Background
In the Netherlands, all pregnant women are offered screening for Human Immunodeficiency Virus (HIV), syphilis and hepatitis B virus (HBV). Participation, including blood sample collection, is according to the opting-out-principle during the first midwife appointment in the first trimester of pregnancy. In case of positive test results, confirmatory tests are performed, and in case of confirmed positive test results women are referred to secondary care for treatment and/or other preventive measures [
1]. In the Netherlands, standard care for sexually transmitted infections includes partner services but execution and data collection of testing and treatment for partners is not embedded in the screening programme. In 2011, an evaluation of the Dutch antenatal screening programme for HIV, HBV and syphilis was published for the years 2006–2008, which concluded that the programme is effective in detecting HIV, HBV and syphilis among pregnant women and in preventing vertical transmission [
2]. It was estimated that 5–10 HIV, 10 syphilis and 50–75 HBV cases in newborns had been prevented annually due to screening [
2].
In 2014, the World Health Organisation (WHO) established a list of validation criteria to facilitate efforts of elimination of mother-to-child transmission (EMTCT) of HIV and syphilis, which were updated in 2017 [
3]. These criteria consist of a global minimum of three impact- and six process indicators, of which an overview is given in Table
1. The WHO has not yet established elaborate guidance on EMTCT of HBV, but it is addressed in the WHO Global health sector strategy on viral hepatitis 2016–2021 [
4]. It states that prevention of MTCT is a core intervention area in ending the hepatitis epidemic, mainly through timely HBV birth-dose vaccination, antenatal testing, and the use of antiviral drugs. In this global strategy, the WHO also set targets to reach EMTCT of HBV in 2030. These additional targets are also included in Table
1.
Table 1
Minimum required WHO indicator criteria for validation of EMTCT of HIV and syphilis, and additional criteria for hepatitis B
Shared indicators HIV and syphilis |
ANC coverage (at least one visit) of ≥95% | 99.2%a | 99.7%a | Yes |
HIV indicators |
Coverage of pregnant women who know their HIV status of ≥95% | 99.2%a | 99.7%a | Yes |
Antiretroviral (ARV) coverage of HIV positive pregnant women of ≥95% (women with suppressed viral load at delivery) | 96% (84%)b | 100% (97%)b | Yes |
Case rate of new paediatric HIV infection due to mother-to-child transmission (MTCT) of HIV of ≤50 cases per 100,000 live births | 0.57b | 0.00b | Yes |
MTCT rate of HIV of < 5% in breastfeeding population OR MTCT rate of < 2% in non-breastfeeding populations | 0.75%b | Yes |
Syphilis indicators |
Coverage of syphilis testing of pregnant women of ≥95% | 99.2%a | 99.7%a | Yes |
Treatment of syphilis-seropositive pregnant women ≥95% (% registered referred to care) | 72.6%a | 70.5%a | Unknown |
Case rate of congenital syphilis ≤50 cases per 100,000 live births | 0.00c-1.14d | 0.00c-0.59d | Presumably |
Hepatitis B indicators |
Hepatitis B virus vaccination: childhood vaccine coverage (third dose coverage) ≥90% | 93.1%e | 92.2%e | Yes |
Hepatitis B virus birth-dose vaccination coverage or other approach to prevent mother-to-child transmission ≥90% | Birth-dose vaccination | 99.4%e | 99.4%e | Yes |
HBIg admission at birth | 99.8%f | 99.8%a |
HBsAg prevalence among children ≤0.1% | 0.0%g | 0.0%g | Yes |
HIV, syphilis and hepatitis B are low prevalent diseases in the Netherlands; hence, EMTCT is likely to be achievable. To assess whether the Netherlands can indeed meet the WHO criteria for EMTCT of HIV and syphilis, we re-evaluated the Dutch antenatal screening programme on HIV and syphilis for the years 2009–2015. To review EMTCT of HBV we used the additional WHO global strategy criteria. In the current study only the minimum indicators are addressed. To achieve official validation, additional data collection and analyses will have to be described in a full country report, such as extensive case studies, evaluation of additional criteria, and an assessment of data- and laboratory quality. Detailed information can be found in the WHO guidance document [
3]. This study can hence be seen as a first exploration of the feasibility of EMTCT, and will provide insight in future efforts that are needed to strengthen EMTCT efforts and surveillance of EMTCT in the Netherlands.
Methods
Participation data and outcomes of the antenatal screening are registered in an electronic database (Praeventis) in the Netherlands. As every year a process evaluation is performed on these data, information on screening uptake and general screening outcomes have been well-reported in the past [
5]. Screening outcomes are reported based on laboratory confirmed positive tests. Screening coverage should be calculated by dividing the number of women screened by the number of pregnancies in a given year. Only an estimate can be provided, as the number of pregnancies in a given year is uncertain. It is estimated by the number of children born alive in the Netherlands half a year later (as reported by the Central Bureau of Statistics (CBS)), assuming that children are born 6 months after screening. As pregnancies may result in multiple births, correction of the number of life births is needed. In the Netherlands, the yearly number of twins and triplets is also registered by the CBS, and results in a correction of minus 1.5–1.7% of the number of children born alive. Another correction is needed for loss of pregnancies. In 2012, we investigated the number of pregnancy losses after the first screening among RhD-negative women, as these women participated to a pilot screening of fetal Rhesus-D-typing in week 27 of pregnancy. Loss of pregnancy was registered in 3.8%. For the denominator of the estimate of screening coverage, the number of children born alive was thus corrected by − 1.5-1.7% and by + 3.8% to get an estimate of the number of pregnancies. The numerator of screening coverage, i.e. the number of women screened in a year, was corrected for double registration of the same pregnancy and for screening of women living abroad, as their children are not in the Dutch CBS statistics of lifeborns. We used additional data from the Praeventis database to describe age and ethnicity of pregnant women with positive test results for HIV, syphilis or HBV. Ethnicity was defined using country of birth of the pregnant women; only first generation migrants were considered non-Dutch. Countries were grouped into regions of origin using the UNSTAT list of geographic regions. In Praeventis, not all additional data needed to evaluate the WHO criteria are available. Therefore, information from other data sources was included, as described below.
HIV
To gain additional information on pregnancies among HIV positive women in the Netherlands, we obtained data from the AIDS therapy evaluation in the Netherlands (ATHENA) cohort, which is maintained by the ‘Stichting HIV monitoring’ (SHM). The ATHENA national observational HIV cohort includes data from all persons with HIV in care in the Netherlands in any of the 26 HIV treatment centres. Pregnancies among HIV positive women are also registered in ATHENA. To be able to compare annual numbers of HIV positive pregnant women between the ATHENA cohort and Praeventis data (which is based on screening date), we estimated the date of screening for the women in the ATHENA cohort. For women who were newly diagnosed with HIV during pregnancy, we used the date of diagnosis as the date of screening. For the other women, we used the date of their 12th week of pregnancy, derived from the estimated due date, as this is the average moment of screening. If the estimated due date was not known, which was mostly the case for pregnancies that did not result in delivery, we used the date of termination of the pregnancy.
Data on children with congenital HIV infection were also obtained from the ATHENA database. We included children who were infected through vertical transmission and who were born in the Netherlands in 2009–2015. To calculate the MTCT rate of HIV, we used the number of children born with HIV divided by the number of registered births given by HIV positive women from the ATHENA database.
Syphilis
To gain insight in the number of children born with syphilis, two alternative data sources were used. First, we used data from the CIb-IDS (Centre for Infectious diseases research, Diagnostics and Screening) laboratory at the National Institute for Public Health and the Environment (RIVM), where immunoglobulin M diagnostics are performed on children aged < 1 years who are suspected of having congenital syphilis. Second, we acquired data for 2009–2015 from the Perinatal Registry of the Netherlands (Perined), which includes congenital syphilis diagnoses reported by paediatricians. Perined contains – among others – diagnoses from all registered children admitted to a paediatric ward within 28 days after birth. In 2015, 86% of Dutch paediatric practices reported data to Perined [
6].
The WHO has established a global surveillance case definition for congenital syphilis, to be used with the validation criteria: 1) A stillbirth, live birth, or foetal loss at > 20 weeks of gestation or > 500 g to a syphilis-seropositive mother without adequate syphilis treatment, or: 2) a live birth, stillbirth or child aged < 2 years born to a woman with positive syphilis serology or with unknown serostatus, and with laboratory and/or radiographic and/or clinical evidence of syphilis infection (regardless of the timing or adequacy of maternal treatment) [
3]. While there is registration of pregnancy outcomes in Perined, cause of death is not registered for stillbirths in the Netherlands, nor is there a comprehensive registration of syphilis serostatus of the mother in case of a stillbirth. Therefore, we do not have insight in congenital syphilis resulting in stillbirths, and are not able to follow the exact WHO case definition. In this study, all registered cases of congenital syphilis include only live born children.
Hepatitis B
Acute and chronic HBV infections are notifiable in the Netherlands and are registered in the National Register for Notifiable Diseases (Osiris) at the RIVM. Reason for testing is not well registered in this database, so we could not distinguish newly notified infections from pregnant women. To obtain information on HBV infections among children, we collected data on children aged < 2 years old who were born in the Netherlands and for whom vertical transmission was specified as the most likely route of transmission from the Osiris database. Data on HBV vaccination coverage are reported in the annual RIVM report on the national immunisation programme [
7] and in the prenatal screening process evaluation [
5].
Discussion
The Netherlands meets all of the minimum WHO criteria for EMTCT of HIV and HBV. For syphilis, one of the three criteria was met. The prevalence of HIV, syphilis and HBV is very low and stable among pregnant women, and only very few children with congenital infections have been born in the Netherlands in the past years. In 2015 there were 0 reported cases for all three infections.
The estimated coverage of the antenatal screening was very high, with > 99% of pregnant women screened each year. We therefore have reliable information on the prevalence of HIV, syphilis and HBV among pregnant women in the Netherlands. The use of multiple additional data sources enabled us to gain more insight into infections among both pregnant women and children born in the Netherlands, and allowed for assessment of almost all of the WHO EMTCT criteria. However, there are some limitations to address.
First, comparison of the different data sources could only be done on group level instead of individual level. Privacy regulations did not allow for linkage of records across different databases. Therefore we could not gain detailed insight into the overlap and completeness of the different data sources. We could not distinguish between women who were familiar with HIV/HBV infection and women newly diagnosed during screening, and had therefore limited information about the women that were (or were not) linked to care. Second, no data were available of women or children who were not in care. This could have caused a slight underestimation of disease prevalence and/or incidence. Last, there was no information available on voluntary abortions in most included data sources. For the screening coverage calculations, we do not expect voluntary abortions to be of significant influence. Especially since most abortions occur before screening.
The main goal of this study was to assess whether the Netherlands meets the WHO criteria for EMTCT of HIV, syphilis and HBV. For HIV, all data needed to evaluate the WHO criteria was available and of good quality. We can therefore be certain that the criteria for HIV are met in the Netherlands. However, we lack insight in the exact proportion of newly diagnosed women at screening, in whether the women refusing HIV testing at screening might be mostly women who are already diagnosed with HIV, and in whether positively screened women are all in care at an HIV treatment centre, as databases could not be linked. As policy in the Netherlands is to guide all newly diagnosed persons with HIV directly to care in one of the 26 officially recognised HIV treatment centres and start immediate treatment [
10], we expect this is well organised for pregnant women. This information can provide important insights into the efficiency of the present health care system, and it would contribute to a more precise estimation of the number of pregnant women newly diagnosed with HIV in the Netherlands.
Also for HBV all data needed were available. Vaccination coverage is well-known due to the national vaccination registration. Furthermore, HBV is a notifiable disease in the Netherlands, and therefore all HBV infections among children should be reported. A decrease in yearly HBV infections among children was seen between 2009 and 2015 (from 2 in 2009 to 0 since 2012). This decrease could be due to increased antiretroviral treatment during pregnancy, which is an effective method to prevent breakthrough infections [
11], but it could also be a consequence of changes in the child healthcare system. In 2011, the responsibility for serological testing of children born to HBV positive mothers was transferred to the child’s general practitioner, instead of being nationally coordinated. An evaluation showed that the implementation of this process was suboptimal, and that only half of the children born in 2012 from HBV positive mothers who were included in the evaluation received serological testing [
12]. This could have caused breakthrough infections to be missed.
For syphilis, validation of WHO criteria was more difficult due to incompleteness of data. Syphilis is not a notifiable disease in the Netherlands. Although the number of syphilis-positive pregnant women could be extracted from Praeventis, this number is subject to uncertainty due to the aforementioned discrepancy in registration of positive syphilis tests versus clinically relevant (active) syphilis infections. This also hampers the collection of information about accurate treatment of pregnant women with an active syphilis infection. Furthermore, reliable data about congenital syphilis infections were lacking. The two data sources that were used reported different numbers of yearly infections. The RIVM CIb-IDS reference laboratory should receive samples from all children with suspected congenital syphilis. However, in Perined there were more cases of congenital syphilis reported than by the CIb-IDS laboratory. It is unknown which number is true, and whether this difference is seen because of registration errors in Perined or because not all cases of suspected congenital syphilis are being forwarded to the CIb-IDS reference laboratory. Finally, no systematic information about cause of death for stillborn children is collected in the Netherlands, causing a lack of insight into the number of pregnancies that do not result in delivery due to syphilis infection resulting in the inability to evaluate the WHO definition of congenital syphilis.
Until now, ten countries have officially validated EMTCT for both HIV and syphilis. Furthermore, EMTCT of HIV has been reached in Armenia and EMTCT of syphilis was validated in Moldova [
3]. No Western-European countries have thus far officially validated EMTCT of HIV, syphilis and/or HBV. However, two evaluation studies were performed in the United Kingdom, which show that the UK meet the minimum WHO criteria for syphilis [
13], and that HIV MTCT rates are also very low [
14]. Even though in most European countries prevalence of HIV, syphilis and HBV will be relatively low among pregnant women and children, averting also the last cases of vertical transmission is very important.
In case the Netherlands decides to apply for official validation, a validation committee and -team will have to be established, and additional indicators will have to be evaluated such as more in-depth assessments of data and laboratory quality. Furthermore, the WHO recommends for countries with low HIV and syphilis prevalence (such as the Netherlands) to include extensive case studies in the validation report. We believe it will be possible to collect more in-depth data needed for validation in the Netherlands, but this will require great efforts and extensive collaborations between (local) governments, research (and surveillance) organisations, health-care providers and laboratories. Given that an official validation process would therefore be very costly, it is the question whether a low-prevalence country such as the Netherlands should engage in it. On the other hand, evaluating the screening programme and its outcomes in light of the WHO criteria provided us with new recommendations for improvement, mainly concerning limitations in data availability and the surveillance of syphilis. In a progress evaluation, the WHO acknowledges that in most countries case reporting and surveillance of (congenital) syphilis (particularly stillbirths) are less developed than that for HIV [
15], which is also the case in the Netherlands. Improvements need to be made especially to the current congenital syphilis surveillance, as at this moment there is no insight in the exact number of children with congenital syphilis per year. The United States and the United Kingdom have reported recent increases in the number of congenital syphilis, indicating that also in developed countries surveillance and EMTCT efforts remain a necessity [
16,
17]. Validation of EMTCT of HIV, syphilis and hepatitis B in the Netherlands might thus seem a disproportionate exercise, but this first explorative evaluation showed that even in a low-prevalence country with an extensive screening programme there are still improvements to be made.
Acknowledgements
The authors thank Kim Vos for her help with interpreting the data and providing additional information on the Dutch antenatal screening, and Petra Oomen for her help with and provision of Praeventis data.