Abstract
Objectives:
Hepatitis B virus (HBV) infection remains a global health problem. Risk factors for HBV infection are usually assessed in prevalent rather than incident infections. To identify demographic and behavioral risks associated with incident HBV among South African blood donors.
Methods:
A case-control study was performed between November 2014 and January 2018. Cases were blood donors testing positive for HBV DNA with or without hepatitis B surface antigen but negative for antibody to hepatitis B core antigen. Participants completed an audio computer-assisted structured interview on exposures during the previous 6 months. Sex-specific multivariable logistic regression yielded independent associations between risks and HBV infection.
Results:
56 females and 37 males with incident HBV were compared to 438 female and 439 male controls, respectively. For females, risk factors were accepting money or goods for sex, using agents to prepare one’s anus prior to anal sex, penetrating injury, non-Black race, and lower educational status. Men reporting homosexual or bisexual orientation or sex with other men, previous injury, referral for HBV testing, or lack of medical insurance were at increased risk. For both sexes, having more than two male sexual partners increased risk.
Conclusions:
Sexual behaviors predominated over parenteral exposures as risks for incident HBV in both female and male blood donors.
Keywords: Hepatitis B virus, Transfusion, Blood donors, Risk factors
Introduction
Hepatitis B virus (HBV) infection remains a serious health problem globally, with 3.5% (296 million) of the adult population estimated to have chronic infection. Of 1.5 million new infections in 2019, most have been acquired in Africa [1]. Furthermore, HBV continues to be endemic in Africa and the Western Pacific countries, which carry the highest global burden with a hepatitis B surface antigen (HBsAg) prevalence of 6% each [2–4].
Worldwide, the Expanded Programme on Immunization (EPI) for HBV has been the single most important strategy, leading to a significant reduction of incident HBV infections in children under 5 years of age to less than 1% in 2019. Thereby, the global goal of the “Sustainable Development Goals” has been reached, but major gaps remain in the same regions that carry the highest burden of HBV [1]. In addition, HIV and HBV co-infections account for 2.7 millions of all global cases, highlighting that efforts directed against both epidemics remain essential [2].
Prevalent infection is the proportion of the population with mostly chronic HBV, while incidence counts persons with recently acquired HBV in a defined unit of time. Within the South African blood donor population, the prevalence of HBV among first-time blood donors decreased from 0.84–0.66% between 2011 and 2019, and there was a significantly lower HBV prevalence in the post- vs pre-HBV EPI birth cohorts (0.14% vs 1.29%) [5]. Similarly, HBV incidence in first-time donors has also decreased from 3.35 to 1.80 (1.33–2.45) per 1000 person-years for the same time periods [5]. Blood donor prevalence and incidence are significantly lower than reported for other populations in South Africa, most likely due to donor selection criteria and the healthy donor effect [6,7].
Factors associated with the prevalence and incidence of HBV infection may not be the same because of differences in the types of behavioral exposures during childhood and adulthood and changes in individual behaviors over time. Previously reported risk factors for prevalent HBV infections in the blood donor setting in Africa include being male, rural residence, younger age, lower education status, ethnic tattooing, being a first-time or family replacement donor as well as co-infection with HIV and hepatitis C virus (HCV) [8–14]. Reliable data on risk factors for incident HBV infections in both male and female adults are sparse but may provide better measures of contemporary behavioral risks in adults.
Current laboratory screening of blood donations in South Africa can differentiate between prevalent and incident HBV infections, allowing the study of demographic and behavioral risk factors associated with incident infection. We hypothesized that greater number of recent sexual partners, recent scarification/tattoo/body markings, and a history of personal contact with an HBV-infected person would be associated with incident HBV infection in both sexes.
Methods
Setting
The South African National Blood Service (SANBS) collects over 900,000 units of whole blood each year from a pool of approximately 350,000 donors. These donations are collected in eight of the nine provinces of the country, covering both urban and rural settings. Several strategies have been successfully implemented by SANBS to reduce the risk of transfusion-transmitted infections caused by HIV, HBV, HCV, and syphilis. SANBS performs health and infectious risk factor screening of all prospective donors by questionnaire and face-to-face interview, including questions on sexual behaviors and parenteral exposures linked to viral infection. In those deemed eligible for donation, a sample from each donation is tested for these viral infections by serology and individual donation (ID) nucleic acid testing (NAT) performed in parallel (Syphilis testing is by serology only).
Study design and population
We conducted a case-control study of incident HBV (cases) compared to infectious marker negative (controls) South African blood donors between November 2014 and January 2018. Study participants were identified from SANBS blood collection sites except for those in the Free State and Northern Cape area, where it was not possible to conduct study procedures for logistical reasons.
Cases were allogeneic whole blood donors who tested positive for incident HBV infection during the study period. Incident HBV-positive donors were defined as those who were HBV-NAT positive, HBsAg positive or negative, and total antibody to hepatitis B core antigen (anti-HBc) negative. Our definition requiring negative anti-HBc total (without testing for anti-HBc immunoglobulin M) was a more stringent approach to ensure that we identified and included cases who had been infected in the previous 6 months. Co-infection with HIV and/or HCV, whether incident or prevalent, was not an exclusion criterion. Controls were allogeneic blood donors negative for HIV, HBV, HCV, and syphilis, who were matched to HIV cases as part of a parallel study of risk factors for incident HIV infection cases (results of a parallel case-control study of incident HIV infection will be reported in a separate manuscript). For the risk factors in incident HIV infection, controls were frequency matched at a 3:1 ratio to incident HIV cases by race, age, and geographic location using information in the blood center database and then recruited either by recall or on the same day of their index blood donation to participate as a comparison group. The same controls were used for the analysis of risk factors for incident HBV. All controls were included, leading to a ratio of ∼9:1 controls per HBV case given lower rates of incident HBV infection in the study population.
Research nurses coordinated participant recruitment, information sharing about the study, consent, and enrollment. HBV cases were identified from their screening test results associated with their donation. Additional confirmation of HBV infection was conducted on a follow up sample post donation, when available. Potential cases were recalled and counseled about their donation results and then invited to participate. Controls were enrolled immediately after donation if all their markers were negative. All participants had to be able and willing to provide informed consent and complete the audio computer-assisted structured interview (ACASI) in English.
Donors were ineligible if they provided autologous or directed blood donations; if they were deferred from donation for health or infectious risk factors, were less than age 18, otherwise unable to provide informed consent, or had insufficient volume of blood collected to complete their screening tests.
Risk questionnaire
Eligible cases and controls completed a confidential ACASI on motivations for blood donation and behavioral factors. This risk factor interview was developed using content from previous studies of infection risks in blood donors conducted in Brazil using the same ACASI technique [15]. The interview content was modified to include reported and putative risk factors in Sub-Saharan Africa for both HIV and HBV infection. The instrument included questions on previous blood donations, perceived risk behaviors, sexual orientation and behavior, alcohol and drug usage, and questions on invasive interventions (medical procedures, scarifications). Risk exposures were assessed over two intervals, lifetime and within the 6 months preceding the donation.
Laboratory testing methods
All donations are routinely screened in parallel by ID-NAT using the Procleix Ultrio Plus (Elite) assay on Tigris (Panther), (Grifols, Sant Cugat del Vallès, Barcelona, Spain), and HBsAg on the PRISM-nEXT (Abbott, Wiesbaden, Germany). Donations that tested positive by NAT or NAT and HBsAg were further tested for anti-HBc total on the Roche Cobas e411 analyzer (Roche Diagnostics, Basel, Switzerland) to identify incident infections.
Data analysis
Questionnaire responses were linked with the participant’s laboratory and demographic data from the blood center’s operational databases and then extracted for analysis. Donation status was classified as either first time, lapsed (previous donation but none in the past year), or repeat (previous donation within the last year). The risk factor analysis by sex focused on behaviors reported in the 6 months before donation. Characteristics of the participants, by case vs control status, were stratified by sex and were reported using descriptive statistics. Frequencies and measures of statistical association for risk behaviors were compared for female and male cases and controls separately and are reported with a P <0.05 level of significance.
We used the Fisher’s exact test to test associations between categorical variables because of sparse observations in some strata.
Sex-specific data was further analyzed using multivariable logistic regression models for women and men separately. There were 17 variables for the female model. These variables were identified through the bivariate analysis and subject area knowledge to have potentially important relationships to HBV status. We used statistical learning to screen all possible models using these 17 variables. The algorithm created all possible 1-variable models, all possible 2-variable models, all possible 3-variable models, etcetera, and up to the single 17-variable model, for a total of 131,071 potential models for females. Following the same process used for the females, there were 18 variables and 262,143 potential models for males. For this study, we restricted the male models to have nine variables or less, for a total of 155,381 potential models. Misclassification rates for each model were calculated using five-fold cross-validation. For males and females separately, we investigated a small subset of all the potential models by focusing on models with the lowest misclassification rates. From this small subset of potential models, we selected a final model for each sex based on consensus among the researchers.
Results
During the study period, 2,707,385 blood donations were made of which 2429 (0.09%) donors tested HBV positive, including 217 (8.9%) incident HBV infections. Of these, 93 cases (46%) were successfully enrolled. Various factors contributed to potential cases not enrolling, including being untraceable, refusal to participate, change of mind after initial agreement, and not attending the research study appointment. Compared to those enrolled, those not enrolled were more likely to be male, of older age, of white/Asian race, and less likely to have donated in the Eastern Cape zone.
The 93 incident HBV cases were compared to 877 enrolled controls (Table 1), thus the ratio of controls to incident HBV cases for this analysis was approximately 9:1. For both females and males, cases and controls had similar demographic characteristics. Enrolled cases were predominantly of Black race (88%), and most were female (60%). Sixty-eight percent of enrolled cases were repeat donors, and 57% were in the age group 21–30 years. Two HBV incident cases had longstanding HIV co-infection.
Table 1.
Demographic comparison of incident hepatitis B virus cases and controls stratified by gender.
| Total | Females |
Males |
Totals |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cases | % | Control | % | Cases | % | Control | % | Cases | % | Control | % | |
| 56 | (100.0) | 439 | (100.0) | 37 | (100.0) | 438 | (100.0) | 93 | (100.0) | 877 | (100.0) | |
| Age | ||||||||||||
| <21 | 10 | (17.9) | 60 | (13.7) | 1 | (2.7) | 36 | (8.2) | 11 | (11.8) | 96 | (10.9) |
| 21–30 | 31 | (55.4) | 248 | (56.5) | 22 | (59.5) | 236 | (53.9) | 53 | (57.0) | 484 | (55.2) |
| 31–40 | 12 | (21.4) | 89 | (20.3) | 10 | (27.0) | 99 | (22.6) | 22 | (23.7) | 188 | (21.4) |
| 41+ | 3 | (5.4) | 41 | (9.3) | 4 | (10.8) | 66 | (15.1) | 7 | (7.5) | 107 | (12.2) |
| Missing | 0 | (0.0) | 1 | (0.2) | 0 | (0.0) | 1 | (0.2) | 0 | (0.0) | 2 | (0.2) |
| Race | ||||||||||||
| White | 3 | (5.4) | 17 | (3.9) | 2 | (5.4) | 18 | (4.1) | 5 | (5.4) | 35 | (4.0) |
| Black | 50 | (89.3) | 400 | (91.1) | 32 | (86.5) | 403 | (92.0) | 82 | (88.2) | 803 | (91.6) |
| Asian/Indian | 0 | (0.0) | 3 | (0.7) | 1 | (2.7) | 4 | (0.9) | 1 | (1.1) | 7 | (0.8) |
| Colored | 3 | (5.4) | 19 | (4.3) | 2 | (5.4) | 13 | (3.0) | 5 | (5.4) | 32 | (3.6) |
| Donor Type | ||||||||||||
| New | 15 | (26.8) | 85 | (19.4) | 6 | (16.2) | 35 | (8.0) | 21 | (22.6) | 120 | (13.7) |
| Lapsed | 6 | (10.7) | 101 | (23.0) | 3 | (8.1) | 43 | (9.8) | 9 | (9.7) | 144 | (16.4) |
| Repeat | 35 | (62.5) | 253 | (57.6) | 28 | (75.7) | 359 | (82.0) | 63 | (67.7) | 612 | (69.8) |
| Unknown | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | 1 | (0.2) | 0 | (0.0) | 1 | (0.1) |
| Zone | ||||||||||||
| Egoli/Vaal | 21 | (37.5) | 128 | (29.2) | 13 | (35.1) | 124 | (28.3) | 34 | (36.6) | 252 | (28.7) |
| Northern | 9 | (16.1) | 57 | (13.0) | 3 | (8.1) | 64 | (14.6) | 12 | (12.9) | 121 | (13.8) |
| KwaZulu Natal | 8 | (14.3) | 104 | (23.7) | 12 | (32.4) | 113 | (25.8) | 20 | (21.5) | 217 | (24.7) |
| Mpumalanga | 6 | (10.7) | 75 | (17.1) | 4 | (10.8) | 89 | (20.3) | 10 | (10.8) | 164 | (18.7) |
| Eastern Cape | 12 | (21.4) | 75 | (17.1) | 5 | (13.5) | 48 | (11.0) | 17 | (18.3) | 123 | (14.0) |
The bivariate risk behavior analyses for females are reported in Table 2a. For females, significant (P <0.05) risk factors were as follows: becoming sexually involved with partners who provide some material benefit such as food, shelter, transport, school fees, or other goods (accepting material goods for sex), having more than one male sexual partner, using anything to prepare one’s anus before or after having sex in the 6 months before donation, a percutaneous injury such as a knife or stab wound or being in an accident with loss of blood in the 6 months before donation, not being HBV vaccinated, and having an education level less than attaining a college or university qualification. Personal contact with an HBV-infected person yielded a borderline risk (P = 0.05).
Table 2a.
Potential risk factors for HBV infection in female cases vs controls.
| Female cases | % | Control | % | P-value | |
|---|---|---|---|---|---|
| All females | 56 | (100.0) | 439 | (100.0) | |
| Race | 0.74 | ||||
| White | 3 | (5.4) | 17 | (3.9) | |
| Black | 50 | (89.3) | 400 | (91.1) | |
| Asian/Indian | (0.0) | 3 | (0.7) | ||
| Colored | 3 | (5.4) | 19 | (4.3) | |
| Education category | 0.09 | ||||
| Up to Grade 10/Standard 8 | 5 | (8.9) | 27 | (6.2) | |
| Up to Grade 12/Standard 10 | 34 | (60.7) | 213 | (48.5) | |
| Tertiary educationa | 17 | (30.4) | 199 | (45.3) | |
| Medical aid | 0.19 | ||||
| No | 39 | (69.6) | 262 | (59.7) | |
| Yes | 17 | (30.4) | 177 | (40.3) | |
| Advised by health care worker to be tested for HBV | 0.10 | ||||
| No | 51 | (91.1) | 423 | (96.4) | |
| Yes | 5 | (8.9) | 14 | (3.2) | |
| unknown | 2 | (0.5) | |||
| Sexual orientation | 0.29 | ||||
| Straight/Heterosexual | 53 | (94.6) | 416 | (94.8) | |
| Bisexual/Gay/Homosexual | 2 | (3.6) | 22 | (5.0) | |
| Unknown | 1 | (1.8) | 1 | (0.2) | |
| Number of sexual partners in the last 6 months | <0.00 | ||||
| 0 | 9 | (16.1) | 175 | (39.9) | |
| 1 | 40 | (71.4) | 250 | (56.9) | |
| 2 to 3 | 6 | (10.7) | 13 | (3.0) | |
| 4+ | 1 | (1.8) | 1 | (0.2) | |
| Sexual involvement with partners who would provide material benefit b | <0.000 | ||||
| No Partners | 40 | (71.4) | 257 | (58.5) | |
| Some Partners | 3 | (5.4) | 2 | (0.5) | |
| All Partners | 4 | (7.1) | 5 | (1.1) | |
| unknown | 9 | (16.1) | 175 | (39.9) | |
| Endoscopy or colonoscopy in the last 6 months before the donation | 0.18 | ||||
| No | 54 | (96.4) | 434 | (98.9) | |
| Yes | 2 | (0.5) | |||
| unknown | 2 | (3.6) | 3 | (0.7) | |
| New tattoo or had one re-applied in the 6 months before donation | 0.57 | ||||
| No | 55 | (98.2) | 433 | (98.6) | |
| Yes | (0.0) | 3 | (0.7) | ||
| unknown | 1 | (1.8) | 3 | (0.7) | |
| Anus preparation practices before sex in the last 6 months c | 0.01 | ||||
| Every time | 7 | (12.5) | 14 | (3.2) | |
| Some times | 4 | (7.1) | 9 | (2.1) | |
| Once | 1 | (1.8) | 6 | (1.4) | |
| Never | 35 | (62.5) | 305 | (69.5) | |
| Not Applicable | 7 | (12.5) | 77 | (17.5) | |
| unknown | 2 | (3.6) | 28 | (6.4) | |
| Received Raatib, ritual scarring, piercing, circumcision, blood sharing, or been stabbed in the last 6 months | 0.69 | ||||
| No | 56 | (100.0) | 429 | (97.7) | |
| Yes | 2 | (0.5) | |||
| unknown | 8 | (1.8) | |||
| Injury in the last 6 months before donation d | 0.01 | ||||
| No | 49 | (87.5) | 425 | (96.8) | |
| Yes | 7 | (12.5) | 12 | (2.7) | |
| unknown | 2 | (0.5) | |||
| Intravenous drug usage | 1.00 | ||||
| Yes | |||||
| No | 56 | (100.0) | 437 | (99.5) | |
| Unknown | 2 | (0.5) | |||
| HBV vaccination | 0.03 | ||||
| Yes | 2 | (3.6) | 62 | (14.1) | |
| No | 51 | (91.1) | 364 | (82.9) | |
| Unknown | 3 | (5.4) | 13 | (3.0) | |
| HBV contact e | 0.05 | ||||
| Yes | 4 | (7.1) | 10 | (2.3) | |
| No | 46 | (82.1) | 417 | (95.0) | |
| Unknown | 6 | (10.7) | 12 | (2.7) |
HBV: hepatitis B virus.
Definitions:
Tertiary education means at least a college or university qualification.
Becoming sexually involved with partners who provide some material benefit such as food, shelter, transport, school fees, or other goods.
Frequency of using anything to dry, clean, or tighten anus before or after having sex.
An injury such as a knife or stab wound or being in an accident with loss of blood in the last 6 months before donation.
Hepatitis contact means having been in close contact in the last 6 months with anybody who has hepatitis (living or working with people who have hepatitis).
The multivariable analysis for females (Table 3a) found that accepting money or goods for sex with some partners (adjusted odds ratio [aOR] = 6.0 95% confidence interval [CI] 1.7–20.6), the use of agents to prepare the anus prior to anal intercourse occasionally (aOR = 4.0 95% CI 1.2–12.7) or always (aOR = 3.0 95% CI 1.0–9.0), percutaneous injury (aOR = 4.0 95% CI 1.3–12.5), and the number of male sexual partners (aOR = 10.23 95% CI 3.1–34.2 for 2+ vs no partners) conferred the highest odds of incident HBV infection. Being of another race (compared to Black race) was also associated with incident HBV infection (aOR = 2.9 95% CI 1.1–8.1). A higher level of education was borderline protective. Note, personal contact with an HBV-infected person was not retained in the final model.
Table 3a.
Independent risk factors for recent HBV infection in females.
| Variable | Adjusted odds ratios | Lower 95% confidence limit | Upper 95% confidence limit |
|---|---|---|---|
| Sexual involvement with partners who would provide material benefit | |||
| Not accepting | 1.00 | — | — |
| Accepting | 6.00 | 1.74 | 20.61 |
| Anus preparation practices | |||
| Never | 1.00 | — | — |
| Not reported | 1.05 | 0.45 | 2.48 |
| Occasionally | 3.96 | 1.24 | 12.65 |
| Every time | 3.04 | 1.03 | 8.95 |
| Injury in past 6 months | |||
| No | 1.00 | — | — |
| Yes | 3.95 | 1.25 | 12.53 |
| Race | |||
| Black | 1.00 | — | — |
| Other | 2.92 | 1.06 | 8.07 |
| Educational status | |||
| Primary or secondary | 1.00 | — | — |
| Tertiary | 0.53 | 0.27 | 1.02 |
| HBV vaccinated | |||
| No | 1.00 | — | — |
| Yes | 0.25 | 0.05 | 1.15 |
| Number of male sexual partners in the last 6 months | |||
| Zero | 1.00 | — | — |
| One | 3.15 | 1.42 | 6.98 |
| Two or more | 10.27 | 3.08 | 34.24 |
HBV: hepatitis B virus.
For males, potential risk factors from the bivariable analysis were reporting homo- or bisexual orientation, not having medical insurance, and having been told by a healthcare worker to be tested for HBV. No significant association was identified for having had contact with an HBV-infected person (Table 2b). The multivariable analysis for males (Table 3b) found that reporting homosexual or bisexual orientation (aOR = 8.0 95% CI 2.2–28.3), having more than two male sexual partners (aOR = 3.6 95% CI 0.9–13.8), having been told by a healthcare worker to be tested for HBV (aOR = 5.3 95% CI 1.7–16.2), not having medical insurance (aOR = 3.1 95% CI 1.3–7.6), and previous percutaneous injury (aOR = 6.1 95% CI 1.0–36.8) increased the odds of incident HBV. Being of another race (compared to Black race) was borderline associated with incident HBV infection (aOR = 2.8 95% CI 0.8–9.1).
Table 2b.
Potential risk factors for HBV infection in male cases vs controls.
| Male cases | % | Control | % | P-value | |
|---|---|---|---|---|---|
| All males | 37 | (100.0) | 438 | (100.0) | |
| Race | 0.26 | ||||
| White | 2 | (5.4) | 18 | (4.1) | |
| Black | 32 | (86.5) | 403 | (92.0) | |
| Asian/Indian | 1 | (2.7) | 4 | (0.9) | |
| Colored | 2 | (5.4) | 13 | (3.0) | |
| Education category | 0.80 | ||||
| Up to Grade 10/Standard 8 | 2 | (5.4) | 26 | (5.9) | |
| Up to Grade 12/Standard 10 | 20 | (54.1) | 208 | (47.5) | |
| Tertiary educationa | 15 | (40.5) | 203 | (46.3) | |
| Unknown | 1 | (0.2) | |||
| Medical aid | 0.01 | ||||
| No | 30 | (81.1) | 263 | (60.0) | |
| Yes | 7 | (18.9) | 175 | (40.0) | |
| Advised by health care worker to be tested for HBV | 0.06 | ||||
| No | 32 | (86.5) | 418 | (95.4) | |
| Yes | 5 | (13.5) | 19 | (4.3) | |
| unknown | 1 | (0.2) | |||
| Sexual orientation | 0.01 | ||||
| Straight/Heterosexual | 32 | (86.5) | 427 | (97.5) | |
| Bisexual/Gay/Homosexual | 5 | (13.5) | 9 | (2.1) | |
| Unknown | 2 | (0.5) | |||
| Number of sexual partners in the last 6 months | 0.17 | ||||
| 0 | 7 | (18.9) | 150 | (34.2) | |
| 1 | 27 | (73.0) | 253 | (57.8) | |
| 2 to 3 | 3 | (8.1) | 29 | (6.6) | |
| 4+ | 6 | (1.4) | |||
| Sexual involvement with partners who would provide material benefit b | 0.11 | ||||
| No Partners | 30 | (81.1) | 278 | (63.5) | |
| Some Partners | (0.0) | ||||
| All Partners | 11 | (2.5) | |||
| unknown | 7 | (18.9) | 149 | (34.0) | |
| Endoscopy or colonoscopy in the last 6 months before the donation | 0.08 | ||||
| No | 36 | (97.3) | 436 | (99.5) | |
| Yes | 2 | (0.5) | |||
| unknown | 1 | (2.7) | (0.0) | ||
| New tattoo or had one re-applied in the 6 months before donation | 0.15 | ||||
| No | 36 | (97.3) | 437 | (99.8) | |
| Yes | 1 | (2.7) | 1 | (0.2) | |
| Anus preparation practices before sex in the last 6 months c | 0.34 | ||||
| Every time | 5 | (13.5) | 24 | (5.5) | |
| Some times | 1 | (2.7) | 5 | (1.1) | |
| Once | (0.0) | 2 | (0.5) | ||
| Never | 17 | (45.9) | 211 | (48.2) | |
| Not Applicable, no anal sex | 11 | (29.7) | 153 | (34.9) | |
| unknown | 3 | (8.1) | 43 | (9.8) | |
| Received Raatib, ritual scarring, piercing, circumcision, blood sharing, or been stabbed in the last 6 months | 1.00 | ||||
| No | 37 | (100.0) | 434 | (99.1) | |
| Yes | 1 | (0.2) | |||
| unknown | 3 | (0.7) | |||
| Injury in the last 6 months before donation d | 0.10 | ||||
| No | 35 | (94.6) | 433 | (98.9) | |
| Yes | 2 | (5.4) | 5 | (1.1) | |
| Intravenous drug usage | 0.08 | ||||
| Yes | |||||
| No | 36 | (97.3) | 438 | (100.0) | |
| Unknown | 1 | (2.7) | |||
| HBV vaccination | 0.77 | ||||
| Yes | 3 | (8.1) | 40 | (9.1) | |
| No | 32 | (86.5) | 382 | (87.2) | |
| Unknown | 2 | (5.4) | 16 | (3.7) | |
| HBV contact e | 0.37 | ||||
| Yes | 1 | (2.7) | 5 | (1.1) | |
| No | 34 | (91.9) | 423 | (96.6) | |
| Unknown | 2 | (5.4) | 10 | (2.3) |
HBV: hepatitis B virus.
Definitions:
Tertiary education means at least a college or university qualification.
Becoming sexually involved with partners who provide some material benefit such as food, shelter, transport, school fees, or other goods.
Frequency of using anything to dry, clean, or tighten anus before or after having sex in the last 6 months before donation.
An injury such as a knife or stab wound or being in an accident with loss of blood.
Hepatitis contact means having been in close contact in the last 6 months with anybody who has hepatitis (living or working with people who have hepatitis).
Table 3b.
Independent risk factors for recent HBV infection in males
| Variable | Adjusted odds ratios | Lower 95% confidence limit | Upper 95% confidence limit |
|---|---|---|---|
| Sexual orientation | |||
| Heterosexual | 1.00 | — | — |
| Homosexual / bisexual | 7.97 | 2.24 | 28.34 |
| Injury in past 6 months | |||
| No | 1.00 | — | — |
| Yes | 6.13 | 1.02 | 36.79 |
| Advised by health care worker to be tested for HBV | |||
| No | 1.00 | — | — |
| Yes | 5.26 | 1.71 | 16.19 |
| Race | |||
| Black | 1.00 | — | — |
| Other | 2.75 | 0.83 | 9.09 |
| Having medical aid | |||
| Yes | 1.00 | — | — |
| No | 3.09 | 0.13 | 7.70 |
| Number of male sexual partners in the last 6 months | |||
| Zero | 1.00 | — | — |
| One | 2.55 | 1.04 | 6.25 |
| Two or more | 3.59 | 0.93 | 13.80 |
| HBV vaccinated | |||
| No | 1.00 | — | — |
| Yes | 1.03 | 0.25 | 4.21 |
HBV: hepatitis B virus.
Self-reported HBV vaccination status was not significantly associated with infection in either sex, although the odds ratio trended toward protection in women.
In our analyses, we also assessed other potential risk factors such as tattooing, traditional scarification, circumcision, iatrogenic/medical exposures, and intravenous drug use. We only had small numbers for these categories and found that none of these exposures were associated with incident HBV in either sex.
Discussion
This study assessed risk exposures associated with incident HBV infections in South African blood donors and identified primarily sexual risks for both males and females, similar to those reported for prevalent HBV infections in high-income countries. Of note, men who reported homo- or bisexual orientation or having sex with men (MSM) had increased risk of incident HBV. Women who reported accepting money or goods for sex and use of agents to prepare the anus before anal sex were at increased risk of incident HBV. For both sexes, percutaneous injuries were found to be significantly associated with incident HBV. We did not identify other parenteral or iatrogenic exposures as potential risk factors and had insufficient statistical power to detect rare potential risk factors such as scarifications.
The epidemiology of prevalent HBV infections in South Africa has changed in the last 30 years from a highly prevalent infection in rural children, associated with horizontal transmission, and in Black adult men, associated with scarification [16,17]. At the time, Karim et al. [18] found no association with homosexual orientation. Data collected in 2014 and 2015 from adults aged 15–49 years, residing in KwaZulu Natal demonstrated a much lower overall HBV prevalence of 4%, with no difference between adults living in rural and urban areas. However, males across all age groups remained more at risk than females, and among those co-infected with HIV, the prevalence of HBV increased to 6.2%. Furthermore, it has been shown that HBsAg positivity is 5% in HIV-infected persons and significantly higher than in non-HIV-infected individuals [19].
Publications on risk factors associated with incident HBV infections are scarce. We identified a survey of acute hepatitis conducted in the US general population in 2007, which found that HBV incidence varied with age and sex over the years. The highest rate (2.9 cases per 100,000 population) was reported among persons aged 25–44 years, and the lowest (0.02 cases per 100,000 population) was reported among persons aged <15 years, with the greatest percentage decline among persons aged <15 years (from 1.2 cases per 100,000 population in 1990 to 0.02 cases per 100,000 population in 2007) [20]. While before 2007, the female-to-male ratio was stable (1.5–1.8), in 2007, the rate for males was about 1.6 times higher than that for females. Contrary to the US study, which was not among blood donors but rather reflects cases reported to the Centers for Disease Control and Prevention, we found more incident HBV infections in young female blood donors than in male donors, but in keeping with the US findings, sexual exposures were the predominant risk factors in both sexes. A South African study reporting on HIV and HBV co-infections found that male sex and active tuberculosis were risk factors for incident infection [7]. We only had two HIV-HBV co-infections and could not analyze HIV’s effect on HBV infection. However, tuberculosis infection was not assessed in our study but was unlikely due to study population being pre-screened blood donors who are generally healthy.
We identified some unexpected risk factors in both male and female donors. Among males, the association with MSM was unexpected in the South African context as heterosexual behaviors are recognized as driving both the HIV epidemic as well as HIV/HBV co-infections. In addition, in females, exchanging sex for material gain and the use of anal preparation practices were risk factors that have not previously been described as a risk factor for incident HBV. However, it is perhaps expected as HIV transmission is driven by similar “blesser blesse” behaviors. This term describes relationships where older men have sexual relationships with younger women, who are given material benefit in exchange for sex, and is a common phenomenon in South Africa [21,22]. In a recent paper, Doyisa et al. [21] describe the HIV risk that such relationships posed to university students in Durban.
Self-reporting of percutaneous injuries was associated with incident HBV in both sexes. Other than intravenous drug use and needle sharing, well-known HBV risk factors in high-income countries, we did not identify reports of other percutaneous injury as risk factors [20]. Cultural scarification, tattooing, and piercings are hypothesized to be Africa-specific risks; however, injuries were not mentioned [13,17]. Unfortunately, our sample size was too small to definitively assess the latter. The reduced risk of HBV infection of lapsed blood donors is an additional unexpected finding as these donors have been found to be at higher risk for most other transfusion-transmitted viruses when compared to repeat donors. The finding may be due to chance because of the relatively small numbers of lapsed donors among the cases or the fact that there were relatively more first-time donors in the HBV cases vs controls (in contrast with more lapsed donors in the controls vs cases).
Some of the findings in our study differ from those found elsewhere. Incident HBV infection was not associated with lower education as reported from studies in African countries [9,12]. However, in our study, females (but not males) who had tertiary education were marginally protected from incident HBV, possibly indicating that education and socioeconomic status are important. Furthermore, we did not find a significant protective association of vaccination with incident HBV, although there was a trend toward protection in women. The lack of significant effects in our study was likely due to recall bias relating to childhood vaccination or the fact that up to half of the incident HBV cases were born prior to EPI implementation in 1995 and unlikely to have benefited from EPI as described by Prabdial-Sing et al. [23]. Finally, although contact with an HBV-infected person was of borderline significance in the bivariate analysis for females, there were very few positive responses, and so we did not find evidence for this being a risk factor for incident HBV.
Potential limitations of the study include participant selection and participation bias. The blood donor selection process is specifically aimed at excluding persons at risk of transfusion-transmissible infections, and the demographic and geographic distribution of those who enrolled differed somewhat from those who did not enroll. Furthermore, the number of incident HBV cases was not large enough to provide adequate power to detect lesscommon risk factors. Although we enrolled almost 50% of incident HBV cases, overall HBV incidence was lower during enrollment than during the design stage of the study. In addition, controls were matched to incident HIV cases and not incident HBV cases, although this does not appear to have significantly affected the demographic comparability of HBV cases and controls. Finally, although the use of a confidential, self-administered ACASI probably enhanced disclosure of risk factors, there may still have been recall bias as in any retrospective study.
In conclusion, this study demonstrates that sexual exposures are the primary risk factor for incident HBV infection in South African blood donors, including the novel findings of MSM and sex for material gain in women. This re-enforces the need for continuous education and promotion of HBV preventative strategies to sexually active adults. The sex-specific risk factors for HBV infection identified in this study may be useful in making the donor history questionnaire more sensitive for excluding at-risk female and male donors in the local context. Currently, the SANBS donor history questionnaire does not enquire about “blesser-blessee” relationships, nor does it enquire about anal sexual practices or penetrating physical injuries. Finally, at least in South Africa, blood services may be less concerned about including questions on traditional scarification and tattooing in their donor history questionnaires.
Acknowledgments
REDS-III South Africa program was the responsibility of the following principal investigators (PIs), committee chairs, and program staff. Field sites: Edward L. Murphy (PI, UCSF) and Ute Jentsch (PI, South African National Blood Service [SANBS]). Data Coordinating Center: Donald Brambilla and Marian Sullivan (co-PI’s Research Triangle International). Central Laboratory: Michael P. Busch (PI, Vitalant Research Institute). Steering Committee and Publications Committee chairs: Steven Kleinman and Roger Dodd. National Heart, Lung, and Blood Institute (NHLBI) program staff: Simone Glynn, Kelli Malkin, and Benyam Hailu for their oversight and research leadership. SANBS Virology Reference Laboratory staff for the management and co-ordination of laboratory testing for this study and the SANBS Nurses for enrollment and administration of the audio computer-assisted structured interview (ACASI). We thank Dr. Nareg Roubinian for his critical reading of the manuscript.
Funding
The work was supported by research contracts from the National Heart, Lung, and Blood Institute (NHLBI) of the National Institutes of Health for the Recipient Epidemiology and Donor Evaluation Study-III International program: HHSN268201100009I (to UCSF and the South African National Blood Service), HHSN268201100002I (to RTI International), and HHSN268201100001I (to Vitalant Research Institute). UJ, KvdB, BC, and ELM were also supported by the National Institutes of Health Fogarty International Center research training grant D43-TW010345.
Footnotes
Declaration of competing interest
The authors have no competing interests to declare.
Research ethics
Ethical committee review approval was provided by the South African National Blood Service (SANBS: approval 2013/013) institutional review board (IRB), the University of California San Francisco Committee on Human Research (approval 14-13943), and from the Research Triangle International (RTI) IRB. The study was also overseen by an Observational Study Monitoring Board constituted by the US National Heart, Lung, and Blood Institute, National Institutes of Health.
References
- [1].World Health Organization Global progress report on HIV, viral hepatitis and sexually transmitted infections. Accountability for the global health sector strategies 2016–2021: actions for impact. Geneva: World Health Organization; 2021. [Google Scholar]
- [2].World Health Organization Global hepatitis report 2017. Geneva: World Health Organization; 2017. [Google Scholar]
- [3].Basnayake SK, Easterbrook PJ. Wide variation in estimates of global prevalence and burden of chronic hepatitis B and C infection cited in published literature. J Viral Hepat 2016;23:545–59. doi: 10.1111/jvh.12519. [DOI] [PubMed] [Google Scholar]
- [4].Zampino R, Boemio A, Sagnelli C, Alessio L, Adinolfi LE, Sagnelli E, et al. Hepatitis B virus burden in developing countries. World J Gastroenterol 2015;21:11941–53. doi: 10.3748/wjg.v21.i42.11941. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Vermeulen M, Swanevelder R, Van Zyl G, Lelie N, Murphy EL. An assessment of hepatitis B virus prevalence in South African young blood donors born after the implementation of the infant hepatitis B virus immunization program: implications for transfusion safety. Transfusion 2021;61:2688–700. doi: 10.1111/trf.16559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Moonsamy S,Pillay P, Prabdial-Sing N (2023) Hepatitis B infection status among South Africans attending public health facilities over a five-year period: 2015to2019. PLOS Glob Public Health 3 (9): e0000992. doi: 10.1371/journal.pgph.0000992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Msomi N, Naidoo K, Yende-Zuma N, Padayatchi N, Govender K, Singh JA, et al. High incidence and persistence of hepatitis B virus infection in individuals receiving HIV care in KwaZulu-Natal, South Africa. BMC Infect Dis 2020;20:847. doi: 10.1186/s12879-020-05575-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Abdella S, Moshago Berheto T, Tolera G, Belete W, Deressa T, Feleke A, et al. Sero-prevalence of transfusion transmittable infections: HIV, hepatitis B, C and Treponema pallidum and associated factors among blood donors in Ethiopia: a retrospective study. PLoS One 2020;15:e0241086. doi: 10.1371/journal.pone.0241086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Biadgo B, Shiferaw E, Woldu B, Alene KA, Melku M. Transfusion-transmissible viral infections among blood donors at the North Gondar district blood bank, northwest Ethiopia: a three year retrospective study. PLoS One 2017;12:e0180416. doi: 10.1371/journal.pone.0180416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Deressa T, Birhan W, Enawgaw B, Abebe M, Baynes HW, Desta M, et al. Proportion and predictors of transfusion-transmissible infections among blood donors in North Shewa Zone, Central North Ethiopia. PLoS One 2018;13:e0194083. doi: 10.1371/journal.pone.0194083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Eko Mba JM, Bisseye C, Ntsame Ndong JM, Mombo LE, Bengone C, Mouelet Migolet G, et al. Prevalent hepatitis B surface antigen among first-time blood donors in Gabon. PLoS One 2018;13:e0194285. doi: 10.1371/journal.pone.0194285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Motayo BO, Faneye AO, Udo UA, Olusola BA, Ezeani I, Ogiogwa JI. Sero-prevalence of transfusion transmissible infections (TTI), in first time blood donors in Abeokuta, Nigeria. Afr Health Sci 2015;15:19–24. doi: 10.4314/ahs.v15i1.3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Nagalo MB, Sanou M, Bisseye C, Kaboré MI, Nebie YK, Kienou K, et al. Sero-prevalence of human immunodeficiency virus, hepatitis B and C viruses and syphilis among blood donors in Koudougou (Burkina Faso) in 2009. Blood Transfus 2011;9:419–24. doi: 10.2450/2011.0112-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Noubiap JJ, Joko WY, Nansseu JR, Tene UG, Siaka C. Sero-epidemiology of human immunodeficiency virus, hepatitis B and C viruses, and syphilis infections among first-time blood donors in Edéa, Cameroon. Int J Infect Dis 2013;17:e832–7. doi: 10.1016/j.ijid.2012.12.007. [DOI] [PubMed] [Google Scholar]
- [15].de Almeida-Neto C, Goncalez TT, Birch RJ, de Carvalho SM, Capuani L, Leão SC, et al. Risk factors for human immunodeficiency virus infection among Brazilian blood donors: a multicentre case-control study using audio computer-assisted structured interviews. Vox Sang 2013;105:91–9. doi: 10.1111/vox.12028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Abdool Karim SS, Coovadia HM, Windsor IM, Thejpal R, van den Ende J, Fouche A. The prevalence and transmission of hepatitis B virus infection in urban, rural and institutionalized black children of Natal/KwaZulu, South Africa. Int J Epidemiol 1988;17:168–73. doi: 10.1093/ije/17.1.168. [DOI] [PubMed] [Google Scholar]
- [17].Abdool Karim SS, Thejpal R, Singh B. High prevalence of hepatitis B virus infection in rural black adults in Mseleni, South Africa. Am J Public Health 1989;79:893–4. doi: 10.2105/ajph.79.7.893. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Samsunder N, Ngcapu S, Lewis L, Baxter C, Cawood C, Khanyile D, et al. Seroprevalence of hepatitis B virus: findings from a population-based house-hold survey in KwaZulu-Natal, South Africa. Int J Infect Dis 2019;85:150–7. doi: 10.1016/j.ijid.2019.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Firnhaber C, Reyneke A, Schulze D, Malope B, Maskew M, MacPhail P, et al. The prevalence of hepatitis B co-infection in a South African urban government HIV clinic. S Afr Med J 2008;98:541–4. [PMC free article] [PubMed] [Google Scholar]
- [20].Centres for Disease Control and Prevention Surveillance for acute viral hepatitis - United States, 2007. Coordinating center for health information and service. Atlanta: Centers for Disease Control and Prevention (CDC), United States Department of Health and Human Services; 2009. [Google Scholar]
- [21].Doyisa N, Maharaj P, Dunn S. Older men, younger women: blesser-blessee relationships among university students in Durban. South Africa. Cult Health Sex 2023;25:1530–42. doi: 10.1080/13691058.2022.2164065. [DOI] [PubMed] [Google Scholar]
- [22].Mah TL, Maughan-Brown B. Social and cultural contexts of concurrency in a township in Cape Town. South Africa. Cult Health Sex 2013;15:135–47. doi: 10.1080/13691058.2012.745951. [DOI] [PubMed] [Google Scholar]
- [23].Prabdial-Sing N, Makhathini L, Smit SB, Manamela MJ, Motaze NV, Cohen C, et al. Hepatitis B sero-prevalence in children under 15 years of age in South Africa using residual samples from community-based febrile rash surveillance. PLoS One 2019;14:e0217415. doi: 10.1371/journal.pone.0217415. [DOI] [PMC free article] [PubMed] [Google Scholar]