Hepatitis B virus immunity prior to and after administration of a ‘booster’ dose of vaccine among health-care students at a South African university

Highlights

• •Health-care students (HCS) are at-risk of occupational exposure to hepatitis B.
• •Following a ‘booster’ hepatitis B vaccine, ∼7% of participants remained non-immune.
• •Pre-vaccination immunity contributed significantly to post-vaccination response.
• •Appropriate management of susceptible individuals is currently not being pursued.
• •Immunity testing following vaccination is highly recommended for health-care students.

Abstract

Background

Health-care students (HCSs) are at risk of occupational exposure to hepatitis B virus (HBV) infection despite an effective hepatitis B vaccine (HepB) being available. The majority of current HCSs are born after HepB was introduced into the South African Expanded Programme on Immunisation in 1995. Thus, it is assumed that having received HepB in infancy, a single ‘booster’ dose would suffice. This study aimed to investigate HBV immunity prior to and after administration of a HepB ‘booster’ dose.

Methods

Hepatitis B surface antibody (anti-HBs) levels were determined in first year HCSs at the University of the Witwatersrand, before and after receiving the ‘booster’. Participant demographics and HepB history were captured using a structured questionnaire.

Results

Before receiving the ‘booster’, 56% (101/180) had anti-HBs < 10 mIU/mL and were non-immune. A further 35% had anti-HBs levels of 10 – 99 mIU/mL, and 9% had ≥100 mIU/mL. <30% of HCSs self-reported completion of a three-dose primary series, which was significantly associated with higher baseline anti-HBs levels compared to those with a partial schedule (p = 0.045). Following vaccination, 39% (71/180) returned for follow-up with a significant median (IQR) increase of 476 (151 – 966) mIU/mL (p < 0.001). Of the 45 students who had non-immune baseline levels, 73% (33/45) responded with ≥100 mIU/mL, 16% (7/45) with 10 – 99 mIU/mL and 11% (5/45) remained non-immune. Levels of ≥100 mIU/mL were achieved by 100% of students with baseline levels ≥10 mIU/mL (n = 26).

Conclusion

More than half of the HCSs were not immune to HBV prior to receiving the recommended ‘booster’ vaccine. Following vaccination, 7% (5/71) remained unprotected. This study highlights that in the absence of vaccination records and without confirming the immune status of HCSs, it cannot be assumed that HCSs will be protected following a ‘booster’. Policy reform and inclusion of serological tests for immunity prior to HCSs initiating clinical exposure are recommended.

Abbreviations

anti-HBs

hepatitis B surface antibody

BBF

blood and bodily fluids

CHWC

Campus Health and Wellness Centre

CLIA

chemiluminescent immunoassay

DoH

Department of Health

EPI

Expanded Programme on Immunisation

FHS

Faculty of Health Science

HBsAg

hepatitis B surface antigen

HBV

hepatitis B virus

HCS

health-care student

HCW

health-care worker

HepB

hepatitis B vaccine

HepB-BD

hepatitis B vaccine birth dose

HepB3

third dose of hepatitis B vaccine

IPC

infection prevention and control

IQR

inter-quartile range

NHLS

National Health Laboratory Services

SD

standard deviation

SST

serum separator tube

WHO

World Health Organization

Wits

University of the Witwatersrand

Introduction

Hepatitis B Virus (HBV) infection is a global health problem that accounts for nearly one million deaths annually through complications of HBV-induced liver diseases, such as hepatocellular carcinoma and cirrhosis [1]. Despite the availability of a safe and highly effective vaccine for the past three decades, HBV remains endemic in many regions, including sub-Saharan Africa, where vaccine coverage remains sub-optimal [2].

Since the early 1990s, the World Health Organization (WHO) has recommended universal vaccination of all infants against HBV, with a minimum of three doses of HepB administered at least four weeks apart [3]. Furthermore, the WHO recommends that an additional dose be administered to neonates within 24 h of birth (HepB-BD) as the risk of developing chronic HBV infection is greatest amongst infants infected during the first year of life (80–90%) – this risk decreases exponentially if infection is acquired in adulthood (1–5%) [4], [5]. Therefore, global efforts towards HBV infection prevention and control have primarily focused on this younger cohort, resulting in significant reductions in children under 5 years becoming chronically infected from 4.7% to 1.3% [6].

Despite these successes, the prevalence of HBV infection amongst high-risk groups, such as health-care workers (HCWs) and those with immune-compromised states, remains high in many parts of the world [7], [8]. These groups include persons who are at increased risk of exposure to HBV because of occupational, behavioural, and biological susceptibility. To reduce the likelihood of infection among these high-risk groups, the WHO has further endorsed at least three doses of HepB in previously unvaccinated persons and booster vaccine doses in previously vaccinated individuals to reinstate immunity [1]. In addition, global organizations recommend serological testing to confirm immunity following vaccination, particularly in HCWs and students in training, prior to potential occupational exposure [9]. A measurement of antibodies to hepatitis B surface antigen (anti-HBs) of ≥10 mIU/mL, measured 1–3 months after administration of the last vaccine dose, confers immunity [10]. This value is commonly cited as the correlate of protection among healthy vaccinees in most countries. However, some countries regard higher anti-HBs levels of ≥100 mIU/mL as adequate immunity in high-risk groups – particularly HCWs [11]. This higher threshold-value has been adopted on the basis of observational studies where, several years after vaccination, significantly fewer breakthrough infections were seen in the ‘high-responder’ cohort (≥100 mIU/mL), when compared to lower threshold values in the ‘low-responder’ cohort (10 – 99 mIU/mL) [12].

Occupational exposure to infectious blood and bodily fluids (BBF) through muco-cutaneous or percutaneous routes are the major cause of HBV infection among susceptible HCWs [13], [14], occurring at rates up to four times that of the general population [15]. Annually, an estimated 2.1 million HCWs (5.9%) globally, are exposed to HBV through contaminated sharp objects, leading to an estimated 66 000 infections [16]. HCWs from developing countries, largely in Asia and Africa, account for more than 90% of these infections, with an annual prevalence of occupational exposure to infectious BBF among African HCWs reported at almost 50% [17], [18].

Health-care students (HCSs) are exposed to clinical situations alongside their professional counterparts and are thus at similar risk of HBV exposure during their academic training years [19]. Some studies have cited higher rates of needle stick injuries among trainees as a result of their zealous nature, coupled with a lack of procedural skills, incorrect use of personal protective equipment, and poor safety and infection control practices [20], [21]. In addition, students’ fear of repercussions, shame and personal lack of knowledge on post-exposure practices, may further prevent them from reporting the incident and in turn receiving the appropriate management [22].

While several African countries have strengthened policies for the control and prevention of HBV infection amongst HCWs, vaccine coverage in the region has remained sub-optimal [23], [24]. A pooled analysis reported a completed vaccination series in just under 25% of African HCWs, as low as 1% in some regions [23]. Interestingly, recent studies have shown increasing trends in vaccine uptake amongst HCSs - largely attributed to the introduction of national and institutional policies recommending (and in some instances, mandating) vaccination prior to clinical exposure [25], [26]. However, substantial variations are seen in knowledge, timing and provision of vaccination between academic institutions, even when standard national policies exist [27].

In South Africa, HepB was first introduced into the national Expanded Programme on Immunisation (EPI) in 1995 [28], and has seen a significant reduction in the prevalence of hepatitis B surface antigen (HBsAg) from highly endemic levels (>8%) to low carriage levels ranging from 0.0 to 2.7%, even among HIV-positive infants [29], [30]. Despite this, vaccine coverage, officially recorded two years after HepB introduction was 74%, and remained at<90% over the following two decades [31]. Consequently, nearly one-quarter of children born during that period (the current age-group of newly enrolled HCSs) are inadequately protected against HBV infection. In lieu of the anticipated poor vaccination coverage, as well as local HBV endemicity, the South African Department of Health (DoH) recommends the provision of three doses of HepB to previously unvaccinated HCWs and HCSs at-risk of HBV exposure [32]. However, national policy fails to translate these recommendations into practical and cost-effective strategies, leaving many local health authorities and academic institutions to create and implement their own guidelines. Despite protocols and guidelines to protect HCSs from occupational exposure to blood and blood products, up to one quarter of HCSs have reported occupational exposure to infections during their academic training, with more than a third of those exposed admitting to not having followed universal precautions or the recommended post-exposure prophylaxis guidelines [33]. Where vaccine coverage among HCWs in South Africa has been sub-optimal [24], HepB coverage appears promising among HCSs in South Africa (>90%) [34], [35].

The majority of HCSs enrolled in the 2021 health-related courses at the University of the Witwatersrand (Wits) were born after the introduction of HepB in 1995 and are thus presumed to have received adequate vaccination in childhood. On this basis, current vaccination policy at Wits stipulates the need for a single booster dose of the HepB to all HCSs during their first year of study, regardless of written proof of vaccination [36]. A three-dose schedule is reserved for those born before the introduction of hepatitis vaccine into the EPI, unless they can provide proof of completed HepB three-dose vaccination [36]. The coordination of the vaccination policy and administration of recommended vaccines, including HepB, is undertaken by the Wits student health clinic - Campus Health and Wellness Centre (CHWC). The financial cost of the vaccine is incurred by the HCS and billed to their respective fees account, providing the university with the only source of vaccine record for its HCSs. Post-vaccination serological testing is not currently offered nor recommended within the Wits policy.

The paucity of data on vaccination practices and the apparent under-appreciation for the need to confirm immunity following vaccination, requires further exploration and analysis [27]. Limited serological studies have been conducted amongst dental undergraduate HCSs in South Africa to evaluate immunity against HBV [34]. Our study thus aimed to quantify the levels of immunity to HBV among all first year HCSs across the Faculty of Health Science (FHS) at Wits, prior to them receiving the recommended single ‘booster’ dose of HepB (to establish baseline immunity). In addition, we aimed to analyse the effects of the ‘booster’ dose by comparing pre- and post-vaccination immunity levels.

Methods

Study population and recruitment site

We conducted a prospective cohort study of first year HCSs registered at Wits during the 2021 academic year. All registered students (N = 924) over the age of 18 years in the FHS were invited via email to participate in a survey which included a structured questionnaire, and blood sample collection to assess anti-HBs levels prior to (‘pre-vaccination’), and one month following vaccination (‘post-vaccination’). Data collection took place from March to November 2021. All students who took part in the study were actively recruited at the CHWC facility on the day of their scheduled vaccination. Eligible participants were taken through the study protocol to ensure their understanding of the risks and benefits of participation. Written informed consent was obtained from all willing participants.

Ethical approval

This study received ethical approval from the Wits Human Research Ethics Committee (medical) in February 2021 – reference number: M201157. All activities were conducted in compliance with the Declaration of Helsinki. Venesection was performed by registered health personnel in accordance with the National Infection Prevention and Control (IPC) Strategic Framework (2019) as set out by the South African DoH [37]. Any personal identifying information was coded and password-protected, accessible to the approved study researchers only.

Data collection

The structured questionnaire was used to determine participant demographics as well as HepB history. Questions regarding age, gender, childhood living environment and course of study were included in Section 1. Questions pertaining to HepB were included in Section 2. Despite a request by CHWC staff for all attending students to provide documentation of vaccination history, the majority of students did not comply. Therefore, details pertaining to Section 2, relied largely on student or caregiver recall and are reported herein without distinction.

For serological testing, approximately 3–5 mL of venous blood was drawn from each participant and collected in the appropriate serum separator tube (SST). CHWC nursing staff then administered a single dose of HepB [36]. A dose of 1.0 mL (20 μg HBsAg) of Euvax B™ Inj. (LG co., Korea) was injected into the participant’s non-dominant deltoid muscle [38]. Euvax B™ consists of highly purified, non-infectious particles of HBsAg produced by DNA recombinant technology in yeast cells (Saccharomyces cerevisiae). Aluminum hydroxide gel is used as the adjuvant and preserved with thimerosal [38].

All samples were delivered to the National Health Laboratory Services (NHLS), on-campus testing facility, within three hours of collection. Quantification of anti-HBs was done using validated chemiluminescent immunoassay (CLIA) technology. The assay used throughout this study was the Roche Elecsys® Anti-HBs II (Roche Diagnostics, Germany) according to the manufacturer's instructions. Overall sensitivity and specificity of this assay is estimated at 100% and 99.45%, respectively [39].

Participants with non-reactive anti-HBs concentrations of < 10 mIU/mL were considered ‘non-immune’. Those with reactive sample levels of ≥10 mIU/mL were stratified into two immune groups: 10 – 99 mIU/mL (‘immunity’) and levels ≥100 mIU/mL (‘high immunity’) baseline categories. Levels of anti-HBs < 10 mIU/mL or ≥1000 mIU/mL were not quantified further due to assay limitations and were therefore assigned values of 0 mIU/mL and 1000 mIU/mL, respectively, during data analysis. Official anti-HBs results, including interpretation of the findings, were emailed to the respective participants. At this time, participants were invited for post-vaccination anti-HBs testing (scheduled at least one month following vaccination) to assess the peak antibody response to the vaccine dose. Returning participants were again classified into three categories based on their anti-HBs response levels: <10 mIU/mL (‘non-immune’), 10 – 99 mIU/mL (‘low responders’), ≥100 mIU/mL (‘high responders’). The term ‘non-immune’ rather than ‘non-response’ was chosen in the case following vaccination, as prior vaccine history could not be verified. The term ‘non-response’ is reserved and internationally used for participants with non-reactive anti-HBs levels following exclusion of chronic hepatitis B infection and two three-dose courses of HepB, measured four weeks after the sixth HepB dose [40].

In cases where the anti-HBs concentrations remained < 10 mIU/mL after administration of the vaccine, respective participants were contacted and advised to return for further testing and management (offered through the study protocol).

Data capture

For each participant, pre- (baseline) and post- vaccination (if available) anti-HBs results, together with the coded responses from the corresponding questionnaire, were captured into an encrypted Microsoft® Excel (2021) document. De-identified hard copies of the questionnaires were scanned and stored on two electronic devices with the original copies held in a sealed cabinet at Wits.

Statistical analysis

Cleaned data was imported into SPSS Statistics version 20.0 (IBM Corp., Armonk, NY, USA) for analysis. Continuous variables were expressed as mean and standard deviation (SD) where appropriate, or median and inter-quartile range (IQR) where the distribution of data deviated from normal. All categorical variables were summarized as frequencies (n) and percentages (%). Pearson’s chi-squared test or Fisher’s exact test (where appropriate) were used to evaluate for significant associations between categorical variables and the main outcome of interest (anti-HBs ‘immune’ categories). Spearman’s rho was calculated to determine the correlation between pre- and post-vaccination anti-HBs levels. Due to the low sample size of returning participants, McNemar’s test for paired nominal data was used to determine the change from pre- to post- vaccination immune categories. Wilcoxon signed-rank tests for paired data were used to determine the significance of changes to anti-HBs levels following vaccination within immune categories. These tests were chosen as data was not normally distributed as raw values or when log-transformed. All tests were two-tailed, and a p-value of less than 0.05 was considered statistically significant. Graphs and tables were compiled using the GraphPad Prism software (version 5, USA).

Results

Invitations were emailed to 924 HCSs, of which 230 were screened at the CHWC. Of these, 222 participants met the eligibility criteria for inclusion in the study. Following discussion and written informed consent, 191 (86.0%) students were enrolled in the study. Anti-HBs results were not available for six students, and a further five did not complete the questionnaire portion of the study (Fig. 1). As a result, 180 (81.1%) students were included in the pre-vaccination analysis. Only 71 (39.4%) of these students returned following vaccination and completed both pre- and post- vaccination procedures as stipulated in the study protocol (Fig. 1).

Fig. 1.

Flow diagram of study participants for inclusion in pre-vaccination (N = 180) and post-vaccination (N = 71) data analysis anti-HBs (hepatitis B surface antibody); HBV (hepatitis B virus).

Pre-vaccination findings

The mean (SD) age of the study population was 19.0 (1.3) years, with the majority identifying as female (70%). Black students accounted for 80% (144) of the sample, and most (96.7%) of the participants were born in South Africa. Six of the nine undergraduate health-related courses at Wits were represented in the study, with no participants from nursing, physiotherapy, or occupational therapy disciplines (Table 1).

Table 1.

Sociodemographic characteristics of study participants from the Faculty of Health Science at the University of the Witwatersrand - Pre-vaccination survey (N = 180).

Characteristics	Category	n (%)
Date of birth	Before (and including) 1995	2 (1.1)
	1996 onwards	178 (98.9)
Sex	Female	126 (70.0)
	Male	54 (30.0)
Race	Black	144 (80.0)
	White	18 (10.0)
	Mixed race	6 (3.3)
	Indian	12 (6.7)
Nationality	South African	174 (96.7)
	Foreign national	6 (3.3)
Childhood residence*	Urban	110 (61.1)
	Rural	69 (38.3)
Course enrolment	Bachelor of Health Sciences	59 (32.8)
	Bachelor of Medicine and Bachelor of Surgery	48 (26.7)
	Bachelor of Pharmacy	28 (15.6)
	Bachelor of Clinical Medical Practice	20 (11.1)
	Bachelor of Dental Science	14 (7.8)
	Bachelor of Oral Health Sciences	11 (6.1)

Values reported as frequency (n) and percentages (%).

*N = 179 for childhood residence (single data point missing).

The results of the pre-vaccination testing revealed that most students (n = 101; 56.1%) had no immunity to HBV, with non-reactive (<10 mIU/mL) anti-HBs levels (Table 2). Of those students with reactive anti-HBs levels (n = 79), 16 students (20.3%) had high levels of immunity (≥100 mIU/mL) with the remainder of 63 (79.7%) students having lower levels of immunity (10 – 99 mIU/mL). Pre-vaccination anti-HBs levels showed no association with sex (p = 0.743), nationality (p = 0.664) or childhood residence (p = 0.396). A significant association was found between race and baseline anti-HBs (p = 0.034). Further subgroup analysis (through pairwise comparisons) revealed this difference was primarily due to higher pre-vaccination levels seen among the Indian students compared to the Black students (p = 0.009; adjusted by the Bonferroni correction for multiple tests) (Table 2 – Section 1).

Table 2.

Pre-vaccination anti-HBs immunity described according to sociodemographic characteristics and hepatitis B vaccine history (N = 180).

Category		Pre-vaccination anti-HBs titre			P value†
		non-immune	low immunity	high immunity
		(<10 mIU/mL)	(10–99 mIU/mL)	(≥100 mIU/mL)
SECTION 1. Demographics of Whole Population (N = 180)
Sex	Female	73 (72.3)	42 (66.7)	11 (68.8)	0.743
	Male	28 (27.7)	21 (33.3)	5 (31.2)
Race	Black	88 (87.1)	45 (71.4)	11 (68.8)	0.034‡ (0.009)§
	White	8 (7.9)	9 (14.3)	1 (6.3)
	Mixed race	2 (2.0)	2 (3.2)	2 (12.5)
	Indian	3 (3.0)	7 (11.1)	2 (12.5)
Nationality	South African	98 (97.0)	61 (96.8)	15 (93.8)	0.664
	Foreign national	3 (3.0)	2 (3.2)	1 (6.3)
Childhood residence	Urban	58 (58,0)	40 (63.5)	12 (75.0)	0.396
	Rural	42 (42.0)	23 (36.5)	4 (25.0)
Previously vaccinated against HBV	Yes*	38 (37.6)	24 (38.1)	4 (25.0)	0.226¶	0.262#
	No	14 (13.9)	3 (4.8)	1 (6.3)
	Unsure	49 (48.5)	36 (57.1)	11 (68.8)
Total 180		101 (56.1)	63 (35.0)	16 (8.9)
SECTION 2 - Vaccine History of Previously Vaccinated (n = 66*)
Completed three-dose schedule	Yes	27 (71.1)	22 (91.7)	2 (50.0)	0.045
	No	11 (28.9)	2 (8.3)	2 (50.0)
Time since last dose	1–5 years ago	0 (0.0)	1 (4.2)	1 (25.0)	0.028||	0.019**
	>10 years ago	33 (86.8)	20 (83.3)	1 (25.0)
	Unsure	5 (13.2)	3 (12.5)	2 (50.0)
Total 66		38 (57.6)	24 (36.4)	4 (6.0)

Anti-HBs (hepatitis B surface antibody); HBV (hepatitis B virus).

All values reported as frequency (n) and percentages (%); n (%).

* branching logic used to answer question on completion of vaccine schedule and timing since last dose, taken from those who answered ‘yes’ to previous vaccination (n = 66).

† Chi square testing (or Fischer’s exact test where needed) to determine p-value for all variables.

‡ Chi square test to determine significance between all categories of race and pre-vaccination immune category.

§ Pair-wise analysis between each race group and pre-vaccination immune category Chi-square test to determine significance adjusted with Bonferroni correction for multiple tests.

¶ Test included those who were previously vaccinated in the category ‘yes’ and ‘no’ only (n = 84).

# Test included those who were previously vaccinated in the category ‘yes’, ‘no’ and ‘unsure’ (N = 180).

|| Test included those who knew the time elapsed since their last dose in the category ‘1–5 years’ and ‘>10 years’ only (n = 56).

** Test included those who knew the time elapsed since their last dose in the category ‘1–5 years’, ‘>10 years’ and ‘unsure’ (n = 66).

The majority of students (96/180; 53.3%) were uncertain of their prior vaccination status, with 36.7% (66/180) students stating they had been previously vaccinated, and 10% (18/180) never having been vaccinated. A minority provided proof of vaccination (data not shown). No significant difference was seen in level of anti-HBs and previous vaccination status (p = 0.226) among these participants (n = 84). Accounting for the additional 96 students who were ‘unsure’ of their vaccine history, again no association was found between self-reported vaccine status (vaccinated, unvaccinated and unsure) and baseline anti-HBs levels when compared across all three groups (n = 180; p = 0.262) (Table 2 – Section 2).

Among students who reported that they were previously vaccinated (n = 66), the majority (51; 77.3%) had completed a three-dose schedule and were more likely to have reactive pre-vaccination anti-HBs levels (≥10 mIU/mL) compared to those who had not completed the schedule (p = 0.045). This difference was most notable in the lower immunity group (10–99 mIU/mL). Recent vaccination was significantly associated with higher immunity levels (p = 0.028). However, only two students had received a dose within the past five years, with both displaying immunity to HBV, compared to 54 students who received their last vaccine dose more than 10 years prior (Table 2 – Section 2).

Post-vaccination findings

Following vaccination, all students were invited to return for post-vaccination anti-HBs testing, regardless of their baseline levels. Approximately 40% (n = 71) of students returned for follow-up after a median (IQR) interval of 48 (41–57) days.

The mean (SD) age of the follow-up study population was 18.7 (0.6) years. Similar socio-demographic characteristics were seen in these students compared to the baseline, apart from students in the Bachelor of Oral Health Science discipline not being represented in post-vaccination analysis (Table 3).

Table 3.

Sociodemographic characteristics of study participants from the Faculty of Health Science at the University of the Witwatersrand - Post-vaccination survey (N = 71).

Characteristics	Category	n (%)
Date of birth	Before (and including) 1995	0 (0)
	1996 onwards	71 (100)
Sex	Female	49 (69.0)
	Male	22 (31.0)
Race	Black	59 (83.1)
	White	7 (9.9)
	Mixed race	3 (4.2)
	Indian	2 (2.8)
Nationality	South African	69 (97.2)
	Foreign national	2 (2.8)
Childhood residence	Urban	44 (62.0)
	Rural	27 (38.0)
Course enrolment	Bachelor of Health Sciences	32 (45.1)
	Bachelor of Medicine and Bachelor of Surgery	21 (29.6)
	Bachelor of Pharmacy	8 (11.3)
	Bachelor of Clinical Medical Practice	6 (8.5)
	Bachelor of Dental Science	4 (5.6)
	Bachelor of Oral Health Sciences	0 (0.0)

Values reported as frequency (n) and percentages (%).

Similar to the pre-vaccination results, post-vaccination anti-HBs groups showed no association with sex, nationality or childhood residence. In contrast to baseline results, no associations between post-vaccination anti-HBs groups and race, prior completion of the three-dose vaccine schedule, or timing of the last received dose were noted (Table 4).

Table 4.

Post-vaccination anti-HBs response described according to sociodemographic characteristics and hepatitis B vaccine history in students registered to the Faculty of Health Science, at the University of the Witwatersrand (N = 71).

Category		Post-vaccination anti-HBs titre			P value†
		non-immune	low responder	high responder
		(<10 mIU/mL)	(10–99 mIU/mL)	(≥100 mIU/mL)
SECTION 1. Demographics of Whole Population (N = 71)
Sex	Female	3 (60.0)	6 (85.7)	40 (67.8)	0.682
	Male	2 (40.0)	1 (14.3)	19 (32.2)
Race	Black	4 (80.0)	6 (85.7)	49 (83.0)	0.601
	White	0 (0)	1 (14.3)	6 (10.2)
	Mixed race	1 (20.0)	0 (0)	2 (3.4)
	Indian	0 (0)	0 (0)	2 (3.4)
Nationality	South African	5 (100.0)	7 (100.0)	57 (96.6)	0.811
	Foreign national	0 (0)	0 (0)	2 (3.4)
Childhood residence	Urban	1 (20.0)	4 (57.1)	39 (66.1)	0.109
	Rural	4 (80,0)	3 (42.9)	20 (33.9)
Previously vaccinated against HBV	Yes*	38 (37.6)	24 (38.1)	4 (25.0)	0.086‡	0.023§
	No	14 (13.9)	3 (4.8)	1 (6.3)
	Unsure	49 (48.5)	36 (57.1)	11 (68.8)
Total 71		5 (7.0)	7 (9.9)	59 (83.1)
SECTION 2 - Vaccine History of Previously Vaccinated (n = 28*)
Completed three-dose schedule	Yes	0 (0.0)	2 (66.7)	19 (79.2)	0.253
	No	1 (100.0)	1 (33.3)	5 (20.8)
Time since last dose	1–5 years ago	0 (0.0)	0 (0.0)	1 (4.2)	0.937¶	0.270#
	>10 years ago	1 (100)	2 (66.7)	23 (95.8)
	Unsure	0 (0.0)	1 (33.3)	0 (0.0)
Total 28		1 (3.6)	3 (10.7)	24 (85.7)

anti-HBs (hepatitis B surface antibody); HBV (hepatitis B virus).

All values reported as frequency (n) and percentages (%); n (%).

* branching logic used to answer question on completion of vaccine schedule and timing since last dose, taken from those who answered ‘yes’ to previous vaccination (n = 66).

† Chi square testing (or Fischer’s exact test where needed) to determine p-value for all variables.

‡ Test included those who were previously vaccinated in the category ‘yes’ and ‘no’ only (n = 84).

§ Test included those who were previously vaccinated in the category ‘yes’, ‘no’ and ‘unsure’ (N = 180).

¶ Test included those who knew the time elapsed since their last dose in the category ‘1–5 years’ and ‘>10 years’ only (n = 56).

# Test included those who knew the time elapsed since their last dose in the category ‘1–5 years’, ‘>10 years’ and ‘unsure’ (n = 66).

Pre-vaccination anti-HBs categorisation was the only factor that was significantly associated with post-vaccination anti-HBs categorisation (p < 0.001) when using McNemar’s test.

Following vaccination, a significant median (IQR) increase of 476 (151 – 966) mIU/mL was noted (p < 0.001). Of the 71 students who returned post-vaccination, 45 were non-immune at baseline with an overall median (IQR) increase within this group of 198 (66–781) mIU/mL (p < 0.001). Classification into immune response categories revealed 33 of the 45 students (73.3%) achieved high levels of immunity (≥100 mIU/mL) while 7 (15.6%) students had lower levels (10–99 mIU/mL). Of the 45 non-immune students, five (11.1%) remained non-reactive following vaccination. This equates to 7% of the total study sample (n = 5/71) who returned following vaccination. Table 5 below describes the characteristics of these five participants.

Table 5.

Socio-demographic characteristics and vaccine history for students who remained non-reactive (anti-HBs < 10 mIU/mL) following a single dose of hepatitis B vaccine.

ID Code	Age	Sex	Race	Nationality	Childhood residence	Course enrolment	Previous HepB Vaccination	Completed three-dose schedule*	Time since last dose*
1	18	Female	Mixed race	South African	Urban	Bachelor of Clinical Medical Practice	Unsure	N/A	N/A
2	18	Female	Black	South African	Rural	Bachelor of Health Sciences	Unsure	N/A	N/A
3	18	Male	Black	South African	Rural	Bachelor of Health Sciences	Unsure	N/A	N/A
4	19	Male	Black	South African	Rural	Bachelor of Pharmacy	Unsure	N/A	N/A
5	19	Female	Black	South African	Rural	Bachelor of Health Sciences	Yes	No	>10 years ago

anti-HBs (hepatitis B surface antibody); HBV (hepatitis B virus); N/A (not applicable).

*N/A. responses to ‘completion of a three-dose schedule’ and ‘time since last dose’ only apply to students who respond ‘yes’ to having received previous HepB.

Students with reactive baseline anti-HBs levels who returned post-vaccination, regardless of lower (10–99 mIU/mL) or higher (>100 mIU/mL) baseline immunity, all achieved high levels of immunity following the single vaccine dose (n = 26) (Fig. 2). A median (IQR) change of 946 (773–975) mUI/mL (p < 0.001) and 884 (311–891) mIU/mL (p = 0.125) was noted in the two groups, respectively. A statistically significant moderate correlation (r_s = 0.566, p < 0.001) was seen between all pre- and post-vaccination anti-HBs levels (Fig. 3).

Fig. 2.

Hepatitis B immune response following vaccination with a single ‘booster’ dose, grouped according to pre-vaccination (baseline) hepatitis B surface antibody (anti-HBs) immune category. Participants in various categories post vaccination reported as frequency (n) and percentages (%). Non-immune (anti-HBs < 10 mIU/mL), low immunity (anti-HBs 10 − 99mIU/mL), high immunity (≥100 mIU/mL). ^***McNemar’s test for significance (p < 0.001) for paired categorical data.

Fig. 3.

Post-vaccination anti-HBs level in comparison with pre-vaccination anti-HBs level (Spearman’s rank correlation (rho) r_s = 0.566, p < 0.001). Shaded area includes 95% confidence interval for linear plot; horizontal dotted line indicates anti-HBs < 10 mIU/mL (below which values are non-reactive) anti-HBs (hepatitis B surface antibody).

Discussion

This study aimed to determine the effect of a single ‘booster’ dose of HepB administered by the Wits CHWC to all FHS undergraduate students in their first year of study. A pre-vaccination questionnaire was completed by participants, and their baseline anti-HBs titres determined. During the second phase of the study, serology was repeated to evaluate individual pairwise differences in anti-HBs levels stratified according to previous vaccination status.

At the start of the 2021 academic year, more than 900 students registered for various first year health science programs, all of whom were instructed to receive HepB as stipulated by the Wits health policy for the prevention of HBV infection [36]. Approximately 25% of the 900 registered FHS students (n = 230) were sampled for participation in this study at CHWC, with 222 satisfying the selection criteria. This implies that the remainder of HCSs (∼75%) may have received their vaccination on alternative dates (when the study team was unavailable), at private health facilities, or not at all. The CHWC has no current mechanism to record the vaccination practices of the latter two groups, and no active investigation or follow-up is made on their vaccine status by CHWC or other Wits administrators. Pre-vaccination (baseline) anti-HBs testing and analysis was performed on 180 students (81% of the eligible participants, Fig. 1), with only 71 (39%) of these returning for post-vaccination testing. The poor follow-up may be because of a lack of awareness and appreciation for the potential risk of HBV among young HCSs.

The mean age of participants was 19 years and most participants were female, characteristics consistent with the current admission statistics in the FHS undergraduate programs at Wits. The distribution of the different race groups was representative of national South African demographics (80.6% Black and 8.0% White) [41]. No students registered to nursing, physiotherapy and occupational therapy disciplines participated in this study. The reason for this is unclear and raises concerns particularly with regards to nurses, who represent one of the highest exposure-prone HCW groups [42]. Further, the lack of representation limits the generalizability of our results and what inferences we can make about the target population.

Less than half of the participants (44%) had reactive anti-HBs samples pre-vaccination. Further when compared to European anti-HBs threshold values for adequate immunity among HCSs (anti-HBs ≥ 100 mIU/mL), only 9% of our study population met this criteria [11], [12]. These values were selected based on studies which identified lower rates of break-through infection among people with anti-HBs ≥ 100 mIU/mL compared to those in a lower immune group of 10 – 99 mIU/mL. However, these studies were conducted in regions with contrasting hepatitis B variants, endemicity and population characteristics [43], [44]. Further studies would be needed to determine the utility of these thresholds in the South African setting.

Most participants (56%) had non-reactive samples, with anti-HBs levels of < 10 mIU/mL at baseline, and were thus considered not immune to HBV. Baseline levels showed no significant association with sex, nationality, or childhood residence. However, a statistically significant difference between race groups was found. These results should be viewed with caution due to the lower representation of non-Black students. In addition our study does not account for potential confounding factors such as socio-economic status, access to health-care facilities, or other social factors which may account for the race differences seen here rather than any biological or genetic reasons.

The high proportion of non-immune students in our study (56%) could be due to inadequate vaccination (no previous vaccination or incomplete schedules) or the result of waning immunity in those who were previously vaccinated. Anti-HBs titres have been shown to wane from as early as one year following vaccination in immunocompetent responders [1], [45]. Similarly, waning levels of anti-HBs have been seen across various HCS populations several decades following successful complete three-dose HepB schedules. In countries where HBV endemicity is low, including those with higher vaccine coverage, better health infrastructure and explicit safety and prevention policies, individuals show persistent immunity lasting up to 20 years following vaccination [46], [47], [48]. Comparatively, in countries where HBV endemicity ranges from intermediate to high levels, similar to that found among South African populations, immune levels are seen to wane earlier [49], [50]. Newly enrolled HCSs in Hong Kong showed persistent levels of anti-HBs in just 18.9% of previously vaccinated students more than 10 years later [51]. The challenge of waning immunity can be addressed by additional booster vaccination to stimulate immunological memory and elicit an anamnestic response in persons who previously responded. Following booster vaccination with either a single or three booster doses, 85% and 100% of individuals with baseline anti-HBs < 10 mIU/mL responded positively, respectively [51]. A key difference between the latter studies and ours, is the inclusion of only participants who had received documented proof of a complete HepB series, which was not possible in the present study, but rather highlights key weaknesses in current policy.

Despite the request by CHWC for all students receiving vaccination to bring a copy of their childhood immunisation cards to the scheduled vaccination appointment, only a minority of students were able to comply with this request. As such, the vaccinators only administer a single ‘booster’ dose to all students born after the introduction of HepB in the South African EPI, under the assumption that these students had received at least three-doses of HepB in childhood [28]. Considering that vaccine coverage for the third dose of HepB (HepB3) in South Africa, from the time of introduction (1995) up until 2003 (age of the youngest enrolled study participant), has fluctuated between 74% and 80% [31], this policy may be based on inaccurate presumptions and the single ‘booster’ dose of vaccine may be insufficient for at least a quarter of HCSs.

Data on vaccine status was largely dependent upon the participant’s (or their caregiver’s) memory. The dependence on recall introduced bias into our study, meaning our data is less reliable than those corroborated with records and should thus be interpreted with caution. Unfortunately, the absence of health records is not a unique finding to our study, as many low and middle income countries report similar findings. In cases where health records are available, information is often inaccurate and/or incomplete and thus solely relying on records alone is not feasible either [52], [53], [54], [55].

Our study revealed no significant differences between vaccination status (‘yes’, ‘no’ or ‘unsure’) and baseline anti-HBs levels. However, it is difficult to draw conclusions from this finding due to potential biases in self-reported vaccination status. Reasons include the large number of students who were uncertain of their vaccine status, as well as the potential misclassification by some participants (who reported either ‘yes’ or ‘no’) due to recall or social desirability bias. Epidemiological studies such as ours are particularly susceptible to this, where a power dynamic between researcher and participant occurs and information given tends to overestimate ‘good’ health behaviours. In our case, participants may have been reluctant to disclose a history of no or incomplete prior vaccination [56].

Most of the participants (77,2%) who reported previous vaccination had completed a three-dose HepB series. Within this group, significantly higher pre-vaccination anti-HBs titres were found in those with a complete schedule when compared to students with a partial schedule (p = 0.045), accounting for a baseline immune-prevalence of 47% and 27%, respectively. This finding supports the conclusions of other studies which demonstrate significant positive correlations between anti-HBs titres and completion of HepB series up to 20 years following childhood vaccination [47], [48].

Post-vaccination analysis was limited to 71 students who returned between 1- and 3-months following vaccination (39% of study sample). Despite the high number of students who were lost to follow-up (n = 109), there were no significant differences in student characteristics between those who returned and those who did not. The only exception included those enrolled in the Bachelor of Oral Health Sciences course who were not represented in the post-vaccination analysis.

Of concern, this study revealed that 7% (5/71) of students returning for follow-up remained non-reactive following the ‘booster’ dose. All were sero-negative at baseline, and only one student reported receiving a partial vaccine schedule more than 10 years prior. The other four students (80%) were unsure of their previous vaccination status. The absence of any history of vaccination prohibits the attribution of non-response to any particular factors. It is not known whether participants had completed the primary three-dose schedule necessary to accurately define HepB non-response [57]. Furthermore, HBV infection, which may confound the interpretation of serological testing, was not evaluated in this study. Current Wits vaccination policy does not include the testing of HCSs for hepatitis B infection nor immunity following vaccination [36]. As such, students and faculty are not aware of susceptible individuals prior to clinical exposure and urgent policy reform to include such tests are needed. As per this study protocol, all five students were referred to a hepatologist for further investigations and management. The prime objective is to determine their HBsAg status.

Previous vaccine status, completion of the three-dose schedule (partial or complete) and timing of the last dose prior to receiving the ‘booster’ vaccine had no effect on post-vaccination anti-HBs levels. This disconnect may be due to recall bias and misclassification mentioned earlier. Pre-vaccination anti-HBs (baseline immunity) was the only variable that was significantly associated with an immune response following vaccination (p < 0.001) with a moderate positive correlation for the entire returning cohort (r_s = 0.566). The anti-HBs levels of students with non-reactive pre-vaccination serology (<10 mIU/mL) were not reported quantitatively, and thus all taken to be 0 mIU/mL. Consequently, tests of correlation within this subset could not be performed. These findings underscore the utility of pre-vaccination anti-HBs testing as a predictive factor for the effectiveness of a ‘booster’ dose in re-establishing immunity (in the presence of waning immunity), as well as preventing the unnecessary administration of a booster dose in those with already elevated levels of anti-HBs. However, a cost-effectiveness analysis should be conducted to compare universal adult HepB vaccination to the costs of anti-HBs testing prior to vaccination among HCSs.

A further fact to consider is that in our study the vaccine Euvax B™ Inj. (LG co., Korea) was administered as a booster to this cohort of HCSs. Euvax B™ contains HBsAg of serological subtype adr, corresponding to subgenotype C2, shown by epidemiological studies to prevail in the Korean population [58]. Genetic analysis of HBV strains distinguishes 9 genotypes A to I, which differ by at least 8% in complete nucleotide sequence and are further divided into subgenotypes with 4% genome divergence [59], [60]. The different genotypes and subgenotypes, which can have distinct geographical distribution, express different serological subtypes [60]. Yet, the majority of HepB vaccines administered globally contain HBsAg serological subtype adw2, which is found in HBV subgenotypes A1/A2/B1/B2/B6/C5/I1 and genotype G [60]. These subgenotypes circulate mainly in Europe, the United States of America and China, whereas the prevailing genotypes express different serological subtypes. In Western Africa, where HBV is hyperendemic, the prevalent genotype is E (serological subtype ayw4), in South East Asia genotype C (adr) prevails and the cosmopolitan genotype D, expresses ayw [60]. Although the consensus is that there is a high-degree of vaccine cross-protection between HBsAg genotypes and subtypes, there are studies, which have shown better protection against homologous HBV strains, compared to heterologous ones [61]. This cannot be overlooked in the context of South Africa. Administration of the Euvax B™ vaccine (adr) may not be as effective in protecting against the subgenotypes circulating in Africa, namely, subgenotypes A1 (adw2), D3 (ayw2/3) and genotype E (ayw4) [62], [63].

Limitations

Less than half the students enrolled returned for post-vaccination testing introducing attrition bias into the study. Together with the potential of recall and social desirability bias (previously described), the results reported here may not be generalizable to HCSs within the FHS at Wits or at other academic institutions. Factors, such as obesity, chronic illness, immune compromised states, tobacco use, genetic haplotypes [57], [63], [64], known to affect the immunogenicity of the vaccine, were not investigated in the present study. Finally, serological markers of either previous, acute, occult or chronic HBV infection [65] were not evaluated in this study. Therefore, immune status at baseline and following vaccination cannot be attributed to vaccination alone. Despite these limitations, this study provides valuable insights into current practices and policies at a South African tertiary institution, assessing the immune status and effect of ‘booster’ vaccination in at-risk students.

The way forward: recommendations

After receiving a ‘booster’ HepB dose, 7% of HCSs remained non-immune, highlighting that despite receiving the “booster” dose the HCSs were unprotected against HBV. Testing for immunity after receiving the booster is an important step in monitoring and ensuring adequate protection. We therefore recommend strengthening institutional vaccination policies and record keeping systems within the FHS and CHWC at Wits. Childhood vaccination records do not provide reliable proof of protection because immunity is expected to wane with time after a successful primary series. Nevertheless, digital platforms to centralize and store student historical and current vaccine records can provide a database to ensure completion of primary schedules, guide further management for students at risk and assist in quantifying true vaccine coverage rates for HBV amongst FHS students at Wits. Immunity testing prior to administration of HepB vaccination will avoid the administration of unnecessary booster doses to those who are adequately protected. However, a cost-effectiveness study will be required to determine the best approach. On the other hand, the inclusion of timely post-vaccination testing for immunity is highly recommended to ensure adequate protection of students before they enter clinical training. ‘Non-responders’ should be tested for HBsAg and those who are chronically infected counselled, and not employed in positions where they are a risk to their patients. HBsAg-negative non-responders should receive a total of 6 HepB doses to be classified as true non-responders and must receive hepatitis B immunoglobulin post-exposure. Institutional subsidy of immunity testing and immunization will ensure higher vaccine coverage and adequate immunity to ensure that HCSs are adequately protected when undertaking their clinical duties.

Funding

This study was funded by Professor E Song, Administered Fund, University of the Witwatersrand.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.