HCV infection is a global health problem that can elicit liver cirrhosis, hepatocellular carcinoma (HCC), and hepatic failure.¹ The World Health Organization reported that in 2019, approximately 290,000 people died from hepatitis C worldwide, mostly from cirrhosis and HCC.² HCV is primarily a blood-borne virus that is transmitted through unsafe therapeutic injections, injection drug use, blood transfusion, accidental needle stick injury, and sexual contact.³ Despite recent improvements in direct-acting antiviral agents for chronic hepatitis C, the prevention of HCV transmission via unsafe invasive medical practices, including acupuncture, remains inadequate.⁴

Acupuncture was first performed over 2,500 years ago in China and consists of the therapeutic insertion and manipulation of thin needles at more than 2,000 acupuncture points connected by yin and yang pathways.⁵ Over the centuries, acupuncture has extended to other continents as a novel treatment strategy for chronic diseases, including musculoskeletal pain and hypertension.⁶ In recent decades, medical oriental doctors, physiotherapists, and chiropractors worldwide are applying acupuncture needling approaches to treat musculoskeletal pain and other health problems.⁷

Although acupuncture is more widely applied than before, unsafe acupuncture practices in the community have increased the risk of HCV transmission.^8,9 However, previous meta-analyses showed insufficient data and controversial results.^10-13 Therefore, this study comprehensively analyzes and performs an updated meta-analysis to assess the association between acupuncture and the risk of HCV transmission.

1. Search strategy and eligibility criteria

We conducted a systematic review and meta-analysis based on the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement (Supplementary Table 1). From January 2000 to May 2022, two authors (MHH and YH) independently identified eligible articles from databases such as PUBMED, EMBASE, and the Cochrane Library, using the structured keywords “acupuncture,” “sham,” “piercing,” “tattoo,” “hepatitis C virus,” “HCV,” “chronic hepatitis C,” and “CHC.” For electronic scrutiny, all the study references and relevant review articles were manually searched. Disagreements were resolved through discussion and consensus referrals to a third investigator (JA). All the analyzed studies were published in English. Detailed search terms and strategies are listed in Supplementary Table 2.

The full text of potentially relevant publications of clinical studies was analyzed based on the following inclusion criteria: (1) non-randomized studies that compared the clinical effects of acupuncture and HCV transmission as either primary or secondary outcomes; (2) extractable RR or OR and their corresponding CI, or provided with enough data to compute these parameters; (3) participant age >18 years; and (4) study sample size >100. If the cases included in a study were published in different phases of duplication, the most recent publications or the largest cohort was selected. Studies excluded were non-human studies (animal or cell studies), non-peer-reviewed articles (meeting abstracts, case reports, editorials, clinical trial protocols, correspondence letters, or editorials), and studies that did not meet the inclusion criteria.

2. Data extraction and quality assessment

Data extracted included author names, publication year, country, ethnicity, enrollment period, study design, study population, sample size, HCV infection rate, adjusted OR or RR, and 95% CI. Two independent reviewers (MHH and YH) reviewed the selected studies. The study designs were categorized as a cross-sectional case-control study, a cross-sectional cohort study, and a cross-sectional study. Ethnicity was classified as either Asian or non-Asian. The risk of bias was assessed by adapting the quality assessment scale from the modified Newcastle–Ottawa Scale according to three categories: selection (range, 0–5 stars), comparability (range, 0–2 stars), and outcome assessment (range, 0–3 stars).¹⁴ More stars indicate a lower risk of bias. Each item was scored as ‘‘yes (star),’’ ‘‘no,’’ or ‘‘unclear,” and an agreement between the three authors (MHH, YH, and JA) was required.

3. Statistical analysis

A random-effects model was used to calculate the pooled OR and 95% CI for the association between acupuncture and HCV transmission risk due to the predicted high heterogeneity. The analysis of heterogeneity was assessed using the Higgins I² statistic, in which 50% or more was considered significant heterogeneity.¹⁵ Funnel plots and Egger’s test for HCV transmission risk were used to assess publication bias.¹⁶ A subgroup analysis was performed to investigate the source of heterogeneity according to the study design and ethnicity. Statistical significance was set at p<0.05. Statistical analysis was conducted using Review Manager 5.3 (The Cochrane Collaboration).

1. Article selection

Fig. 1 shows the detailed steps of the study selection. Overall, we identified 1,465 potentially relevant studies through database searches and additional records by manual search. After excluding articles subsequent to primary screening, 373 studies were retrieved for full-text review. Of these, 345 articles were excluded as they did not meet the inclusion criteria: 38 conference abstracts only, 245 with other transmission routes, 24 insufficient HCV data, 5 genetic studies, 6 case series, 12 review papers, and '15' no extractable data. Finally, 28 studies with 194,826 participants were included in the systematic review and meta-analysis.^17-44

Figure 1. Study flow chart.

2. Study characteristics and quality assessment

Among the 194,826 participants included in the 28 studies, 178,583 (91.7%) were control participants and 16,243 (8.3%) had undergone acupuncture. Most studies were cross-sectional case-control studies (n=14 [50.0%]),^{17,18,21-24,26,28,29,32,35,38-40} cross-sectional studies (n=12 [42.9%]),^{19,20,25,27,31,33-34,36,37,41,43-44} and cross-sectional cohort studies (n=2 [7.1%]).^30,42 Categorizing by ethnicity, 16 (57.1%) studies included Asian populations, whereas 12 studies included non-Asian populations. Most studies analyzed hospital samples (n=13 [46.4%]) and the general population (n=9 [32.2%]). The majority of the studies collected acupuncture data from questionnaires, except for 3 studies procured from medical records.^26,33,44 The baseline demographics of the included studies are presented in Table 1. The quality assessment results (Supplementary Table 3) adopted the Newcastle–Ottawa scale, which showed that most of the included studies scored more than three stars in the selection and outcome assessment sections. However, the comparability section showed diverse results between studies (0–2 stars).

Table 1

3. Association between acupuncture and HCV transmission

In the overall analysis, acupuncture users showed a higher HCV transmission rate (17.4% [2,941/16,843]) than control participants (4.6% [8,233/178,583]). The pooled analysis showed statistical significance for the HCV transmission rate in acupuncture users with high-grade heterogeneity (OR, 1.84; 95% CI, 1.46–2.32; p<0.001; I²=80%; Fig. 2). Three studies showed a lower rate of HCV infection in control than in acupuncture subjects.^18,33,43 However, there were no consistent findings between the studies. Among the studies analyzed, four studies reported 0 HCV infection in acupuncture patients due to the small study population.^27,33,37,43

Figure 2. Pooled analysis of odds ratio for the association between acupuncture and HCV infection. Each dot represents the OR with a 95% CI. The box size represents the weight of the study in the meta-analysis. Weights were obtained using the random-forest model. M-H, Mantel-Haenszel.

4. Subgroup analysis and publication bias

Fig. 3 shows the pooled analyses of ORs for the association between acupuncture and HCV transmission, according to the study design. Both cross-sectional case-control studies (OR, 1.96; 95% CI, 1.47–2.61; p<0.001; I²=88%) and cross-sectional studies (OR, 1.85; 95% CI, 1.32–2.61; p<0.001; I²=0%) showed significantly higher HCV transmission rates in acupuncture users than in controls. However, cross-sectional cohort studies did not show statistical significance between the two groups (OR, 1.08; 95% CI, 0.73–1.60; p=0.760; I²=0%). As presented in Fig. 4, both Asian and non-Asian populations showed significantly higher HCV transmission rates in acupuncture users than in controls (Asians: OR, 1.98; 95% CI, 1.43–2.76; p<0.001; I²=84%; non-Asians: OR, 1.60; 95% CI, 1.28–1.99; p<0.001; I²=29%). No significant asymmetry was detected in the funnel plots for the association between acupuncture and HCV infection (Egger’s p>0.05; Fig. 5).

Figure 3. Pooled analysis of odds ratio for the association between acupuncture and HCV infection according to study design. Each dot represents the OR with a 95% CI. The box size represents the weight of the study in the meta-analysis. Weights were obtained using the random-forest model. M-H, Mantel-Haenszel.

Figure 4. Pooled analysis of odds ratio for the association between acupuncture and HCV infection according to ethnicity. Each dot represents the OR with a 95% CI. The box size represents the weight of the study in the meta-analysis. Weights were obtained using the random-forest model. M-H, Mantel-Haenszel.

Figure 5. Funnel plot for the association between acupuncture and HCV infection. SE, standard error.

This study reports an updated systematic review and meta-analysis of the effects of acupuncture on HCV transmission. By analyzing 28 studies (194,826 participants), we found that acupuncture users showed significantly higher HCV transmission rates than controls (17.4% vs. 4.6%; OR, 1.84; 95% CI, 1.46–2.32, p<0.001). The risk of acupuncture and HCV infection was consistent with the study design (cross-sectional case-control and cross-sectional studies) and ethnicity (Asian and non-Asian populations).

Acupuncture remains popular worldwide and appears to be a relatively safe treatment for the management of lifestyle risk factors, especially for musculoskeletal diseases.⁴⁵ The major indications for acupuncture include biological effects on local inflammatory responses, pain control, arthritis, and other joint diseases, as well as a broad range of minor diseases (headaches) that do not respond to conventional treatments.^46,47 However, there is a need for safer acupuncture practices using disposable needles and close monitoring of acupuncturists worldwide.⁴¹ As the prevalence of blood transfusion is less than 1% in the general population, parenteral viral exposure (such as during acupuncture) should be a public health priority to prevent HCV transmission.⁴⁸

A previous meta-analysis regarding percutaneous needle injections, including tattoos, piercings, and acupuncture, showed conflicting results from our study.^10-13 Jafari et al.¹⁰ showed that tattoos were associated with a higher risk of HCV infection (83 studies, 132, 145 participants; pooled OR, 2.24; 95% CI, 2.01–2.50; p<0.001). In this study, non-injection drug users showed the strongest association between tattooing and HCV infection (OR, 5.74; 95% CI, 1.981–6.66).¹⁰ However, Tohme and Holmberg¹² reported no definitive evidence for an increased risk of HCV infection if tattoos and piercings were performed by professional parlors (adjusted OR, 0.8; 95% CI, 0.4–1.7), but a significant risk for tattoos performed in nonprofessional settings (adjusted OR, 3.5; 95% CI, 1.4–8.8). Van Remoortel et al.¹³ pooled and analyzed 21 studies and showed that percutaneous needle treatments (including tattooing, acupuncture, and piercing) increased the HCV infection in blood recipients (tattoo: OR, 5.28; 95% CI, 4.33–6.44; p<0.001; acupuncture: OR, 1.56; 95% CI, 1.17– 2.08; p=0.03; piercing: OR, 3.25; 95% CI, 1.68–6.30; p=0.005). The report by Lim et al.¹¹ analyzed 86 studies investigating tattoos and transfusion-transmitted diseases (HBV, HCV, and HIV) and found that the tattooed group showed a higher risk of HCV infection than the non-tattooed group (OR, 2.89; 95% CI, 2.48–3.37). Although previous meta-analyses have assessed tattooing, piercing, and HCV infection risk, the current study is the first and largest study to focus on acupuncture and HCV infection risk.

This study has several limitations. First, the included studies contained observational data from a cross-sectional design. Second, a substantial degree of heterogeneity may be present; therefore, caution should be taken when interpreting the results. Third, insufficient information on other confounding factors for HCV infection between the acupuncture and control groups may serve as potential limitations. Fourth, since a majority of the studies collected acupuncture data by using questionnaires (25/28; 89.3%) without reporting time intervals between acupuncture and HCV infection, the direct correlation between acupuncture and HCV infection is hard to establish. Lastly, meta-analysis of studies is controversial because uncontrolled confounders may affect pooled estimates. It could be argued that pooling analyses might be inappropriate due to possible heterogeneity which can affect the outcomes of interest.

In conclusion, the evidence from this meta-analysis shows that acupuncture potentially increases the HCV transmission rate. Unsafe medical procedures and social practices, including acupuncture, should be performed with caution. Further large-scale, high-quality studies are warranted in the future.

Supplementary material is available at the Korean Journal of Gastroenterology website (https://www.kjg.or.kr/).

This study was supported by grants from the Basic Science Research Program through the National Research Foundation of Korea (RS-2022-00166674).

None