The prevalence of Human Immunodeficiency Virus infection among voluntary blood donors in mainland China: A systematic review and meta‐analysis

Human immunodeficiency virus (HIV) infection through transfusion has been an imperative challenge for blood safety. Despite the implementation of screening strategies, there was still the residual risk of transfusion‐transmitted HIV. Considering that the prevalence of HIV infection in blood donors is significant for evaluating blood safety and potential risks to the population, meta‐analysis was applied to investigate the HIV prevalence among voluntary blood donors during the past 27 years to characterize the epidemiology and related risk factors of HIV in blood donors. The literature concerning the HIV screening reactive rate and prevalence in Chinese voluntary blood donors was collected through the systematic searching of four electronic databases. After integrating data, following the Preferred Reporting of Items for Systematic Reviews and Meta‐Analyses guidelines, data manipulation and statistical analyses were conducted by Stata 12.0. The results indicated that overall HIV prevalence was 0.0178% (95% confidence interval [CI], 0.0169%−0.0187%) with a remarkable rise, which varied from 2000 (0.0034%) to 2015 (0.027%). The HIV window period infection rate was 0.0475‱ (95% CI, 0.0304‱−0.0646‱). Importantly, subgroup analysis revealed the heterogeneity in gender, occupations, education and donation frequency. With the effective control of HIV transmission through blood, HIV prevalence declined in China to some extent in recent years, and the characteristics of HIV epidemic in some provinces have drastically changed. However, remaining relatively high HIV prevalence and overall increased trend of HIV prevalence since the 21th century demonstrates the potential residual risk of blood transfusion, and the whole society is supposed to pay close attention to HIV infection.


| INTRODUCTION
It is known to all that acquired immune deficiency syndrome (AIDS) is one of the most severe pandemics.It was estimated that approximately 38.4 million people were infected with human immunodeficiency virus (HIV) and 1.5 million annually were new infections. 1 In China, there were 60 000 people infected with HIV or diagnosed with AIDS in 2021, 19 000 patients died of AIDS. 2 Since Peking Union Medical College Hospital reported the first case of AIDS in China in 1985, The infection of HIV has spread all over the country. 3 1988, Notice on Rectifying Production Management of Blood Products promulgated by the Ministry of Health required testing of blood donors for HIV antibodies.After the implementation of the blood donation law in China in 1998, the government advocated voluntary blood donation and banned paid blood donations for clinical use, these Policies undoubtedly reduces the risk of HIV transmission through blood transfusion and enhance blood safety through legitimate access to blood collection and normalized blood management. 46][7][8] The routine blood screening included two rounds of parallel post-donation enzyme-linked immunosorbent assay (ELISA) from different manufacturer.If the test results in anti-HIV is reactive for one or two kits, the reexamination will be conducted.When the result is still reactive, the sample will be determined as screening positive, and then will be sent to the local Center for Disease Control and Prevention (CDC) for anti-HIV confirmative test using western blot (WB).
Besides the serological testing (ELISA), the nucleic acid testing (NAT) was introduced into the postdonation screening in 2010 in China.NAT could shorten the window period and reduce the transfusion residual risks.With all these efforts, transfusiontransmitted risk of HIV infection has been reduced compared to the past.However, there is a considerable amount of HIV infection cases in China, and HIV infection through transfusion remains an imperative challenge for blood safety.A wide range of studies aimed at understanding the epidemiology and related risk factors of HIV among voluntary blood donors in China has been published, but some defects were found in these studies, such as contradictory results from different studies and deficient representative samples.Additionally, due to insufficient sample size or invalid statistical methods, some studies were unqualified.Thus, it is essential to examine the epidemiology of HIV among the blood donors in a systematic approach.Meta-analysis is chosen as the most reasonable means to address this issue, because Meta-analysis is a statistical method which is used to combine findings from plentiful independent studies on HIV infection among blood voluntary donors in multiple regions.Although there have been some meta-analysis related to HIV, we paid more attention to the HIV prevalence of voluntary blood donors, so as to provide well-targeted guidance on HIV transmission in transfusion and professional advice based on current data.We excluded studies if they were overlapping or paradoxical (i.e., data from the literature were repeated or inconsistent between the context); they lacked the exact sample size, specific research time or location; the sample size was less than 50 individuals; their objects consisted of plasma donors for manufacturers; or if the unit of measure was a U (related to blood collection).We also excluded studies if their population was not entirely wholly blood voluntary donors.
To avoid that the same blood donors had participated in the different studies, we rigorously and carefully screened the retrieved literature to avoid overlapping counts during the process of literature selection and data extraction.We discriminated participants by city of study, year, the institution at which the work was performed, etc.
When the above information was found to be the same or similar, we would read the full text to identify whether there were the same participants in the different studies, and overlapping counts would be deleted.Moreover, Once the HIV-positive blood donors are identified in China, their information will be reported to the relevant authorities, and they will also be recorded in the blood donation center's information system.As a result, these donors will not be able to donate blood again, ensuring that there are no instances of the same HIV-positive individual donating blood multiple times.
In the Standard Operating Procedure of blood station, ELISApositive samples are sent to qualified confirmation laboratories for Western Blot verification.For NAT pool-positive samples that turn negative upon individual retesting, the samples are sent to confirmation laboratories for more sensitive viral load PCR verification.For NAT positive samples, apart from nested PCR in few collected literature, we carried out follow-up for confirmation in the great majority of studies.Through subsequent follow-up, the cases with ELISA-seronegative and NAT − positive were confirmed.In consideration of nonspecific positive reaction in NAT, NAT-positive donors were tested for anti-HIV ELISA and NAT 1 week, 2 weeks, 1 month, and then every 1 month after blood donation.
In our study, HIV infected donors should be confirmed by NAT or WB tests.Accordingly, it is necessary that the People living with HIV (PLWH) included WB + /NAT + , WB + /NAT -and WB − /NAT + .The prevalence of HIV was calculated by dividing the number of positive (WB or NAT) for the HIV donors by the total number of blood donors.
All titles and abstracts identified in the search were independently examined by two experienced investigators.The same two investigators then assessed the full text of any articles deemed potentially relevant, and then the disagreements between investigators were resolved by consulting a third investigator to reach a consensus.

| Literature search
We commenced designing and conducting this meta-analysis based on the proposed PRISMA guidelines.The search of electronic literatures on HIV epidemiology among blood voluntary donors in mainland China was performed using China National Knowledge Infrastructure, Medline, PubMed, and Wanfang Data from 1996 to 2022 with key words including "HIV," "human immunodeficiency virus," or "AIDS"; "blood"; "blood donors"; "blood centers"; or "donation"; "transfusion transmitted disease"; "screening reactive" or "disqualification"; "epidemiology"; and "China" or "Mainland China".No language restrictions were applied.

| Quality assessment of the studies
In consideration of the applicability of quality assessment, all included studies were assessed for methodological quality using the prevalence critical appraisal instrument developed by Munn and colleagues. 9

| Statistical analysis
Pooled results of HIV prevalence and corresponding 95% confidence were used to assess the condition of HIV infection in mainland China.The heterogeneity among studies was tested by the I 2 test and Q statistical test.On account of the high heterogeneity of collected studies in our research (I 2 > 75%, p < 0.1), [10][11][12][13][14] DerSimonian-Laird random-effects model was used to calculate the pooled odds ratios.Meta-regression analysis based on multiple factors (publication year, province, sex ratio, and sample size) was performed to find the source of heterogeneity.In addition, publication bias was assessed using Egger's tests, in which p < 0.05 represents significant difference.

| The selection of literatures
According to the inclusion and exclusion criteria, 498 studies of individual voluntary blood donors were identified in meta-analysis.
We collected demographic characteristics from these studies including district, sexuality, age, ethnicity, vocation, educational level, marriage, and blood type.Additional information including title, author, study year, sample capacity, the quantity of subjects with PLWH was extracted (Figure 1).

| General information
Our study included approximately 1.47 hundred million voluntary blood donors, demographic data from donors was tabulated, as presented in Table 1.
In this study, all the 31 provinces, autonomous regions and municipalities in mainland China were divided into seven areas: East China, North China, South China, Central China, Northwest China, Southwest China, and Northeast China.

| HIV prevalence
After screening, all the HIV-associated cases were tested by NAT or Western Blot (WB).498 studies were included in the investigation, the pooled statistical result of HIV prevalence is presented in Table 2.The overall pooled prevalence among voluntary blood donors in mainland China was 0.0178% (95% CI, 0.0169%−0.0187%)with a broad range of 0.0027%−0.0721%in different regions.Because I 2 = 95.8% and p < 0.001, there is the high heterogeneity in our research (I 2 > 75%, p < 0.1).There was an evident regional difference in HIV prevalence among blood voluntary donors (p < 0.05), in which Southwest China (0.0392%, 0.0332%−0.0452%)and South China (0.0222%, 0.0195%−0.0248%)had a relatively higher HIV prevalence, while the East China had the lowest HIV prevalence (0.0113%, 0.0102%−0.0124%).Moreover, Figure 2 indicated the geographical distribution of HIV prevalence, and the overall HIV prevalence varied annually (Figure 3).

| Demographics of HIV infection confirmed voluntary donors
Table 3 provides detailed information on HIV prevalence and the actual number of HIV-infected confirmed voluntary donors of subgroups.

| Gender
There were 109 studies brought into the analysis of the correlation between HIV prevalence and sex among voluntary blood donors.The prevalence in male voluntary donors was 0.0273% (95% CI, 0.0246%−0.0300%),and that in female donors was 0.0049% (95% CI, 0.0041%−0.0057%).The difference between the gender was significant (p < 0.05).

| Ethnicity
We investigated the association of HIV prevalence and ethnicity, and acquired HIV prevalence and 95%CI in the blood voluntary groups of which Han 0.0319% (0.0207%−0.0431%)and ethnic minority 0.0345% (0.0247%−0.0442%).Nevertheless, no significant difference was found in Han and ethnic minority donors (p > 0.05).

| Education
The analysis about the relevance of HIV prevalence and the education was conducted in accordance with 52 related studies about education.
Besides, there was a statistical difference between first and repeat donors (p < 0.05).

| Year
The overall HIV prevalence showed a fluctuation trend from 1998 to 2021.As shown in Figure 3

| Window period HIV infection
There were 52 studies concerned about the ELISA negative donors and the NAT positive donors.Through analysis, overall pooled HIV window period (ELISA negative& NAT positive) infection rate was 0.0475‱ (95% CI, 0.0304‱−0.0646‱)

| Meta-regression
A substantial heterogeneity (I 2 > 75%, p < 0.1) existed in included studies about HIV prevalence.The meta-regression analysis based on multiple factors (publication year, province, sex ratio [male/female], and sample size) showed that the variables included in the regression were province and publication year (p < 0.05), which probably explain the part of heterogeneity.

| Publication bias
We assessed the publication bias through Egger's test, acquired the meaningful publication bias (p < 0.05).Subsequently, the detailed analysis of publication bias for subgroups was developed: In geographic locations, the obvious publication bias (p < 0.05) existed in 11 provinces; There was a publication bias in most occupation, except for the soldier.In addition, significant publication bias (p < 0.05) occurred in the groups of gender, age, education, marriage and donation frequency, no significant publication bias (p > 0.05) was found in the rest of the groups or subgroups.
Besides, we investigated statistical characteristics of HIV screening of voluntary blood donors and analyzed associated factors.
See Supporting Information: Appendix for details.

| DISCUSSION
In mainland China, the provision of blood primarily depended on paid and employer-organized donations before 1998.When the "Blood Donation Law of the People's Republic of China" came into force since 1998, the government had banned all kinds of remunerative blood donation and the HIV prevalence rate decreased significantly in blood donors.However, the prevalence later has increased until 2015, which may attribute to an increasing amount of PLWH and the advancement of testing technologies, 15,16 especially the introduction of NAT reduced the window period.HIV prevalence has decreased slowly since 2016, this may be due to implementing the Action plan for AIDS containment and prevention in China in 13th 5-Year.
According to data compiled by Chinese CDC and Prevention and previous reports, the top five provinces with the largest amount of HIV-infected people are Sichuan, Yunnan, Guangxi, Guangdong, and Henan, [17][18][19] which were roughly consistent with the highest prevalence in voluntary blood donors in these provinces.Previous reports by the Chinese CDC and Prevention were conducted only on the general population during specific periods, such as 2010-2017.In this study, however, the range of investigation was broader.Thus, there are some discrepancies between our study and those reports.
On the one hand, although the number of HIV infections in Guangdong ranks among the top five in China, the epidemic in Guangdong is characterized by a low prevalence overall but high prevalence in localized areas.1][22] Our study focused on blood donors, and the number of studies from the above six cities only accounted for approximately 40% in Guangdong Province.Thereby the HIV prevalence among voluntary blood donors in Guangdong is not high.
On the other hand, changes in the epidemiological characteristics and the trend of rapid and then stable growth of HIV in the general population might affect the overall HIV prevalence among blood donors in Henan. 23,24 our study.The detailed information about HIV prevalence in blood donors can be seen in Table 2. Yunnan, Sichuan, Guangxi, with high HIV prevalence were also the labor exporting provinces which resulted in the high mobility of PLWH and the high-risk groups, 25 the government is supposed to strengthen capacity of HIV prevention and monitoring in these regions.For one thing, the deficiency of efficient prevention strategies and attention to high-risk groups might have contributed to the higher prevalence in areas with high incidence rates in China. 26We suggested that the government conducts focused testing for high-risk groups in high prevalence regions.8][29] As shown in In subgroup analysis, there are significant differences in HIV prevalence among voluntary donors regarding occupation, education and donation frequency.1][32] There were many first blood donors with high-risk behaviors seek for HIV test through blood donation, and resulted in an increase in the HIV-positive rate among the first blood donors.Meanwhile, Because of stable health, regular lifestyle and qualified numerous health evaluations before donation, repeat blood donors are low-risk donor group.They are more familiar with knowledge of blood donation and blood safety than first-time donors. 33,34 terms of occupation, HIV prevalence in the business service group was the highest and there was a significant difference between business service and the other occupations (p < 0.05), possibly because of the business service industry was more vulnerable to HIV infection. 35It's worth noting that commercial sexual activity is the primary method of heterosexual transmission. 36Recent studies have shown that the knowledge of HIV and its transmission is related to education status, and the lack of knowledge of HIV could hamper prevention, treatment, and care. 37,38And more attention should be paid to the lower education group and more comprehensive model of education for HIV prevention was urgently needed.
Moreover, there was a discrepancy of HIV prevalence between male and female donors (p < 0.05), and this is most probably due to heterosexual transmission and men who have sex with men (MSM) group.Firstly, the fast-growing share of sexual transmission is serious, 38 and HIV infection cases of Heterosexual transmission account for about 65% of all cases in the national work report of AIDS in 2016. 39,40Secondly, HIV/AIDS epidemic among MSM has been on the rise in mainland China in recent years. 41,42In 2019, the homosexual transmission accounted for one quarter of newly HIV infected. 43Recent studies have revealed the rate of new infections in MSM population was higher in North, East and Southwest China, and the rate in first tier cities was significantly high. 44Because the prevalence of HIV in the MSM group could result in the infection of HIV from the high-risk group to the general population in China, 45 it was necessary to take measures to prevent high risk behaviors among MSM.
There were some blood donors donated in the HIV window period in our study, which were qualified by ELISA, but were NAT reactive.The development of blood from these donors for clinical transfusion could cause HIV infection.There were many cases of HIV infection through transfusion during the window period in the country and abroad. 46,47The window period in HIV testing increases the potential risk of transfusion and seriously threatens blood safety.Overall, the heterogeneity cannot be eradicated or avoided in meta-analysis, particularly those from cross-sectional studies with wide distribution and long-term period. 48,49In regard to heterogeneity in this study, random-effects models was used to eliminate the effect of heterogeneity as much as possible.Additionally, we further analyzed the potential sources of heterogeneity through subgroup with meta-regression method, and the results showed that the province may contribute more heterogeneity than other groups.
In our study, all selected literatures were rigorously screened and were qualified for the meta-analysis.Publication bias was determined in some subgroups.Remarkably, there were less than five published reports in some provinces, which may weaken the effect of the meta-analysis.In addition, this study only included literatures published in Chinese and English, which may relatively play a critical part in the publication bias.
Even so, a comprehensive search strategy combined with the large sample size and various resources of literature in this study may have been conducive to minimization of publication bias.
The meta-analysis results presented here come with some limitations: (1) All studies used cross-sectional observational study design.
(2) Most of the literatures was Chinese language.
(3) Some confusing information from low quality literatures could not be confirmed and had to be excluded for failure to contact the authors.

AUTHOR CONTRIBUTIONS
Miao He and Zhan Gao conceived and designed the study.Yuhui Li, Yang Huang and Lei Gao did the literature search and data analysis, and wrote the first draft of the manuscript, which was reviewed by all other authors.Xinhui Zhang supervised the study and finalized the manuscript, which all authors read and approved.

1 |
Through meta-analysis, we have investigated the HIV blood screening reactive rate and prevalence, and obtained the demographics of blood voluntary donors in China.This study is in regard to a nationwide surveillance of AIDS cases and HIV infections among Chinese voluntary blood donors with the revealing of heterogeneity between previous studies during the past 27 years in mainland China.2| MATERIALS AND METHODS2.Literature selection and data extractionThe inclusion criteria were as follows: (1) The study was conducted only in mainland China (Hong Kong, Taiwan, and Macao were excluded); (2) The HIV screening test should be ELISA or NAT and ELISA reactive donors must be confirmed by WB.The screening strategy is supposed to do conform with relevant standardized donor screening policies; and (3) Experimental objects are blood voluntary donors.
Characteristic-based stratified analysis was performed and the Z or chi-square test was used to calculate statistical difference among subgroups.All data analysis was performed with computer software (Stata Statistical Software Package 12.0, StataCorp) for the statistical analysis.The PRISMA flow diagram and geographical distribution maps were made with Adobe Photoshop 12.0 (http://www.adobe.com/),and the trend plot was made by Joinpoint Regression Program (https:// surveillance.cancer.gov/help/joinpoint).

F I G U R E 2
Provincial distribution of the pooled HIV prevalence in voluntary blood donors in mainland China (A: Provincial distribution of the pooled HIV prevalence as at December 2004; B: Provincial distribution of the pooled HIV prevalence collected a decade ago).The horizontal bars of the different yellow color depth in the lower left corner indicate different ranges of rate (the base map is from the National Administration of Surveying, Mapping and Geoinformation of China [http://www.sbsm.gov.cn/] for free downloading and use).*Because the number of studies in some provinces is missing or few, the prevalence in these provinces are recorded as unknown.HIV, human immunodeficiency virus.

Figure 2 ,
Figure 2, compared with ten years ago, the HIV prevalences and characteristics of the epidemic in some provinces have drastically changed, such as Xinjiang and Henan.The situation of HIV infection was mitigated in Xinjiang.In addition, with the effective control of At present, NAT has been fully conducted for HIV testing of blood centers in China.Compared with serological method, NAT could effectively shorten the risk of transmitting HIV by transfusion during the window period.Most importantly, our findings indicated that HIV prevalence in voluntary blood donors has risen steadily and then declined, and this trend was practically consistent with the data from Chinese annual national notifiable infectious disease report.The situation of HIV infection of voluntary blood donors could indirectly reflect that of the general population.Although the residual risk of HIV infection in blood transfusion still is a great challenge, thanks to a new generation of ELISA reagents and NAT, the blood transfusion in China still safe.
Pooled HIV prevalence in voluntary blood donors in different areas of Chinese mainland.
T A B L E 2 a Data are reported as percent (95% CI).
T A B L E 3 Demographic characteristics of HIV infection confirmed voluntary donors.