetamine users [17,119]. More, the prevalence of HCV Ab is different by the recruitment setting and there is a differential distribution of studies across time by the recruitment setting (e.g., more studies in DICs in recent years) causing a declining trend in the pooled results for HCV Ab prevalence. Therefore, it is needed to be confirmed by a cohort study thoroughly the impact of the harm reduction program on the incidence of HCV infection among PWID. To the best of our knowledge, no study in Iran measured the incidence of HCV infection among PWID. Van Den Berg et al. [34] confirmed in the Amsterdam Addiction Cohort study that the incidence of HCV infection among PWID who received both OST and high coverage NSP was approximately one-third lower than that of those who received either OST or NSP alone. In addition, Sharifi et al. [120] indicated that HIV infection incidence among PWID in 2014 was 5.39 (95% CI 4.71, 6.16) per This study suggested that after an increase in the 2000s, the HIV infection incidence decreased and stabilized among PWID and prisoners in Iran. This could be explained by expanding the preventive interventions, e.g., an increasing number of harm reduction centers for PWID and scaling up free harm reduction services such as increased coverage of the NSP among PWID. The above-mentioned studies suggested that a harm reduction intervention is needed to reach PWID early on to be effective in reducing the risk of HCV transmission. This is especially applicable to Iran if the overall coverage of the NSP and OST programs has reached sufficient among PWID. Other studies have confirmed the effectiveness of the harm reduction program as well [121][122][123][124]. In addition, the study by Rahimi et al. [125] reported that the prevalence of HIV was 14.3% before 2007 and 9.7% after 2007. These studies are approving the findings of the current systematic review, which showed sufficient evidence on the effectiveness of harm reduction program in the containment of HCV epidemics among PWID in Iran.
Some limitations of our study were differences in the studies carried out in different provinces and insufficient data from several provinces. The prevalence of HCV Ab in PWID had a wide range in different provinces (11-89%). The regional differences, at least somewhat, might be as a result of variations in time of the study and recruitment settings. It might also be due to differences in socio-demographic characteristics, access conditions for study participants, and high-risk behaviors of PWID in provinces of Iran. In addition, the number of studies conducted in each province is different. In a study [126] conducted in China, the prevalence of HCV Ab was different in various geographical regions. It has been stated that this finding may be due to the differences in the pattern of high-risk behaviors, poverty, and ethnicity in different provinces. These findings indicated that the burden of HCV is still high in some areas, and the scale-up Finally, the quality of tests used to determine HCV varied in the included studies and the proportions of the participants receiving the OST and NSP may influence the results of this review; however, we were unable to conduct additional subgroup analysis by the quality of tests and coverage of OST and NSP in this meta-analysis because of the lack of data. Despite the above-mentioned limitations, the results of this meta-analysis are valuable due to its large sample size, offering evidence to support the hypothesis that after the extensive implementation of harm reduction programs in Iran, the HCV epidemic among PWID has been controlled. We also performed a cumulative meta-analysis by the study date to investigate a trend of the HCV Ab prevalence, and the result indicated that our conclusion was robust when more new studies were added.

Conclusions
There is a large burden of HCV infection among PWID in Iran. The prevalence of HCV Ab in PWID decreased after 2005, and although not significant, there seemed to be lower prevalence estimates in later years. There are great variations in the prevalence of HCV Ab between different provinces. However, there is no sufficient

Method section: Search strategy Appendix 2
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

Method: Figure 1
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

Method section: Study selection, par 1
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

Method section:
Eligibility criteria, par 2

Risk of bias in individual studies
12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

Method section:
Data extraction, quality assessment and risk of bias, par 1 Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means).

Method section: Data analysis, par 1
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I 2 ) for each meta-analysis.

Method section: Data analysis, par 2
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).

Method section: Data analysis, par 1
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.

Study selection
17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.

Results: paragraph 1
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.

Results: paragraph 2
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).

Results: Risk of bias in studies, par 1
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency.
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15).

Results: publication bias and Figure 9 Appendix
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]).

Summary of evidence
24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

Discussion: paragraph 1, 2 and 3
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).

Discussion: paragraph 4 and 5.
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research.

FUNDING
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review.  Trend of HCV Ab prevalence among PWID in the hospital/healthcare center setting, the Cochran-Armitage test, P value < 0.001. c Trend of HCV Ab prevalence among PWID in the DIC setting, the Cochran-Armitage test, P value = 0.01