An assessment of tne successes and shortfalls of the national birth defects databases and an improved data collection method of the databases.

Abstract

The Constitution of South Africa (SA) together with the National Health Act (63 of 1977) govern the provision of health services to the residents of the country. With a three-tier system of governance consisting of national, provincial and local (district) government, each tier functions autonomously, though in unison. The National Health Act outlines the health system whilst specifying services per level of governance. In this document, medical genetic services, amongst others, are included as a health issue that needs to be addressed as part of the functions of the National Department of Health (NDOH). At this level, these services form part of the Maternal and Child Health services. Although neglected, medical genetics services are important for the prevention and management of congenital disorders (CDs) in the community. These services are implemented through the development and implementation of policy guidelines. Data on CDs form the basis for policy development, decision making and planning for services. Without empirical data, services for this vulnerable group of individuals, cannot be adequately provided. Collection of CD data was initiated in 1980, with multiple surveillance systems available in the country by the early 1990s. One system in particular (Birth Defects Surveillance System-BDSS) was successful, with its data (from 1992-2004) being submitted to the International Clearinghouse for Birth Defects Monitoring Systems (ICBDMS), whose functions include the exchange of CD birth prevalence among member countries and the promotion of epidemiologic studies. In 2006, the NDOH developed the standardized birth defect notification tool (BDNT), with the intention of substituting all existing CD surveillance systems with one notification tool and system for the entire country. The primary objective of this study was to measure the effectiveness of this system, taking into account the challenges experienced in the reporting period. This was done by analyzing the data and comparing it to other surveillance systems globally and locally. The secondary objective was to develop an improved surveillance system for the country. Data from the BDNT was analyzed for a nine year period, 2006 until 2015. Analyzed data included the number of CDs reported per year, per province and per district. CDs of all pregnancy outcomes were included e.g. live births, stillbirths, terminated pregnancies and miscarriages. Birth prevalence was calculated based on aetiology, (whether the CD was genetically or environmentally determined), and per priority condition (This list of priority conditions was defined by the NDOH in their 2001 Birth defects policy guidelines. The number of priority CDs reported by doctors versus nurses was also determined). Priority conditions are, Down syndrome, fetal alcohol syndrome, neural tube defects, oculocutaneous albinism orofacial clefts and talipes equinovarus. A total of 17 761 notifications were received from across SA, of which 16 395 (92.13%) were CD notifications and 1 366 (7.69%) were zero notifications (these are notification tools that were completed by the health facility in months when no CDs were identified by the health viii facility). Compliance was erratic with KwaZulu-Natal province reporting the most CDs, n=9 732 (59.36%), and Western Cape province reporting the least, n=389 (2.37 %). KwaZulu-Natal province’s success is largely attributed to the good medical genetics services that were administered by Professor William Winship while he was alive. Overall, the districts where medical genetics facilities are located reported more CDs. When compared to modelled estimates, the BDNT surveillance system showed an underreporting rate of 98%. Malformations accounted for most of the reported CDs with a birth prevalence of 1.02 per 1 000 live births. Birth prevalence for CDs categorized by aetiology were: single gene disorders 0.07 per 1 000 live births; chromosomal disorders 0.13 per 1 000 live births; multifactorial disorders 0.09 per 1 000 live births; CDs caused by Rh (rhesus factor) incompatibility 0.00 (0.0032) per 1 000 live births and 0.01 per 1 000 live births for CDs caused by teratogens. Birth prevalence for each priority CD was as follows: Down syndrome 0.12 per 1 000 live births, fetal alcohol syndrome 0.01 per 1 000 live births, neural tube defects 0.09 per 1 000 live births, oculocutaneous albinism 0.03 per 1 000 live births, orofacial clefts 0.10 per 1 000 live births and talipes equinovarus 0.10 per 1 000 live births. Over half (57.80%) of all reported CD cases were diagnosed by nursing staff. Following analysis of data from the BDNT, a new surveillance system was developed containing the following factors: the types of CDs to be monitored, approaches to data collection, classification of collected data and the use of data received. Initially, the new CD surveillance system was going to be integrated into the national notifiable medical conditions surveillance system. In addition, an electronic system (with a backup paper-based notification system) was developed together with colleagues responsible for notifiable medical conditions surveillance and the National Institute for Communicable Diseases (NICD) which is yet to be piloted. Upon further research, certain elements were lacking in the system which could negatively impact upon implementation. As a way forward, certain considerations were identified for future implementation of a CD surveillance system. These were categorized into mandatory and elective factors. The former includes political commitment to CDs as a health issue, legislation prescribing CD services including surveillance, vital registration of CDs at birth and death, and NDOH facilitating the coordination of CD surveillance systems in the country. The latter includes the use of a district based approach to data collection with specific personnel identified to collect data using an electronic system. This study lays the foundation for national CD surveillance in SA. Various surveillance systems or patient registries are available, but none operate data at a national level. This study further identified the need for coordination between the different surveillance systems and/or patient registry data sets (e.g. non-governmental organisations and laboratories) which are not included in the BDNT. The national CD surveillance system could serve as a link between the various stakeholders (provinces, academic institutions, laboratories and non-governmental institutions), allowing each entity to have a system that is suitable to their needs while collating data from these systems. The CD surveillance system should also follow patients from the point of diagnosis to treatment/management and/or death, allowing for the true burden of CDs to be measured.