The world prevalence, associated risk factors and mortality of hepatitis C virus infection in hemodialysis patients: a meta-analysis

Abstract

Background

The worldwide burden of HCV infection among hemodialysis patients has not been systematically examined.

Methods

A systematic literature search was conducted in MEDLINE and Scopus to determine the worldwide prevalence of HCV infection, risk factors, and clinical outcomes among hemodialysis patients. Random-effect models and meta-regressions were used to generate pooled estimates and assess heterogeneity.

Results

Four hundred and seven studies with 1,302,167 participants were analyzed. The pooled prevalence of HCV infection was 21%. The highest prevalence was observed in Africa (28%) and low-income countries (48.5%). A significant prevalence decline was observed following the publication year and was also inversely related to GDP and total population of each country. Factors associated with HCV positivity included younger age, longer dialysis duration, more blood transfusions, and dialyzer reuse. The pooled unadjusted hazard ratio for all-cause mortality was 1.12 (95% CI 1.03–1.22), and the adjusted hazard ratio was 1.21 (95% CI 1.12–1.30) in HCV-infected compared to non-HCV infected patients.

Conclusions

HCV infection among hemodialysis patients is a worldwide shared burden and is associated with a higher risk of death. Avoiding unnecessary blood transfusion and dialyzer reuse should be encouraged to prevent HCV transmission in hemodialysis units.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40620-022-01483-x.

Introduction

Hemodialysis is the most prevalent modality of renal replacement therapy for patients with end-stage kidney disease worldwide [1]. Since the early 1990s when diagnostic testing for HCV became accessible, there was rising concern over the high prevalence of hepatitis C virus (HCV) within dialysis units [2]. The Dialysis Outcomes and Practice Patterns Study (DOPPS) demonstrated a mean hemodialysis center prevalence of HCV infection of 13.5% (range 2.6–22.9%) on data collection in France, Germany, Italy, Japan, Spain, the United Kingdom, and the United States between 1998 and 2001 [3] and a 9.5% seroprevalence among hemodialysis patients in 12 nations enrolled from 1996 to 2011 [4]. The DOPPS data also illustrated that HCV infection in hemodialysis patients was linked to higher mortality and hospitalization, as well as lower quality of life scores [5].

HCV is a hepatotropic RNA virus of the family Flaviviridae. The severity of the illness ranges from minor to severe, including liver cirrhosis and hepatocellular carcinoma (HCC) [6]. Globally, approximately 115 million people (1.6%) have chronic HCV infection [7]. HCV is primarily transmitted through percutaneous blood exposure as a result of sharing contaminated devices for drug injection or medical procedures, blood transfusion before the start of blood screening in the 1990s, and long-term hemodialysis. The age-adjusted HCV-related mortality rate increased each year from 2010 through 2013 but began to decline in 2014 after the emergence of direct-acting antiviral (DAA) treatments, which achieve exceptionally high rates of sustained virological response (SVR) [8]. However, access to diagnosis and treatment of HCV infection is limited in low-income countries, and there is currently no effective vaccine against HCV. On World Hepatitis Day 2021, the theme “Hepatitis can’t wait” underlined the importance of the urgency of eradicating hepatitis in order to meet the 2030 elimination targets [9].

Risk factors for HCV transmission among hemodialysis patients are blood transfusion, contamination of dialysis systems, inadequate disinfection and cleaning of environmental surfaces, improper contact of health care personnel with equipment and patients, and mishandling of parenteral medications [10, 11]. Infection control is the primary factor for preventing HCV transmission in dialysis units. Since 2001, the Centers for Disease Control and Prevention (CDC) has issued recommendations to minimize the spread of HCV infections among chronic hemodialysis patients [12]. Current guidelines for preventing HCV transmission in the hemodialysis unit (KDIGO 2018) recommend taking strong actions to improve hygiene in dialysis units, and include injection safety and environment cleaning whenever there is a new case of HCV infection [13]. The reuse of HCV-infected patients’ dialyzers is accepted if standard infection control protocols are followed.

To date, the worldwide prevalence of HCV infection among hemodialysis patients has not been systematically examined. This meta-analysis was conducted to estimate the prevalence of HCV infection in hemodialysis populations around the globe and to demonstrate geographic variations according to regions and their economies. The risk factors for HCV transmission and clinical outcomes, including mortality and hospitalization, were evaluated. This study aims to raise awareness of HCV infection among hemodialysis patients and provide a resource for developing public health policies.

Materials and methods

Data sources and searches

This meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table S1) [14]. The electronic searches were conducted in MEDLINE and Scopus databases (through March 2021) to identify eligible studies. For the search, the following Medical Subject Headings (MeSH) search terms were used: prevalence AND hepatitis C AND hemodialysis. Reference lists and all prior systematic reviews and meta-analyses were searched manually to identify additional eligible studies. The protocol was registered in PROSPERO (International Prospective Register of Systematic Reviews; ID CRD42022319084). The search was limited to publications in English.

Eligibility criteria

Studies evaluating patients (age ≥ 18 years) with end-stage renal disease undergoing long-term hemodialysis were included. Cohort, case–control, and cross-sectional studies were considered eligible for inclusion in the analysis. The studies had to report on HCV prevalence with crude number of total patients tested for HCV and number of positive results. Patient outcomes included risk factors for HCV infection and death; cause of death was also explored. We excluded abstracts or interim reports. Duplicate articles were identified and eliminated. Patients undergoing peritoneal dialysis, kidney transplantation, and palliative care (defined as patients who refused dialysis) were excluded.

Study selection

Two authors (PG and TL) independently screened the titles and abstracts of all electronic citations, and full-text articles were retrieved for a comprehensive review and independently rescreened. Each study was independently assessed by two authors (PG and TL). Disagreements were resolved through adjudication by a third author (TT).

Data extraction

The following data were independently extracted from the included studies by pairs of authors (PG and TL): first author’s name, year of publication, country of study, study design, type and technique of HCV testing, sample size, prevalence of patients with HCV infection, HCV genotype, age, hemodialysis duration, risk factors including tattoo, dialyzer reuse, blood transfusion, and laboratory investigations including liver function and serum albumin. Clinical outcomes were also extracted to evaluate pooled unadjusted hazard ratios, including hospitalization, all-cause mortality, infection-related mortality, malignancy-related mortality, and cardiovascular mortality (Supplemental Table S2). Studies demonstrating adjusted hazard ratios were also recorded separately. In addition to grouping the countries of the study by geographic zones, we also grouped countries by gross national income per capita derived from the World Bank’s classification of income of economies into low-, middle-, and high-income countries to assess the impact of a country’s economic status on the prevalence of HCV infection among hemodialysis patients. Low-income economies are defined as those with a gross domestic product (GDP) (representing the value of all final goods and services produced within a nation in a given year) of $1085 or less in 2021; $1086 and $4255 for lower middle-income; $4256 and $13,205 for upper middle-income; $13,205 or more for high-income [15]. The countries were also classified according to GDP [16].

Quality assessment

Pairs of authors (PG and TL) independently assessed study quality with disagreements resolved by discussion and consensus. The quality of the included literature was evaluated using the Newcastle–Ottawa Scale (NOS). NOS scores of 0–3, 4–6, and 7–9 were classified as poor, fair, and good quality for cohort and case–control studies, respectively. Study quality scores for cross-sectional prevalence studies were defined as unsatisfactory (0–4), satisfactory (5–6), good (7–8), and very good quality (9–10) [17, 18].

Statistical analysis

For the studies available for inclusion in the meta-analysis, random-effects models were conducted to generate pooled prevalence rates, risk factors, laboratory values, and clinical outcomes of hepatitis C infection in hemodialysis patients. Pooled unadjusted hazard ratios were calculated for all clinical outcomes and reported separately from pool adjusted hazard ratios derived from available studies. All pooled estimates are provided with 95% confidence intervals (95% CIs). Heterogeneity was assessed using the I² index and the Q test p value with a significant I² index of ≥ 75%, indicating medium-to-high heterogeneity [19]. Subgroup and meta-regression analyses were also conducted to examine global patterns of HCV prevalence, risk factors, and mortality by geometric zones, country economic status, GDP, year of publication, and total population of each country. Publication bias was formally assessed using Egger’s test [20]. The meta-analysis and meta-regression analysis were performed by using Comprehensive Meta-Analysis (version 2.0; Biostat, www.meta-analysis.com) and the “metafor” package in the R system software (version 2.14.0), respectively [21].

Results

Study characteristics

A total of 3220 potentially relevant citations were identified and screened. Four hundred and eighteen articles were retrieved for detailed evaluation, of which 407 studies from more than 70 countries worldwide with 1,285,389 participants fulfilled eligibility criteria and were included in the meta-analysis (Fig. 1). The studies spanned 32 years, from 1989 to 2021. There were 36 prospective cohorts, 10 retrospective cohorts, 3 case–control studies, and 358 cross-sectional studies. The characteristics of all included studies and quality assessment using NOS are provided in Supplemental Table S3. Among cohorts and case–control studies, 47 studies (96%) were considered of good quality, while the remaining 2 studies (4%) were of fair quality. For cross-sectional studies, there were 89 studies (24.86%) considered of very good quality, 232 studies (64.8%) of good quality, 35 studies (9.78%) of satisfactory quality, and only 2 studies (0.56%) of unsatisfactory quality.

Fig. 1.

Flow diagram of the study selection

Asia accounted for the most significant proportion of reports (47.17%, 191 reports), followed by Europe (27.52%, 112 reports), South America (10.07%, 41 reports), North America (6.39%, 26 reports), Africa (6.39%, 27 reports), Australia & New Zealand (1.47%, 9 reports), and multicontinental studies (0.98%, 1 report). Most studies (51.35%) originated from high-income countries. Three hundred and fifty-eight articles used anti-HCV antibody tests for the diagnosis of HCV infection, 32 articles used HCV RNA detection primarily by reverse transcription-polymerase chain reaction (RT-PCR), 14 articles used anti-HCV antibody tests or HCV RNA detection, 1 article used branched-DNA signal amplification assay, and 1 article used HCV antigen assay. The diagnostic test was not mentioned in 1 article.

Prevalence of HCV infection and HCV genotypes in hemodialysis patients

Based on all 407 studies, the pooled prevalence of HCV infection was 20.7% (95% CI 18.9–22.6; I² = 99.63%) among the hemodialysis population (Table 1). HCV genotypes were reported in 47 articles (Table 2). The most frequent HCV genotype in hemodialysis patients worldwide was genotype 1b (33.5%), followed by genotype 1a (22.8%), 3 (8.2%), 2 (6%), 4 (5%), and 6 (2.4%).

Table 1.

Pooled prevalence of HCV infection among hemodialysis patients

Subgroup	Studies (n)	Patients with HCV (n)	Patients (n)	HCV prevalence rate (%) (95% CI)	Test for heterogeneity		Assessment of publication bias
					I2 index	Q test, p value	Egger's test, p value
All	407	125,972	1,302,167	20.7 (18.9–22.6)	99.63	< 0.001	< 0.001
World zone
Africa	26	3249	9999	28.0 (21.4–35.7)	97.92	< 0.001
Asia	192	24,576	112,933	22.3 (19.8–25.0)	98.75	< 0.001
Australia & New Zealand	6	281	15,497	3.6 (1.7–7.4)	90.28	< 0.001
Europe	112	18,365	172,621	20.1 (16.7–23.9)	99.38	< 0.001
North America	26	51,320	748,570	16.5 (12.4–21.7)	99.87	< 0.001
South America	41	4389	25,032	19.4 (15.9–23.6)	97.72	< 0.001
Multiple continents	4	23,792	217,515	18.46 (6.33–43.11)	99.98	0.001
Diagnostic test
Antigen	1	11	119	9.2 (5.2–15.9)	0	1
Antibody	358	101,107	902,616	21.0 (19.3, 22.8)	99.47	< 0.001
Antibody or RNA	14	9281	20,970	28.1 (13.0–50.5)	99.79	< 0.001
DNA	1	59	394	15.0 (11.8–18.8)	0	< 0.001
RNA	32	3347	86,405	17.2 (10.6–26.7)	99.48	< 0.001
Not mentioned	1	12,167	291,663	4.17 (4.10–4.24)	0	1
Country income classificationa
High income	209	96,890	1,189,337	19.2 (17.2–21.4)	99.68	< 0.001
Middle income	184	21,157	101,687	21.2 (19.0–23.7)	98.37	< 0.001
Low income	8	803	1714	48.5 (32.0–65.4)	97.31	< 0.001
Not available	6	7122	9429	50.6 (30.9–70.2)	98.45	< 0.001

^aAccording to the World Bank’s classification of income of countries [15]

Table 2.

Pooled prevalence of HCV genotypes among hemodialysis patients

Subgroup	Studies (n)	Patients with HCV (n)	Patients (n)	Genotype prevalence rate (%) (95% CI)	Test for heterogeneity
					I2 index	Q Test, p value
All
Genotype 1a	47	614	2411	22.8 (17.2–29.6)	87.34	< 0.001
Genotype 1b	47	757	2411	33.5 (26.5–41.3)	88.35	< 0.001
Genotype 1 (other)	46	228	2384	4.2 (2.0–8.7)	89.59	< 0.001
Genotype 2	47	138	2411	6.0 (4.1–8.9)	75.23	< 0.001
Genotype 3	46	257	2384	8.2 (5.3–12.5)	86.35	< 0.001
Genotype 4	44	192	2317	5.0 (3.1–8.0)	82.29	< 0.001
Genotype 6	20	107	967	2.4 (0.8–6.6)	82.16	< 0.001
Mixed genotype	20	74	967	5.2 (2.7–9.6)	73.47	< 0.001
World zone
Africa
Genotype 1a	1	1	27	3.7 (0.5–22.1)	0	1
Genotype 1b	1	5	27	18.5 (7.9–37.5)	0	1
Genotype 2	1	1	27	3.7 (0.5–22.1)	0	1
Genotype 4	1	20	27	74.1 (54.7–87.1)	0	1
Asia
Genotype 1a	20	221	967	22.8 (15.4–32.3)	81.26	< 0.001
Genotype 1b	20	331	967	36.5 (25.1–49.7)	88.44	< 0.001
Genotype 1 (other)	20	70	967	3.8 (1.0–13.3)	87.73	< 0.001
Genotype 2	20	28	967	4.5 (2.8–7.2)	31.88	0.086
Genotype 3	20	110	967	7.5 (3.4–15.5)	87.18	< 0.001
Genotype 4	20	26	967	3.2 (1.4–7.1)	72.67	< 0.001
Genotype 6	20	107	967	2.4 (0.8–6.6)	82.16	< 0.001
Mixed genotype	20	74	967	5.2 (2.7–9.6)	73.47	< 0.001
Europe
Genotype 1a	15	273	1045	15.0 (7.7–27.0)	92.36	< 0.001
Genotype 1b	15	350	1045	38.9 (25.9–53.7)	91.98	< 0.001
Genotype 1 (other)	15	51	1045	2.7 (0.7–9.8)	88.12	< 0.001
Genotype 2	15	85	1045	9.3 (4.3–18.9)	88.36	< 0.001
Genotype 3	15	105	1045	6.2 (2.5–14.3)	90	< 0.001
Genotype 4	15	144	1045	7.7 (4.0–14.2)	85.33	< 0.001
North America
Genotype 1a	3	40	94	42.5 (32.8–52.8)	0	0.371
Genotype 1b	3	39	94	41.8 (29.1–55.8)	33.86	0.22
Genotype 1 (other)	3	3	94	3.8 (1.3–10.4)	0	0.857
Genotype 2	3	7	94	7.9 (3.9–15.3)	0	0.991
Genotype 3	3	5	94	7.9 (3.6–16.6)	0	0.392
South America
Genotype 1a	8	79	278	34.0 (14.4–61.2)	86.652	< 0.001
Genotype 1b	8	32	278	15.7 (7.6–29.8)	68.806	0.002
Genotype 1 (other)	8	104	278	11.1 (2.1–42.8)	89.949	< 0.001
Genotype 2	8	17	278	6.9 (3.2–14.3)	50.09	0.051
Genotype 3	8	37	278	14.6 (8.1–24.9)	64.51	0.006
Genotype 4	8	2	278	2.9 (1.3–6.5)	0	0.709
Country income classificationa
High income
Genotype 1a	21	222	994	21.4 (14.4–30.6)	81.37	< 0.001
Genotype 1b	21	336	994	35.4 (24.5–48.1)	88.17	< 0.001
Genotype 1 (other)	20	70	967	3.8 (1.0–13.3)	87.73	< 0.001
Genotype 2	21	29	994	4.6 (2.9–7.2)	29.01	0.105
Genotype 3	20	110	967	7.5 (3.4–15.5)	87.18	< 0.001
Genotype 4	21	46	994	4.0 (1.6–9.6)	82.94	< 0.001
Middle income
Genotype 1a	26	392	1417	24.1 (15.8–34.9)	89.83	< 0.001
Genotype 1b	26	421	1417	31.6 (22.7–42.2)	88.89	< 0.001
Genotype 1 (other)	26	158	1417	4.5 (1.7–11.2)	90.58	< 0.001
Genotype 2	26	109	1417	8.2 (4.8–13.6)	82.26	< 0.001
Genotype 3	26	147	1417	8.4 (4.8–14.2)	85.39	< 0.001
Genotype 4	23	146	1323	5.6 (3.1–10.0)	82.21	< 0.001
Genotype 6	2	104	396	5.8 (0.0–88.7)	91.8	< 0.001
Mixed genotype	20	74	967	5.2 (2.7–9.6)	73.47	< 0.001
Low income
Genotype 6	8	3	298	2.9 (1.1–7.5)	24.03	0.238

^aAccording to the World Bank’s classification of income of countries [15]

Risk factors for HCV infection in hemodialysis patients

The mean age of hemodialysis patients with HCV infection was 52 years. Younger age was significantly associated with increasing HCV infection (mean difference − 1.41 years; 95% CI − 2.27 to − 0.56; p = 0.001; I² = 96.46%). Longer dialysis duration was a significant factor for HCV positivity (mean difference 32.87 months; 95% CI 29.44–36.30; p < 0.001; I² = 98.56%). Blood transfusion and dialyzer reuse increased the risk of HCV infection. The pooled odds ratio for HCV infection was 2.13 (95% CI 1.74–2.61; p < 0.001; I² = 75.91%) and 2.12 (95% CI 1.47–3.05; p < 0.001; I² = 0%) from blood transfusion and dialyzer reuse, respectively. Higher numbers of blood transfusions (about 4.37 units) were associated with HCV infection among hemodialysis patients (95% CI 3.15–5.59; p < 0.001; I² = 90.74%). Tattoo was not found to be a significant risk factor for HCV infection in patients undergoing hemodialysis (Table 3).

Table 3.

Risk factors for HCV infection in hemodialysis patients

Risk factor	Studies (n)	Patients (n)	Mean difference (95% CI)	p value	Test for heterogeneity		Assessment of publication bias
					I2 index	Q test, p value	Egger's test
Age (all)	102	927,553	− 1.413 (− 2.269, − 0.558)	0.001	96.46	< 0.001	< 0.001
Subgroup by world zone
Africa	9	1,110	0.382 (− 1.142, 1.905)	0.623	0	0.544
Asia	57	18,699	0.280 (− 1.790, 2.351)	0.791	93.17	< 0.001
Australia & New Zealand	2	76,749	− 4.301 (− 4.701, − 3.901)	< 0.001	0	0.91
Europe	20	7,051	− 2.927 (− 4.810, − 1.045)	0.002	77.28	< 0.001
North America	11	740,985	− 7.164 (− 7.916, − 6.412)	< 0.001	89.98	< 0.001
South America	2	510	− 1.933 (− 5.322, 1.455)	0.263	0	0.6
Multiple continents	1	82,449	− 4.100 (− 4.475, − 3.725)	< 0.001	0	1
Subgroup by country income classificationa
High income	57	918,813	− 1.852 (− 2.916, − 0.788)	0.001	97.75	< 0.001
Middle income	44	8,689	− 0.813 (− 2.052, 0.426)	0.198	56.53	< 0.001
Low income	1	51	9.00 (0.613, 17.387)	0.035	0	1
Hemodialysis duration (all)	103	183,655	32.87 (29.44, 36.30)	< 0.001	98.56	< 0.001	0.49
Subgroup by world zone
Africa	7	578	29.636 (13.911, 45.361)	< 0.001	91.14	< 0.001
Asia	60	16,759	28.286 (23.996, 32.575)	< 0.001	97.52	< 0.001
Australia & New Zealand	2	76,749	29.250 (5.776,52.724)	0.015	48.18	0.17
Europe	24	5,007	52.570 (35.642, 69.499)	< 0.001	98.63	< 0.001
North America	4	1,124	20.791 (4.942, 36.640)	0.010	84.66	< 0.001
South America	5	989	25.524 (16.528, 34.520)	< 0.001	75.51	0.003
Multiple continents	1	82,449	32.400 (31.692, 33.108)	< 0.001	0	1
Subgroup by country income classification a
High income	56	174,539	35.383 (30.92, 39.85)	< 0.001	98.72	< 0.001
Middle income	43	8,618	29.87 (24.98, 34.76)	< 0.001	96.98	< 0.001
Low income	2	97	5.67 (2.74, 8.61)	< 0.001	0	0.46
Not available	2	401	44.98 (13.52, 76.44)	0.005	93.94	< 0.001
Number of blood transfusions	38	6,695	4.369 (3.151, 5.586)	< 0.001	90.74	< 0.001	< 0.001
Subgroup by world zone
Africa	5	412	4.762 (0.244, 9.280)	0.039	70.64	0.009
Asia	21	3,588	3.628 (2.111, 5.145)	< 0.001	93.69	< 0.001
Europe	10	2,518	5.523 (3.397, 7.650)	< 0.001	31.03	0.16
North America	1	90	43.00 (20.536, 65.464)	< 0.001	0	1
South America	1	87	7.00 (3.986, 10.014)	< 0.001	0	1
Subgroup by income
High income	22	4,342	5.044 (3.308, 6.780)	< 0.001	92.36	< 0.001
Middle income	14	2,256	3.565 (1.709, 5.420)	< 0.001	87.03	< 0.001
Low income	2	97	4.039 (− 2.496, 10.575)	0.226	58.38	0.121
			Odd ratio (95% CI)
History of blood transfusion	64	18,489	2.128 (1.738, 2.605)	< 0.001	75.91	< 0.001	0.06
Subgroup by world zone
Africa	2	220	5.047 (0.789, 32.290)	0.087	79.74	0.026
Asia	38	11,327	1.974 (1.503, 2.593)	< 0.001	77.37	< 0.001
Australia & New Zealand	1	60	3.437 (0.180, 65.594)	0.412	0	1
Europe	15	3,743	2.647 (1.915, 3.659)	< 0.001	51.64	0.011
North America	2	563	1.074 (0.626, 1.843)	0.796	0	0.56
South America	6	2,576	1.853 (0.763, 4.499)	0.173	88.16	< 0.001
Subgroup by country income classificationa
High income	40	11,339	1.951 (1.556, 2.446)	< 0.001	71.52	< 0.001
Middle income	24	7,150	2.472 (1.614, 3.788)	< 0.001	81.48	< 0.001
Tattoo	6	1,928	1.520 (0.473, 4.887)	0.482	52.28	0.063	0.94
Dialyzer reuse	4	1,846	2.116 (1.467, 3.053)	< 0.001	0	0.498	0.38

^aAccording to the World Bank’s classification of income of countries [15]

Laboratory results

Hemodialysis patients with HCV infection had significantly higher levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), and total bilirubin (TB) than non-HCV patients in pooled analyses (Supplemental Table S4). There was significantly lower serum albumin in the HCV-positive group than in the HCV-negative group (Fig. 2).

Fig. 2.

Pooled prevalence of HCV infection among hemodialysis patients by world zones

Clinical outcomes

The pooled rate of all-cause hospitalization

According to 3 studies shown in Table 4, there was no significance in HCV-associated all-cause hospitalization among hemodialysis patients (unadjusted hazard ratio 1.006, 95% CI 0.857–1.180, p = 0.945; I² = 97.43%).

Table 4.

Pooled hazard ratio of HCV-associated mortality and hospitalization in hemodialysis patients

Clinical outcome	Studies (n)	Hazard ratio (95% CI)	p value	Test for heterogeneity		Assessment of publication bias
				I2 index	p value	Egger's test, p value
Unadjusted hazard ratio
All-cause hospitalization	3	1.006 (0.857, 1.180)	0.945	97.43	< 0.001	0.09
All-cause mortality	13	1.123 (1.033, 1.221)	0.006	81.673	< 0.001	0.06
Subgroup by world zone
Asia	2	1.296 (1.195, 1.405)	< 0.001
Australia & New Zealand	2	1.015 (0.953, 1.080)	0.645
Europe	6	1.205 (1.076, 1.348)	0.001
North America	3	1.016 (0.913, 1.130)	0.768
Infection-related mortality	8	1.160 (1.026, 1.313)	0.018	0	0.866	0.33
Subgroup by world zone
Asia	3	1.358 (1.030, 1.789)	0.03
Australia & New Zealand	2	1.078 (0.912, 1.273)	0.378
Europe	1	1.388 (0.340, 5.674)	0.648
North America	2	1.198 (0.933, 1.538)	0.156
Malignancy-related mortality	5	1.390 (1.019, 1.896)	0.038	0	0.628	0.69
Subgroup by world zone
Asia	3	1.573 (1.095, 2.259)	0.014
Australia & New Zealand	1	0.880 (0.454, 1.705)	0.705
Europe	1	1.735 (0.397, 7.581)	0.464
Cardiovascular mortality	8	0.998 (0.927–1.074)	0.952	4.241	0.397	0.91
Adjusted hazard ratio
All-cause mortality	7	1.207 (1.119, 1.302)	< 0.001	75.59	< 0.001	0.001
Subgroup by world zone
Asia	1	1.480 (1.132, 1.934)	0.004	0	1
Australia & New Zealand	2	1.156 (1.049, 1.274)	0.004	38.26	0.2
Europe	2	1.309 (1.173, 1.460)	< 0.001	0.14	0.32
North America	2	1.155 (1.011, 1.318)	0.033	83.35	0.014
Cardiovascular mortality	2	1.186 (0.983, 1.431)	0.075	60.24	0.113	Not available

The pooled unadjusted all-cause and cause-specific mortality

Among 13 studies, the pooled unadjusted hazard ratio for all-cause mortality in hemodialysis patients with HCV infection was 1.12 (95% CI 1.03–1.22, p = 0.006; I² = 81.67%) compared to the patients without HCV infection (Fig. 3). Death from infection and malignancy in HCV-infected patients was significantly higher than in non-HCV patients. The pooled unadjusted hazard ratios of infection-related mortality and malignancy-related mortality were 1.16 (95% CI 1.03–1.31, p = 0.018; I² = 0%) and 1.39 (95% CI 1.02–1.90 p = 0.038; I² = 0%), respectively. There was no significance in unadjusted HCV-associated cardiovascular mortality (Table 4).

Fig. 3.

Pooled unadjusted hazard ratio for all-cause mortality in hemodialysis patients with HCV infection

The pooled adjusted all-cause and cause-specific mortality

Meta-analysis of 7 studies showed that the pooled adjusted hazard ratio for all-cause mortality in HCV-positive patients undergoing hemodialysis was 1.207 (95% CI 1.12–1.30, p < 0.001; I² = 75.59%) compared with HCV-negative patients (Fig. 4). Two studies reported a trend toward a higher risk of adjusted cardiovascular mortality but were not statistically significant.

Fig. 4.

Pooled adjusted hazard ratio for all-cause mortality in hemodialysis patients with HCV infection

Investigations of heterogeneity

Anti-HCV antibody tests identified more hemodialysis patients with HCV infection than other tests, followed by HCV RNA detection (21% vs. 17.2%). When examined according to geographic regions of the world, the pooled prevalence of HCV infection appeared highest in Africa (28%), followed by Asia (22.3%), Europe (20.1%), South America (19.4%), North America (16.5%), and Australia & New Zealand (3.6%) as shown in Table 1 and Fig. 2. The highest prevalence was demonstrated in low-income countries (48.5%).

The distribution of HCV genotypes varied among geometric zones (Table 2). HCV genotype 1 was the most common in all regions except in Africa. Genotype 4 was the most common in Africa. In North America, the highest prevalence of genotype was 1a (42.5%), followed by 1b (41.8%), 2 (7.9%), and 3 (7.9%). Genotype 1a was also the most frequent in South America (34%), followed by 1b (15.7%), 3 (14.6%), 2 (6.9%), and 4 (2.9%). Genotype 1b was the most prevalent in Europe (38.9%), followed by 1a (15%), 2 (9.3%), 4 (7.7%), and 3 (6.2%). The highest occurrence of genotype in Asia was 1b (36.5%), followed by 1a (22.8%), 3 (7.5%), 2 (4.5%), 4 (3.2%), and 6 (2.4%). Genotype 4 was the most common in Africa (74.1%), followed by 1b (18.5%), 1a (3.7%), and 2 (3.7%). There was no report of HCV genotype in Australia & New Zealand. According to subgroup analysis by country economic status, HCV genotype 1b predominated in high and middle-income countries, with prevalence rates of 35.4% and 31.6%, respectively. HCV genotype 6 (2.9%) was found only in low-income countries.

HCV-infected patients undergoing hemodialysis were younger than non-HCV patients in North America, Australia & New Zealand, and Europe. In the subgroup analysis by countries’ economies, there were younger patients with HCV infection among high-income countries. In contrast, older patients had a higher risk of developing HCV infection in low-income countries. Longer hemodialysis duration was a significant risk factor for HCV infection in all regions and economic statuses. The mean difference in dialysis duration associated with HCV infection was lesser in countries with lower income. History of blood transfusion was correlated with an increased risk of HCV infection in Europe and Asia. In addition, more units of blood transfusions increased the risk in all regions (Table 3).

In the subgroup analysis by world zone, the pooled unadjusted hazard ratio for all-cause mortality from HCV infection was highest in Asia, followed by Europe. Nevertheless, the pooled adjusted hazard ratio for all-cause mortality in HCV-infected patients was significantly higher than in non-HCV patients in all regions. HCV-associated mortality from infection and malignancy were found to be statistically significant only in Asia (Table 4).

Meta-regression illustrated a significant decline in the prevalence of HCV infection among hemodialysis patients associated with the year of study publication (p < 0.001). The HCV prevalence was also inversely related to each country's total population (p < 0.001). Countries with higher GDP had a significant trend towards a lower prevalence of HCV-infected patients (p = 0.04).

Assessment of publication bias

Egger’s regression model was used to detect publication bias. The result presented significant publication bias in HCV prevalence. Evaluation of publication bias in other outcomes is shown in Tables 3 and 4.

Discussion

The objective of our meta-analysis was to estimate the global prevalence of HCV infection in hemodialysis populations and to demonstrate geographic variations according to regions and their economies. A total of 407 observational studies published since 1989, when HCV was first discovered, were evaluated, representing more than 1 million patients. To our knowledge, this is the largest meta-analysis that has systematically examined the global prevalence of HCV infection among the hemodialysis population. Included studies were diverse in terms of nation, participants, and findings. Mainly, the quality of the studies was good.

Our pooled prevalence of HCV infection in hemodialysis patients was 20.7% (Table 1) and was far higher than the previously reported global seroprevalence of HCV in the general population, which ranged from 1.3 to 2.1% [7]. Repeated exposure to the dialysis machine, frequent blood transfusions due to anemia, and drug injection in dialysis-dependent patients always carry a minor but consequential risk of HCV transmission [22]. It is also important to note that the prevalence of HCV infection in the studies included in our analysis was essentially varied, depending on investigating site and year. It ranges from 3.6% in Australia and New Zealand to the highest, 28% in Africa. When stratified by country’s income and GDP, the lower-income countries had greater prevalence than the higher group. These findings indicated that health policy, public hygiene, socioeconomic status, and cultural factors contribute to the variation in HCV epidemiology. Our results also showed the decline in the prevalence of HCV infection among hemodialysis patients in more recently published studies. The reduction in infection rates might reflect the improved implementation of standard infection control procedures since the CDC provided recommendations regarding HCV infection control practices for the dialysis unit in 2001 [12].

Several tests are used to diagnose HCV infection. The majority of studies (87.96%) in our meta-analysis used serological tests to evaluate the prevalence of HCV infection in the hemodialysis population (Table 1). Anti-HCV tests, however, are unable to differentiate between a resolved and an active HCV infection. [23]. Nucleic acid test (NAT) technologies are used to detect HCV viremia with detection limits of about 10–20 international units [IU]/ml [24]. HCV antigen tests that detect core antigen alone or in conjunction with additional HCV proteins are less expensive than NAT but have a less sensitive detection limit (150–3000 IU/ml) [25, 26]. According to our findings, the anti-HCV antibody test detects more HCV infections than the HCV RNA method (21% vs. 17.2%). The results in the present study were supported by previous HCV prevalence studies which showed that approximately 50–90% of seropositive patients had HCV-RNA detection [10, 27, 28]. This could result from a previously resolved infection, fluctuating viremia levels, or RNA levels below the diagnostic assay's detection limit. Umlauft et al. discovered that 35% of hemodialysis patients had a fluctuating pattern of HCV viremia with intervals of undetectable HCV RNA [29]. To screen patients before starting in-center hemodialysis, the KDIGO 2018 Guidelines [13] recommend using NAT alone or an immunoassay followed by NAT if the immunoassay is positive. Initial NAT testing should be considered in dialysis units with a high prevalence of HCV.

Globally, the most common HCV genotype among hemodialysis patients in our pooled prevalence is genotype 1b (33.5%), followed by genotype 1a (22.8%), 3 (8.2%), 2 (6%), 4 (5%), and 6 (2.4%). Significant variations across regions are shown in Table 2 and Fig. 2B, C. HCV genotype 1 is the most common in all regions except Africa, However, only 1 article in Africa investigated HCV genotype and reported that the most prevalent is genotype 4. In addition, there are no reports in Australia & New Zealand. Subtype 1a dominates in North America (42.5%) and South America (34%), whereas subtype 2b is the most frequent in Europe (38.9%) and Asia (36.5%). HCV genotype 6 is reported only in Asia (2.4%). In the subgroup analysis by country economic status, HCV genotype 1b is the most common in high- and middle-income countries. This HCV genotype distribution in the hemodialysis population is similar to that in the general population, as per a previous report [7].

On the basis of our meta-analysis, the risk factors associated with HCV infection in patients undergoing hemodialysis are younger age, longer dialysis duration, more blood transfusions, and dialyzer reuse. Nevertheless, patients on hemodialysis are also at the same risk of HCV infection as the general population as regards unprotected intercourse, unsafe tattooing, or interfamilial transmission. Our finding of dialyzer reuse as a risk of HCV infection highlights a virtual knowledge gap. Considering dialysis-related factors, according to the KDIGO 2018, individuals with HCV infection can reuse their dialyzers as long as regular infection control protocols are followed [13]. Indeed, the recommendation was based on two observational studies that did not show dialyzer reuse to be a risk factor for HCV transmission [30, 31]. However, these two studies were conducted in developed countries, while our meta-analysis included studies from countries with various socioeconomic statuses. According to the KDOQI 2020, it is suggested that the dialyzers of HCV-infected patients can be reused when necessary [32]. Therefore, the suggestion of dialyzer reuse in HCV-infected patients might depend on the country’s context.

HCC and cirrhosis were the leading causes of death in patients with HCV infection. The risk of HCC in HCV-infected patients is increased by 15- to 20-fold and the incidence of HCV-related HCC remains high despite the advent of effective treatment [33]. However, the incidence of HCV-related HCC and cirrhosis were underreported. Studies by Faustini et al. and Nakayama et al. reported a higher incidence of HCC and cirrhosis in dialysis-dependent patients with HCV infection, but both studies were single-center and small [34, 35].

The pooled all-cause mortality, infection-related mortality, and malignancy-related mortality are higher in HCV-positive patients than in HCV-negative patients undergoing hemodialysis (Table 4). A previous study also observed these relationships in a non-dialysis population [36]. A meta-analysis by Petta et al. reported a significant pooled effect of HCV infection on cardiovascular mortality from two observational studies in a normal population (Odds ratio 1.65; 95% CI 1.07–2.56; p = 0.02) [37–39]. According to a recent meta-analysis by Fabrizi et al., the adverse impact of HCV on cardiovascular mortality was also demonstrated in the hemodialysis group [40]. However, our investigation, which included all data from a prior meta-analysis and more studies published after that, found no link between HCV infection and cardiovascular mortality (Table 4). Since patients on long-term hemodialysis have a high number of cardiovascular risk factors, the impact of HCV infection on cardiovascular mortality may be subtle.

In subgroup analysis, the Asian population had a higher hazard ratio of all-cause mortality from HCV infection and was the only population exhibiting significant infection-related mortality and malignancy-related mortality (Table 4). These findings require further investigation on the access to HCV treatment in Asia. Measures for preventing HCV transmission in the hemodialysis unit in Asia might focus on reducing blood transfusions by introducing erythropoiesis-stimulating agents and intensive HCV screening for blood donors since history and number of blood transfusions are identified as significant risk factors for HCV infection in Asia. Nevertheless, none of the studies provided the mortality rate in Africa, and South America, which needs further investigation.

Direct Acting Antiviral-based therapy is the current standard of care for HCV infection, which achieves high rates of SVR and is associated with a significant reduction in HCV mortality [41]. The introduction of DAAs since 2014 may also have contributed to a considerable decline in our pooled HCV prevalence among hemodialysis patients. However, there are still some issues with DAA treatments in hemodialysis patients. Because DAAs have variable renal elimination, kidney function is a key factor in treatment selection. HCV genotype is another aspect to be considered in treating patients with renal impairment. For all HCV genotypes, the combination of glecaprevir and pibrentasvir is the preferred regimen in severe renal impairment and in those requiring hemodialysis (Grade of evidence A, Grade of recommendation 1) [42]. Grazoprevir/elbasvir is also the regimen of choice for severe renal impairment and hemodialysis patients infected with HCV genotypes 1a, 1b, and 4 (Grade of evidence A, Grade of recommendation 1) [42]. Sofosbuvir is eliminated mainly by renal route (80%) and is only approved for use in patients with eGFR ≥ 30 ml/min/1.73 m² [43]. However, several studies [44–46] found that sofosbuvir was safe and effective in hemodialysis patients. Nowadays, sofosbuvir can be given without dose adjustment in hemodialysis patients when alternative treatment is unavailable [42, 47]. Hemodialysis patients with decompensated cirrhosis (Child–Pugh B or C) also should be treated with the fixed-dose combination of sofosbuvir and velpatasvir without ribavirin [42]. Despite pan-genotypic DAAs, identification of individual genotypes before beginning first-line therapy is still valuable for choosing the best treatment regimens and may be essential in some countries where drug procurement or price dictates genotype-specific treatment. More virological studies, particularly in low-income countries, are required to understand the epidemiology, distribution, and prevalence of HCV subtypes to optimize treatment decisions without the need for individual HCV genotype and subtype determination in the future.

The main strength of our meta-analysis is that it includes a large number of studies. Some limitations should be considered. First, several diagnostic tests were performed to identify the prevalence of HCV infection. Most studies used an anti-HCV antibody test; however, a positive result may indicate a current HCV infection (acute or chronic), a past resolved infection, or a false-positive result [23]. A false-negative result is also found in immunocompromised patients and those with possible HCV exposure during the prior 6 months [48]. Second, the sensitivity and specificity of the serological tests are different between the old and new generations. Third, some potential confounders regarding hemodialysis unit policy should be considered, such as room sharing, machine disinfection, and isolation procedures. These factors, and to what extent they could lower HCV transmission in dialysis units should be further evaluated. Fourth, we excluded non-English articles; therefore, some studies that published local data may not have been identified. In addition, only two databases including MEDLINE and Scopus were used but they should cover most leading journals. Fifth, there is a publication bias in HCV prevalence, presumably leading to a more considerable HCV prevalence in hemodialysis patients than actually exists. In addition, few articles studied HCV-related clinical outcomes and genotypes. There is a need for more studies in developing and lower-income countries.

Conclusion

HCV infection affects 1 in 5 patients undergoing hemodialysis and is associated with a higher risk of death. Our meta-analysis provides a platform to facilitate discussions among health care professionals, the public, and policymakers in order to raise awareness about HCV infection among hemodialysis patients and its associated health care burden. In addition, this study might carry implications for future development of public health policies in preventing HCV transmission in hemodialysis units.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HCV

Hepatitis C virus

GDP

Gross domestic product

DAA

Direct-acting antiviral

SVR

Sustained virological response

CDC

Centers for Disease Control and Prevention

NOS

Newcastle–Ottawa Scale

95%CI

95% Confidence intervals

AST

Aminotransferase

ALT

Alanine aminotransferase

ALP

Alkaline phosphatase

TB

Total bilirubin

Author contributions

Conceptualization: PG, TT, KT, JP, PS. Methodology: PG, TT, PS. Data curation: PG, TL, TT, PS. Formal analysis: PG, TL, TT, PS. Funding acquisition: PS. Investigation: PG, TL, TT, PS. Project administration: PG, TL, PS. Software: PS. Supervision: KT, JP, YA, SE-O, YA, PS. Validation: PG, TL, TT, PS. Visualization: PG, TL, TT, PS. Writing – original draft: PG, TL, TT, PS. Writing—review & editing: PG, TL, TT, YA, KP, SE-O, PS.

Declarations

Conflict of interest statement

The authors have no relevant financial or non-financial interests to disclose.

Ethical approval

This article does not contain any studies with human participants or animal performed by any of the authors.

Funding

Research Unit for Metabolic Bone Disease in CKD patients, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand.

Data availability statement

The datasets generated or analyzed during this study are included in this published article and supplementary table S3. Completed data extraction tables are available for interested reader upon request.