Professional Documents
Culture Documents
1
These authors contributed equally to Methods
this article. We searched Medline, Global Health, Embase, African Journals Online and
African Index Medicus. We included observational or interventional studies
that enrolled infants of HBV-infected women, and that tested for HBsAg or
Publication data
Submitted 11 April 2016 HBV DNA between 3 and 12 months of age.
First decision 28 May 2016
Resubmitted 15 August 2016 Results
Accepted 17 August 2016 Fifteen articles from 11 African countries were included. Among HBeAg-posi-
EV Pub Online 15 September 2016
tive women, the pooled risk was 38.3% (95% CI: 7.0–74.4%) without prophy-
As part of AP&T’s peer-review process, a
laxis, which was significantly lower than the lower bound of 70–90% risk in the
technical check of this meta-analysis literature (P = 0.007). Among HBeAg-negative women, the pooled risk was
was performed by Dr Y. Yuan. The 4.8% (95% CI: 0.1–13.3%) without prophylaxis, which lays within the lower
Handling Editor for this article was range of the 5–30% risk in Asia. By extrapolating the pooled transmission risks
Professor Geoffrey Dusheiko, and it was
to the number of births to infectious mothers, an estimated 1% of newborns
accepted for publication after full
peer-review. (n = 367 250) are annually infected with HBV at birth in sub-Saharan Africa.
Conclusions
Compared to Asia, the risk of mother-to-child transmission is low in sub-Saharan
Africa. However, the annual number of infants perinatally infected with HBV is
twice the number of incident paediatric HIV infections in sub-Saharan Africa
(n = 190 000). This highlights the importance of preventing mother-to-child
transmission of HBV in sub-Saharan Africa, which has been long neglected.
form (Data S3). Extracted information included study compared with the lower boundary of the transmission
design, settings, any preventive measures [hepatitis B risk frequently quoted in the literature: 70% for
vaccine, hepatitis B immunoglobulin (HBIG), anti-viral HBeAg-positive mothers and 5% for HBeAg-negative
therapy], type of serological/virological assays, maternal mothers.4, 11–13
information (median/mean age, HBV genotypes, HIV Subgroup analyses were performed to explore the
co-infection, obstetric complications, mode of delivery), sources of between-study heterogeneity in the risk of
infant information (type of HBV marker assessed and mother-to-child transmission without any prevention
age at assessment), and outcomes (the number of eligi- from HBeAg-positive mothers and HBeAg-negative
ble children with HBsAg-positive or HBV DNA-posi- mothers. The following potential sources of heterogeneity
tive mothers, the number followed up, the number were defined a priori and a test of heterogeneity was car-
infected) separated by maternal HBeAg status. When- ried out using the ‘metaprop’ command20: study region;
ever possible, the outcomes were further stratified by type of serological assay [moderately sensitive assays:
maternal HBV DNA status. Nine corresponding radioimmunoassay (RIA), haemagglutination assay (HA)
authors were contacted for further information, five and immunochromatography (IC); or highly sensitivity
replied and three provided numerical data that were assays: enzyme immunoassay (EIA) and chemilumines-
not presented in the published papers. Where studies cent immunoassay (CIA)]; sample size (N ≤ 10, 11–30,
did not mention whether the hepatitis B vaccine was >30); maternal virological status (HBV DNA levels,
administered or not, we ascertained, via the WHO/ genotypes, HIV co-infection); and infant information
UNICEF report, whether the vaccine was integrated in [type of HBV marker assessed and age at assessment
the national programme in this country at that time. (<6 months or ≥6 months)]. A multivariable meta-
The included studies were also evaluated for the risk regression was not conducted due to the small number
of bias (Data S4).19 of studies included.
The effect of each preventive measure was assessed by
Statistical analysis comparing the pooled transmission risks between those
The principle summary measures were the risk of with and without the intervention, after stratification by
mother-to-child transmission without any preventive maternal HBeAg status.
intervention in infants of mothers positive for both Finally, we estimated the number of infants infected
HBsAg and HBeAg, and in infants of mothers positive with HBV at birth each year in sub-Saharan Africa. We
for HBsAg but negative for HBeAg. This was defined first obtained the number of births to infectious mothers
as the proportion of the number of infants positive by extrapolating the prevalence of HBsAg in each coun-
for HBsAg and/or HBV DNA divided by the total try1 and prevalence of HBeAg in women of child-bearing
number assessed. The transmission risk from mothers age in SSA24 to the number of births in each country as
with occult HBV infection (defined as negative HBsAg per the UN Population Division25 for the period of
and detectable HBV DNA) was also estimated. If a 2010–2015 as below.
study reported children’s outcomes at ≥2 time points
between 3 and 12 months of their age, the latest time ðNo. of birth to HBsAg-positive HBeAg-positive mothersÞ
point was used. To pool the risks, meta-analysis was ¼ ðannual number of birthsÞ ðprevalence of HBsAgÞ
carried out using ‘metaprop’ command20 with STATA ðprevalence of HBeAgÞ
13.1 (STATA Corporation, College Station, TX, USA).
After the variance of the proportions was stabilised ðNo. of birth to HBsAg-positive HBeAg-negative mothersÞ
using Freeman-Tukey double arcsine transformation,21 ¼ ðannual number of birthsÞ ðprevalence of HBsAgÞ
estimates were pooled using the DerSimonian-Laird ð1 prevalence of HBeAgÞ
random-effects model.22 The confidence intervals for
the individual studies were presented with the score- Then, we extrapolated the estimated transmission risks
test while those for the pooled estimates were pre- by type of prevention (no prophylaxis, timely birth dose
sented with the Wald test.20 The percentage of total vaccine or delayed vaccine) to the number of births to
variation between studies due to heterogeneity was HBeAg-positive and HBeAg-negative mothers, after
assessed using the I2 statistic.23 Using the Z test, the taking account of country-specific hepatitis B vaccine
pooled African estimates without any prophylaxis were coverages in 2014 as below.26
ðNo. of infants infected with HBV at birthÞ ¼ ðNo. of birth to HBsAg-positive HBeAg-positive
mothersÞ fðvaccine coverageÞ ðtransmission risk from HBeAg-pos mothers with vaccineÞ
þ ð1 vaccine coverageÞ ðtransmission risk from HBeAg-pos mothers without prophylaxisÞg
þ ðNo. of birth to HBsAg-positive HBeAg-negative mothersÞ fðvaccine coverageÞ
ðtransmission risk from HBeAg-neg mothers with vaccineÞ þ ð1 vaccine coverageÞ
ðtransmission risk from HBeAg-neg mothers without prophylaxisÞg
Seven countries (Botswana, Cabo Verde, Djibouti, The in the former study the information on the control group
Gambia, Namibia, Nigeria and Sao Tome and Principe) was insufficient and we could only include the interven-
have already integrated birth dose vaccine into their tion group in the analysis. Of the 13 observational studies,
national programmes, and thus we used the coverage of six were conducted without any prophylactic mea-
birth dose vaccine and the estimated risk of transmission sures.30, 32, 33, 36–38 Two Gambian studies provided hep-
when the birth dose was administered. For the rest of atitis B vaccine with varying schedules (median: 3 weeks,
the countries, we applied the coverage of three doses of 95% received at >1 week after birth).28, 31 One Nigerian
hepatitis B vaccine and the estimated transmission risk study provided three different prophylactic regimens: hep-
when the vaccine was delayed for >1 week. atitis B vaccine and HBIG within 24 h (n = 16); hepatitis
B vaccine within 24 h without HBIG (n = 12); and hep-
RESULTS atitis B vaccine at 2–7 days without HBIG (n = 11).27
The search strategy identified 4274 papers. After exclud- Four studies exclusively recruited women co-infected with
ing 1452 articles in duplicate, 2822 were screened, and HIV; three of these provided maternal anti-retroviral regi-
69 papers were assessed in full text. Finally, 15 papers mens including lamivudine35, 40, 41 or tenofovir41 during
were included in this review (Figure 1). Reasons for pregnancy, with a median duration of 2,40 341 and
exclusion are presented in Figure 1 and Table S1. 4 months35 before the delivery. The fourth study of co-
After excluding two studies that did not report the infected mothers did not provide any pre-natal maternal
number of women screened,27, 28 a total of 14 239 treatment, however, lamivudine was given to both moth-
women were screened for HBsAg of whom 1143 tested ers and infants for the first 7 days after delivery.39 In all
positive (pooled prevalence of 8.5%, 95% CI: 7.3–9.9%). studies involving co-infected mothers, the infants were
HBeAg was also tested in 951 HBsAg-positive women, given hepatitis B vaccine after 6–8 weeks of birth as part
and 171 were positive (pooled prevalence of 16.2%, 95% of the standard of care. Apart from the Nigerian study,
CI: 10.7–22.6%). Of 1222 infants of HBsAg-positive none of the studies provided HBIG to infants.
mothers, 919 were finally assessed for HBV markers Maternal HBV status was determined prenatally
and 745 infants, of whom 127 were born to HBeAg- (n = 7),27, 28, 33–35, 39, 41 at delivery (n = 7)29, 30, 32, 36–38, 40
positive mothers and 618 to HBeAg-negative mothers, and post-partum (n = 2).31, 34 Maternal HBeAg was
provided at least one data point between 3 and assessed in all the included studies, however, HBV DNA
12 months of age. was examined only in six studies.32, 33, 36, 39–41 HBV
genotype in mothers was only reported in one study from
Characteristics of the included studies Malawi, for which all (100%) were genotype A.39 Women
The majority of the included studies were conducted in with pre-term labour were excluded in one study,37 and
West Africa (n = 9),27–35 followed by East (n = 5)36–40 mode of child delivery was described in two studies.27, 40
and Southern Africa (n = 1)41 (Table 1). There were 13 Two studies employed HBV DNA as a marker of child
observational cohort studies27, 28, 30–33, 35–41 and two infection39, 41 while the rest used HBsAg. The risk of bias
nonrandomised controlled trials:29, 34 in Senegal, Yvonnet in the included studies is summarised in Table S2.
et al. assessed the efficacy of the hepatitis B birth dose
vaccine (<24 h after birth) vs. no intervention29; and in Risk of mother-to-child transmission without
Ivory Coast, Ekra et al. assessed the efficacy of starting the preventive measures
hepatitis B vaccine within 24 h of birth vs. within The pooled risk of mother-to-child transmission, without
6–8 weeks.34 The latter provided two-independent cohorts any preventive measures, in infants of HBeAg-positive
in our analysis (birth dose cohort, 6 weeks cohort), while mothers was 38.3% (total sample size n = 34, 95% CI:
Manually added (N = 2)
7.0–74.4%) (Figure 2). This was significantly lower than strong evidence among HBeAg-negative mothers
the 70–90% risk that has been repeatedly quoted in the (I2 = 71.5%, P = 0.007).
literature (P = 0.007). The pooled risk in infants of
HBeAg-negative mothers was 4.8% (total sample size Subgroup analysis
n = 184, 95% CI: 0.1–13.3%) (Figure 3). This was not Of the eight potential sources of heterogeneity, maternal
different from the lower boundary of the 5–30% risk HBV DNA, genotype and HIV co-infection were not
presented in the literature (P = 0.2). There was good evi- assessed since only few reported on these variables. Simi-
dence of between-study heterogeneity among studies of larly, infant information was not evaluated because the
HBeAg-positive mothers (I2 = 50.6%, P = 0.09), and type of HBV marker and age at assessment did not vary
CIA, chemiluminescent immunoassay; Co, cohort study; HA, haemagglutination assay; EIA, enzyme immunoassay;
IC, immunochromatography; In, interventional study; N/A, not applicable; N/R, not reported; RIA, radioimmunoassay.
* Presented as s.d., or range, or IQR.
† Sixteen neonates received both hepatitis B vaccine and HBIG within 24 h of birth, 12 received only hepatitis B vaccine within
24 h of birth, and 11 received only hepatitis B vaccine at 2–7 days.
‡ The study provided lamivudine to both mothers and infants for 1 week, postnatally.
across the studies. Sample size was found to be a Effect of preventive measures in infants of HBeAg-
source of heterogeneity in the risk of mother-to-child positive mothers
transmission without prophylaxis from HBeAg-positive Among the infants of HBeAg-positive mothers, the
mothers (Table 2 and Figure S1). The risk was lower in pooled risk of mother-to-child transmission in those
studies with larger sample sizes (N > 10: 20.7%, 95% who received hepatitis B vaccine at >1 week (36.6%, 95%
CI: 7.1–38.1%) than those that included few infants CI: 8.3–69.9%) and those with a timely birth dose vac-
(N ≤ 10: 88.4%, 95% CI: 33.4–100%, P = 0.009). The cine (32.4%, 95% CI: 10.2–58.3%) did not differ from
transmission risk did not differ according to the sero- the transmission risk without any prevention (38.3%,
logical assay methods or the ‘region’ in HBeAg-positive 95% CI: 7.0–74.4%, P = 0.9 and 0.8 respectively; Fig-
mothers. In infants of HBeAg-negative mothers, those ure 2). In infants of women co-infected with HIV, the
in East Africa tended to have higher risk (7.5%, 95% pooled risk was higher when their mothers had received
CI: 1.7–15.9%) than in West Africa (0.5%, 95% CI: anti-viral therapy during pregnancy (47.0%, 95% CI:
0.0–4.3%, P = 0.05) (Figure S2). 21.8–72.8%) than when both mothers and infants
No.
Author, infected/No.
year assessed ES (95% CI)
No intervention
Marinier, 1985 3/15 0.20 (0.07, 0.45)
Greenfield, 1986 1/2 0.50 (0.09, 0.91)
Grathwohl, 1992 1/1 1.00 (0.21, 1.00)
Roingeard, 1993 2/2 1.00 (0.34, 1.00)
Menendez, 1999 3/14 0.21 (0.08, 0.48)
Subtotal (I^2 = 50.59%, P = 0.09) 0.38 (0.07, 0.74)
received anti-viral agents in the first week after birth The pooled risk was 0.7% (total sample size n = 11, 95%
(29.4%, 95% CI: 13.3–53.1%, P = 0.002). CI: 0.0–24.1%; Figure 4).
Effect of preventive measures in infants of HBeAg- Number of infants infected at birth each year
negative mothers Of 177 115 000 infants born in sub-Saharan Africa over
Compared to infants of HBeAg-negative mothers with- the period 2010–2015, an estimated 1 836 250 (1.0%)
out any prophylaxis, the pooled mother-to-child trans- are perinatally infected with HBV; annually, 367 250
mission risk was significantly lower in those who infants are infected at birth in sub-Saharan Africa. Esti-
received hepatitis B vaccine at >1 week (0.0%, 95% CI: mates by country are presented in Table S3.
0.0–0.0%, P = 0.01) and those with a timely birth dose
vaccine (0.0%, 95% CI: 0.0–0.0%, P = 0.02; Figure 3). DISCUSSION
Among infants of women co-infected with HIV, the In this meta-analysis, we found that: (i) in the absence
pooled risk was higher in the studies of pre-natal mater- of any preventive measures the pooled transmission risk
nal anti-viral therapy (5.1%, 95% CI: 0.0–29.8%) than in from HBeAg-positive mothers in sub-Saharan Africa was
immediate post-natal anti-viral therapy in mothers and significantly lower than those reported in Asia (38.3% vs.
infants (0.0%, 95% CI: 0.0–13.3%), although the differ- 70-90%); (ii) the transmission risk from HBeAg-negative
ence was not significant (P = 0.1). mothers in sub-Saharan Africa was similar to Asian esti-
mates (4.8% vs. 5–30%); (iii) the administration of hep-
Risk of mother-to-child transmission from mothers atitis B vaccine at birth and at >1 week was associated
with occult HBV with a reduction in transmission risk from HBeAg-nega-
Two studies assessed the risk of mother-to-child trans- tive mothers but not HBeAg-positive mothers and (iv)
mission from mothers with occult HBV infection. In the transmission risk from HBeAg-positive pregnant
both studies, the mothers were also infected with HIV. women co-infected with HIV was substantial even when
No.
Author, infected/No.
year assessed ES (95% CI)
No intervention
Marinier, 1985 0/62 0.00 (0.00, 0.06)
Greenfield, 1986 7/49 0.14 (0.07, 0.27)
Tsega, 1988 1/20 0.05 (0.01, 0.24)
Roingeard, 1993 2/19 0.11 (0.03, 0.31)
Menendez, 1999 1/34 0.03 (0.01, 0.15)
Subtotal (I^2 = 71.47%, P = 0.01) 0.05 (0.00, 0.13)
Table 2 | Potential sources of heterogeneity in the estimates of mother-to-child transmission risk (without any
intervention) by maternal HBeAg status
mothers received anti-viral therapy during pregnancy or of chronic carriers were perinatally infected in Asia while
when infants received anti-viral prophylaxis. in sub-Saharan Africa the vast majority (90%) were
Despite a similarly high prevalence of HBsAg between infected horizontally during childhood and only a
Asia and sub-Saharan Africa, the major mode of trans- minority (10%) had acquired through mother-to-child
mission differs. In the pre-vaccine era, an estimated 40% transmission.42 Indeed, we estimated here that 1% of
No.
Author, infected/No.
year assessed ES (95% CI)
Occult HBV
Table 3 | Risk of mother-to-child transmission by maternal HBeAg and type of preventive measure in Asia and in
sub-Saharan Africa
Timely birth dose vaccine & HBIG 5–10% No data <0.5% 0% 48–50
Timely birth dose vaccine & HBIG & <2%† No data No data No data 51, 52
newborns in sub-Saharan Africa are infected at birth, There are several explanations for this apparent geo-
whereas this has been reported as 3–5% in Asia.12 graphical difference in transmission risk. First, among
The incidence of mother-to-child transmission in an mothers positive for HBeAg, HBV DNA levels may be
area is determined by three factors: the prevalence of lower in sub-Saharan Africa than in Asia. However, previ-
infectious mothers, the risk of transmission from an ous African studies do not support this; median HBV
infectious mother to her infant, and the coverage of DNA levels in HBeAg-positive pregnant women
preventive services. The lower incidence of mother-to- (106–8 IU/mL)43, 44 were similar to those in Asia.45 Sec-
child transmission in sub-Saharan Africa has generally ond, circulating HBV genotypes differ between Africa and
been attributed to a lower prevalence of HBeAg in Asia, and certain genotypes are known to increase the risk
HBsAg-positive African mothers.3 However, we found of mother-to-child transmission (e.g. genotype C com-
that the transmission risk from HBeAg-positive mothers pared to B in Asia).46 Although genotype was not exam-
in the absence of prophylaxis was significantly lower in ined except for one study, we found a significant difference
sub-Saharan Africa (38.3%) than in Asia (70–90%), and in transmission risk from HBeAg-negative mothers
the risk from HBeAg-negative mothers lays within the between West and East Africa. This may be related with a
lower range of the Asian estimates (4.8% vs. 5–30%). variation in predominant genotype: E in West and A in
Although the number of studies in this review was rela- East Africa.9 Third, obstetric complications are known to
tively small, studies with larger sample sizes showed be associated with a transmission from infectious moth-
even smaller transmission risk from HBeAg-positive ers.47 These complications may be more likely to result in
women (20.7%), supporting a robustness of this geo- child deaths in sub-Saharan Africa due to insufficient
graphical difference in the risk of mother-to-child neonatal care, which may lead to a reduction in the num-
transmission. ber of HBV-infected infants aged at 3–12 months.
In Asia, the efficacy of the birth dose vaccine without Our study provided a unique opportunity to estimate
HBIG compared to no prophylaxis has been firmly estab- the risk of HBV mother-to-child transmission from HIV
lished; the risk reduces from 70–90% to 20% in infants of co-infected women treated with anti-virals during preg-
HBeAg-positive mothers, and from 5–30% to <0.5% in nancy but without neonatal immunoprophylaxis, and
those born to HBeAg-negative mothers (Table 3).48, 49 found a substantial transmission risk (47.0% from
However, we found that transmission risk from HBeAg- HBeAg-positive, and 5.1% from HBeAg-negative moth-
positive mothers in sub-Saharan Africa was similar ers). We could not assess the efficacy of this strategy
between the three varied vaccination groups: no vaccine (anti-virals during pregnancy without birth dose vac-
(38.3%), delayed vaccine at >1 week (36.6%), and timely cine), because there was no control group of infants born
birth dose vaccine (32.4%). The only controlled trial for to HIV co-infected mothers without any prophylaxis.
the prevention of mother-to-child transmission in Africa Another study of HIV co-infected mothers did not pro-
(which was included in our pooled analysis) showed a vide any anti-viral therapy during pregnancy, but admin-
substantial transmission risk from HBeAg-positive moth- istered anti-retrovirals, including lamivudine, to both
ers in the timely birth dose group (37.5%).34 We do not mothers and infants in the first week after delivery to
know whether this worrisome finding represents a gen- prevent mother-to-child transmission of HIV through
uine lack of efficacy of timely birth dose vaccine in Africa, breastfeeding.39 Interestingly, the risk of HBV transmis-
or if this is related to the quality of vaccines that were used sion in this study was significantly lower than the studies
in this particular study (e.g. loss of potency due to inad- which gave anti-virals during pregnancy, suggesting that
vertent freezing of the vaccines within the cold chain). anti-viral prophylaxis to neonates, a strategy similar to
Until further data on the efficacy of timely birth dose vac- pre-exposure prophylaxis of infants against HIV, might
cine becomes available in Africa, it is safer to assume that be another option to prevent mother-to-child transmis-
the birth dose vaccine alone may not be sufficient to pre- sion of HBV.
vent transmission from HBeAg-positive mothers in sub- There are few limitations in our study. First, we
Saharan Africa. could not examine the transmission risk by maternal
In contrast, among infants born to HBeAg-negative HBV DNA levels, which are more accurate than
mothers, the pooled transmission risk was reduced from HBeAg to predict the risk of mother-to-child transmis-
4.8% without prophylaxis to zero in those who received sion.53 This is because only few included studies
hepatitis B vaccine at birth or at >1 week after birth. assessed maternal HBV DNA levels, and these studies
One possible explanation for this reduction is that the used different assay methods with a varying limit of
5% of infants infected in the absence of any intervention detection. Nevertheless, our estimates based on HBeAg
might have been contaminated after birth through early are still highly useful in sub-Saharan Africa as the
horizontal transmission. As this review included infants access to HBV DNA assays is often restricted to a
aged up to 12 months, it is possible that some early hor- sophisticated laboratory in large urban centres.54 More-
izontal transmission has been captured; this horizontal over, our analysis clearly demonstrated that maternal
transmission might have been eliminated when a vaccine HBeAg is a perfectly sensitive marker (sensitiv-
was given at birth or even later. ity = 100%) to predict HBV transmission to their chil-
In Asia, additional administration of HBIG to the dren despite timely administration of birth dose
birth dose vaccine has been shown to further reduce vaccine. Second, the efficacy of preventive measures
the transmission risk from HBeAg-positive mothers; should be best estimated by pooling a risk ratio from
from 20% with only the birth dose vaccine, to 5–10% randomised controlled trials. However, we assessed the
when this was combined with HBIG (Table 3).48, 50 efficacy of timely birth dose and delayed hepatitis B
Moreover, to eliminate the risk of immunoprophylaxis vaccine by comparing pooled transmission risk of vacci-
failure, anti-viral therapy has been administered to nated cohorts with that of nonvaccinated cohorts,
highly viraemic women during pregnancy in addition because none of the included studies had a control
to hepatitis B vaccine and HBIG at birth, with a pro- group except one.34 Care must be taken to interpret
ven efficacy.51, 52 In contrast, in sub-Saharan Africa, these results because of their vulnerability to confound-
only one study27 provided HBIG, and anti-viral treat- ing. Finally, there were two papers that did not men-
ment was only administered to pregnant women co- tion whether or not there was administration of
infected with HIV, primarily to prevent mother-to-child hepatitis B vaccine in the study cohort and were con-
transmission of HIV. sidered as ‘received vaccination’ based on the national
REFERENCES
1. Schweitzer A, Horn J, Mikolajczyk RT, infection: epidemiology and vaccination. hepatocellular carcinoma in chronic
Krause G, Ott JJ. Estimations of Epidemiol Rev 2006; 28: 112–25. carriers of hepatitis B virus: a case-
worldwide prevalence of chronic 5. Hyams KC. Risks of chronicity control study in The Gambia. Liver Int
hepatitis B virus infection: a following acute hepatitis B virus 2015; 35: 2318–26.
systematic review of data published infection: a review. Clin Infect Dis 1995; 9. Shimakawa Y, Lemoine M, Njai HF,
between 1965 and 2013. Lancet 2015; 20: 992–1000. et al. Natural history of chronic HBV
6736: 1–10. 6. Chang M-H. Natural history and infection in West Africa: a longitudinal
2. Cowie BC, MacLachlan JH. The global clinical management of chronic population-based study from The
burden of liver disease attributable to hepatitis B virus infection in children. Gambia. Gut 2015; doi:10.1136/gutjnl-
hepatitis B, hepatitis C, and alcohol: Hepatol Int 2008; 2(suppl. 1): S28–36. 2015-309892 [Epub ahead of print].
increasing mortality, differing causes. 7. Shimakawa Y, Yan H-J, Tsuchiya N, 10. WHO. Combating Hepatitis B and C to
Hepatology 2013; 58: 218A–9A. Bottomley C, Hall AJ. Association of Reach Elimination by 2030. Geneva,
3. Kiire CF. The epidemiology and early age at establishment of chronic Switzerland: WHO, 2016.
prophylaxis of hepatitis B in sub- hepatitis B infection with persistent 11. Mast EE, Margolis HS, Fiore AE, et al.
Saharan Africa: a view from tropical viral replication, liver cirrhosis and A comprehensive immunization
and subtropical Africa. Gut 1996; 38 hepatocellular carcinoma: a systematic strategy to eliminate transmission of
(suppl. 2): S5–12. review. PLoS ONE 2013; 8: e69430. hepatitis B virus infection in the United
4. Shepard CW, Simard EP, Finelli L, 8. Shimakawa Y, Lemoine M, Bottomley States: recommendations of the
Fiore AE, Bell BP. Hepatitis B virus C, et al. Birth order and risk of Advisory Committee on Immunization
Practices (ACIP) part 1: immunization 24. Ott JJ, Stevens GA, Wiersma ST. The 35. Ilboudo D, Simpore J, Ouermi D, et al.
of infants, children, and adolescents. risk of perinatal hepatitis B virus Towards the complete eradication of
MMWR Recomm Rep 2005; 16: 1–31. transmission: hepatitis B e antigen mother-to-child HIV/HBV coinfection
12. WHO. Preventing Mother-to-Child (HBeAg) prevalence estimates for all at Saint Camille Medical Centre in
Transmission of Hepatitis B: Operational world regions. BMC Infect Dis 2012; 12: Burkina Faso, Africa. Brazilian J Infect
Field Guidelines for Delivery of the Birth 131. Dis 2010; 14: 219–24.
Dose of Hepatitis B Vaccine. Geneva, 25. United Nations, Population Division, 36. Greenfield C, Osidiana V, Karayiannis
Switzerland: WHO, 2006. Department of Economic and Social P, et al. Perinatal transmission of
13. Visvanathan K, Dusheiko G, Giles M, Affairs. World Population Prospects: hepatitis B virus in Kenya: its relation
et al. Managing HBV in pregnancy. The 2015 Revision, 2015. Available at: to the presence of serum HBV-DNA
Prevention, prophylaxis, treatment and http://esa.un.org/unpd/wpp/Download/ and anti-HBe in the mother. J Med
follow-up: position paper produced by Standard/Fertility/ (accessed 22 January Virol 1986; 19: 135–42.
Australian, UK and New Zealand key 2016). 37. Tsega E, Tsega M, Mengesha B,
opinion leaders. Gut 2016; 65: 340–50. 26. WHO & UNICEF. WHO-UNICEF Nordenfelt E, Hansson BG, Lindberg J.
14. Okada K, Kamiyama I, Inomata M, estimates of vaccine coverage, 2015. Transmission of hepatitis B virus
Imai M, Miyakawa Y, Mayumi M. E Available at: http://apps.who.int/ infection in Ethiopia with emphasis on
antigen and anti-E in the serum of immunization_monitoring/ the importance of vertical transmission.
asymptomatic carrier mothers as globalsummary/timeseries/ Int J Epidemiol 1988; 17: 874–9.
indicators of positive and negative tswucoveragehepb3.html (accessed 22 38. Menendez C, Sanchez-Tapias JM,
transmission of hepatitis B virus to January 2016). Kahigwa E, et al. Prevalence and
their infants. N Engl J Med 1976; 294: 27. Onakewhor J, Charurat M, Matthew O, mother-to-infant transmission of
746–9. Osagie E, Asemota M, Omoigberale A. hepatitis viruses B, C, and E in
15. Beasley RP, Trepo C, Stevens CE, Immunologic pattern of hepatitis B Southern Tanzania. J Med Virol 1999;
Szmuness W. The e antigen and infection among exposed and non- 58: 215–20.
vertical transmission of hepatitis B exposed babies in a PMTCT program 39. Chasela CS, Kourtis AP, Wall P, et al.
surface antigen. Am J Epidemiol 1977; in low resource setting: does every Hepatitis B virus infection among HIV-
105: 94–8. exposed newborn require 200 IU of infected pregnant women in Malawi
16. Stevens CE, Neurath RA, Beasley RP, hepatitis B immunoglobulin? J Vaccines and transmission to infants. J Hepatol
Szmuness W. HBeAg and anti-hbe Vaccin 2013; 4: 207. 2014; 60: 508–14.
detection by radioimmunoassay: 28. Shimakawa Y, Bottomley C, Njie R, 40. Pirillo MF, Scarcella P, Andreotti M,
correlation with vertical transmission of Mendy M. The association between et al. Hepatitis B virus mother-to-child
hepatitis B virus in Taiwan. J Med Virol maternal hepatitis B e antigen status, as transmission among HIV-infected
1979; 3: 237–41. a proxy for perinatal transmission, and women receiving lamivudine-containing
17. Liberati A, Altman DG, Tetzlaff J, et al. the risk of hepatitis B e antigenaemia in antiretroviral regimens during
The PRISMA statement for reporting Gambian children. BMC Public Health pregnancy and breastfeeding. J Viral
systematic reviews and meta-analyses of 2014; 14: 532. Hepat 2015; 22: 289–96.
studies that evaluate health care 29. Yvonnet B, Digoutte P, Denis F, Correa 41. Hoffmann CJ, Mashabela F, Cohn S,
interventions: explanation and P. [Prevention of mother-to-child et al. Maternal hepatitis B and infant
elaboration. J Clin Epidemiol 2009; 62: transmission of hepatitis B virus by infection among pregnant women living
e1–34. vaccination of the newborn]. Med Afr with HIV in South Africa. J Int AIDS
18. Weinbaum CM, Williams I, Mast EE, Noire 1982; 29: 721–31. Soc 2014; 17: 18871. eCollection 2014.
et al. Recommendations for 30. Marinier E, Barrois V, Larouze B, et al. 42. Edmunds WJ, Medley GF, Nokes DJ,
identification and public health Lack of perinatal transmission of O’Callaghan CJ, Whittle HC, Hall AJ.
management of persons with chronic hepatitis B virus infection in Senegal, Epidemiological patterns of hepatitis B
hepatitis B virus infection. MMWR West Africa. J Pediatr 1985; 106: 843–9. virus (HBV) in highly endemic areas.
Recomm Rep 2008; 8: 1–20. 31. The Gambia Hepatitis Study Group. Epidemiol Infect 1996; 117: 313–25.
19. Altman DG. Systematic reviews of Hepatitis B vaccine in the expanded 43. Andersson MI, Maponga TG, Ijaz S,
evaluations of prognostic variables. In: programme of immunisation: the et al. The epidemiology of hepatitis B
Egger M, Smith GD, Altman DG, eds. gambian experience. Lancet 1989; 1: virus infection in HIV-infected and
Systematic Reviews in Health Care: 1057–60. HIV-uninfected pregnant women in the
Meta-Analysis in Context, 2nd ed. 32. Grathwohl J, Ndumbe P, Leke R, Uy A, Western Cape, South Africa. Vaccine
London: BMJ Publishing Group, 2001; Gerlich WH, Repp R. [Risk of perinatal 2013; 31: 5579–84.
228–47. hepatitis B virus transmission in 44. Ducancelle A, Abgueguen P, Birguel J,
20. Nyaga VN, Arbyn M, Aerts M. mothers with low viral DNA levels et al. High endemicity and low
Metaprop : a Stata command to detectable by the polymerase chain molecular diversity of hepatitis B virus
perform meta-analysis of binomial data. reaction]. Monatsschr Kinderheilkd infections in pregnant women in a rural
Arch Public Health 2014; 72: 1–10. 1992; 140: 366–8. district of North Cameroon. PLoS ONE
21. Freeman MF, Tukey JW. 33. Roingeard P, Diouf A, Sankale JL, et al. 2013; 8: 10–2.
Transformations related to the angular Perinatal transmission of hepatitis B 45. Wen W-H, Chang M-H, Zhao L-L,
and the square root. Ann Math Stat virus in Senegal, west Africa. Viral et al. Mother-to-infant transmission of
1950; 21: 607–11. Immunol 1993; 6: 65–73. hepatitis B virus infection: significance
22. DerSimonian R, Laird N. Meta-analysis 34. Ekra D, Herbinger KH, Konate S, et al. of maternal viral load and strategies for
in clinical trials. Control Clin Trials A non-randomized vaccine effectiveness intervention. J Hepatol 2013; 59: 24–30.
1986; 7: 177–88. trial of accelerated infant hepatitis B 46. Wen WH, Chen HL, Ni YH, et al.
23. Higgins JPT, Thompson SG, Deeks JJ, immunization schedules with a first Secular trend of the viral genotype
Altman DG. Measuring inconsistency in dose at birth or age 6 weeks in Cote distribution in children with chronic
meta-analyses. BMJ 2003; 327: 557–60. d’Ivoire. Vaccine 2008; 26: 2753–61. hepatitis B virus infection after
universal infant immunization. Antimicrob Chemother 2015; 70: 396– perinatal hepatitis B virus exposure is
Hepatology 2011; 53: 429–36. 404. dependent on maternal virus load. J
47. Xu D-Z, Yan Y-P, Choi BCK, et al. 50. Chen H-L, Lin L-H, Hu F-C, et al. Infect Dis 1994; 170: 1418–23.
Risk factors and mechanism of Effects of maternal screening and 54. Lemoine M, Eholie S, Lacombe K.
transplacental transmission of hepatitis universal immunization to prevent Reducing the neglected burden of viral
B virus: a case-control study. J Med mother-to-infant transmission of HBV. hepatitis in Africa: strategies for a
Virol 2002; 67: 20–6. Gastroenterology 2012; 142: 773–81. e2. global approach. J Hepatol 2015; 62:
48. Lee C, Gong Y, Brok J, Boxall EH, 51. Chen HL, Lee CN, Chang CH, et al. 469–76.
Gluud C. Hepatitis B immunisation for Efficacy of maternal tenofovir disoproxil 55. UNAIDS. Fact Sheet 2015, 2015.
newborn infants of hepatitis B surface fumarate in interrupting mother-to- Available at: http://www.unaids.org/
antigen-positive mothers. Cochrane infant transmission of hepatitis B virus. sites/default/files/media_asset/
Database Syst Rev 2006; (2): CD004790. Hepatology 2015; 62: 375–86. 20150901_FactSheet_2015_en.pdf
49. Machaira M, Papaevangelou V, 52. Pan CQ, Duan Z, Dai E, et al. (accessed 1 April 2016).
Vouloumanou EK, Tansarli GS, Falagas Tenofovir to prevent hepatitis B 56. Shimakawa Y, Toure-Kane C, Mendy
ME. Hepatitis B vaccine alone or with transmission in mothers with high viral M, Thursz M, Lemoine M. Mother-to-
hepatitis B immunoglobulin in neonates load. N Engl J Med 2016; 374: 2324–34. child transmission of hepatitis B in sub-
of HBsAg+/HBeAg- mothers: a 53. Burk RD, Hwang L-Y, Ho GYF, Saharan Africa. Lancet Infect Dis 2016;
systematic review and meta-analysis. J Shafritz DA, Beasley RP. Outcome of 16: 19–20.