Background: In 2010, the World Health Assembly passed a resolution calling upon countries to prevent birth defects where possible. Though birth defects surveillance programs are an important source of information to guide implementation and evaluation of preventive interventions, many countries that shoulder the largest burden of birth defects do not have surveillance programs. This paper shares the results of a hospital-based birth defects surveillance program in Uganda which, can be adopted by similar resource-limited countries. Methods: All informative births, including live births, stillbirths and spontaneous abortions; regardless of gestational age, delivered at four selected hospitals in Kampala from August 2015 to December 2017 were examined for birth defects. Demographic data were obtained by midwives through maternal interviews and review of hospital patient notes and entered in an electronic data collection tool. Identified birth defects were confirmed through bedside examination by a physician and review of photographs and a narrative description by a birth defects expert. Informative births (live, still and spontaneous abortions) with a confirmed birth defect were included in the numerator, while the total informative births (live, still and spontaneous abortions) were included in the denominator to estimate the prevalence of birth defects per 10,000 births. Results: The overall prevalence of birth defects was 66.2/10,000 births (95% CI 60.5-72.5). The most prevalent birth defects (per 10,000 births) were: Hypospadias, 23.4/10,000 (95% CI 18.9-28.9); Talipes equinovarus, 14.0/10,000 (95% CI 11.5-17.1) and Neural tube defects, 10.3/10,000 (95% CI 8.2-13.0). The least prevalent were: Microcephaly, 1.6/10,000 (95% CI 0.9-2.8); Microtia and Anotia, 1.6/10,000 (95% CI 0.9-2.8) and Imperforate anus, 2.0/10,000 (95% CI 1.2-3.4). Conclusion: A hospital-based surveillance project with active case ascertainment can generate reliable epidemiologic data about birth defects prevalence and can inform prevention policies and service provision needs in low and middle-income countries.
The birth defects surveillance system uses active case ascertainment and includes obtaining demographic and basic medical information for all births delivered at the participating hospitals. All newborns are examined for major external birth defects by trained midwives. A woman who delivers a baby with a major external birth defect is asked if photographs can be taken of her newborn to help with the diagnosis of the birth defect. Written informed consent is obtained before photographs are taken. If a photograph is not possible, midwives draw and write a detailed description of the defect. Surveillance data are obtained from multiples sources, which include review of patient medical records, interviewer administered questionnaires, and newborn physical examination findings. Figure 1 illustrates the surveillance system activity flow. Surveillance system activity flow All required ethics approvals were obtained as per the Uganda National Council for Science and technology (UNCST) guidelines [14]. This surveillance study was approved by the Joint Clinical Research Centre institutional review board/ethics committee and the US Centers for Disease Control and Prevention Institutional Review Board (IRB) (protocol # 6606.0). The surveillance was also approved by the Uganda National Council of Science and Technology (Ref: HS 1693), The surveillance system is being conducted at four hospitals, including one public/government hospital, Mulago National Referral Hospital and three faith-based private not-for-profit hospitals (Mengo hospital, St. Francis Hospital, Nsambya and Uganda Martyrs Hospital, Lubaga), in Kampala, Uganda. They were selected based on findings from a review of 2012 annual health data from the Ministry of Health, the 2012 annual hospital reports and the Uganda Demographic and Health Survey (UDHS) 2011. While the UDHS 2011 estimated 93% of births in Kampala were health-facility based, the Ministry of Health, 2012 annual health report and the 2012 annual hospital reports revealed 55% of the births in Kampala were at these four hospitals. The four hospitals included in this surveillance project have approximately 50,000 births annually, Mulago National Referral Hospital contributes 60.0% of births while Mengo Hospital, St. Francis Hospital, Nsambya and Uganda Martyrs Hospital, Lubaga contribute 12.0, 13.4 and 14.0% respectively. The time period for this surveillance project is approximately 4 years, during which we would expect to capture approximately 200,000 births. Assuming a birth defects prevalence range of 13.0 per 10,000 births for central nervous system defects to 87.0 per 10,000 births for musculoskeletal defects [4], we expect between 260 and 1740 newborns with each major external birth defect during this time period. All informative births (live, stillbirths and spontaneous abortions), regardless of gestational age, at the four hospitals are included in the birth defects surveillance system. Informative births are those in which the newborn is well formed enough to ascertain the presence or absence of an external birth defect. Birth defects must be diagnosed at birth, during the newborn hospitalization period, or before discharge from the hospital. If prenatal diagnosis of birth defects is available, confirmation must be done at birth. In Uganda, elective termination of pregnancies is not legal except when it preserves maternal life and with consent by two registered physicians. However, all informative spontaneous abortions are included regardless of gestational age. Births outside the four surveillance hospitals and un-informative macerated stillbirths are excluded. All live births and stillbirths are examined by a trained surveillance midwife within 2 h of birth or as soon as feasible, without interrupting the first breastfeeding or preparation for burial. A systematic examination that is “head to toe” and “front to back” is used for every live birth and stillbirth to identify major external birth defects. During this examination, standard measurements are collected, including weight, head circumference and body length. Examination of all newborns is conducted by the surveillance midwife in the presence of the mother and/or relative where possible. All care of live newborns with birth defects is provided through routine care by the hospitals, which includes referral to specialists when available. Data are collected by the surveillance midwife within 24 h after delivery using android-based tablets with paper forms as a back-up. All data collection forms were programmed using Open Data Kit (ODK), an open source data collection platform. All data entered on the tablet are encrypted and only completed forms are transcribed and transmitted to the main application server via internet. To ensure confidentiality, all tablets are password protected, and data are encrypted during transmission to protect it from unauthorized access. For every birth in the four participating hospitals, a surveillance form is completed by the midwife. Surveillance data include maternal demographic data such as age, tribe, address at the time of conception and current residence; brief maternal pregnancy history such as antenatal visit history, parity and history of previous birth defects; HIV sero-status and ART exposure; newborn characteristics such as sex, gestational age, anthropometric measures and presence or absence of birth defects; birth outcome such as live birth, stillbirth and spontaneous abortion. Each mother/newborn pair is assigned a unique study identification number that is generated automatically by the tablets. For multiple deliveries, a form is completed for each newborn. A mother who delivers a child with any major external birth defect is asked if photographs of her child can be taken. If she is willing, written informed consent is obtained and photographs are taken by the trained surveillance midwife using the Android-based tablet. Photographs are taken from several views, including a view of the entire fetus or newborn plus several focused views of the birth defect(s). Many major external birth defects are identified; however this surveillance focuses on birth defects of interest listed in Table 1. The birth defects of interest to the surveillance system a10th International Classification of Diseases modified by the Royal College of Paediatrics and Child Health adaptation bCNS Central nervous system In addition to taking photographs, the examining midwife writes a narrative description of the birth defect(s), detailing the location, size, appearance and other specific details necessary for an independent person who has not seen the infant to envision the birth defect(s) and make a diagnosis. The surveillance midwife then requests a study physician to do an independent bedside examination of any birth that she suspects has a birth defect. The study physician makes an independent diagnosis and prepares an additional independent narrative description. Both the surveillance midwife and the study physician assign diagnosis codes based on the 10th International Classification of Diseases modified by the Royal College of Paediatrics and Child Health adaptation [15], which are either pre-programed in the tablet or available for reference on the tablet. The photographs and the two narrative descriptions are reviewed by the study review team (co-principal investigator and program manager), who may modify the narrative description, diagnosis and diagnostic code when necessary. The final decision is sent to CDC for confirmation of final diagnosis and code assignment. In situations where mothers do not provide consent for photographs to be taken of their newborns, the surveillance midwives make illustrations of the birth defects, write detailed narrative descriptions of the birth defects, and photograph the illustrations. The data collection software (ODK) links the photographs of the infant or the illustration with the mother’s surveillance information. Quality control and assurance are addressed in several ways. To standardize study activities, all study staff were trained on the principles of Good Clinical Practice [16], the study protocol, and how to conduct the surveillance activities. In addition, standard operating procedures were developed to ensure systematic collection of data and reduce interpersonal and inter-site variability. Study data undergo three levels of quality control. Quality control level 1 combines use of real-time electronic and manual data checks. Quality control level 2 is a manual data check completed by study research assistants to ensure data completeness and validity. Quality control level 3 combines both manual and electronic reviews that are completed by the data managers, program manager and investigators. To ensure inclusion of all births, hospital delivery registers are reconciled with the information in the database on a regular basis. Data from births that may have been discharged from the hospital and not included by surveillance midwives are abstracted from the patient medical file and entered by surveillance research assistants. Quality assurance activities are regularly conducted by the program manager and CDC project monitors. The program manager regularly reviews 10.0% of randomly selected data collected in the past month to assess quality, completeness and regulatory compliance and implements corrective and preventive measures. These data include all newborns with birth defects, all cases and controls and a simple random sample of data from other participants not in the mentioned categories to add up to 10.0% data collected each month. Descriptive statistics of the population included in the surveillance system were generated, including maternal characteristics and infant characteristics. This includes the distribution of birth defects by maternal age, parity, maternal HIV sero-status, newborn sex, birth outcome, type of pregnancy, and mode of delivery. The birth defects prevalence at the four surveillance hospitals was calculated for each major birth defect by aggregating the number of birth defect cases as the numerator and the total number of informative births (live births, stillbirths and spontaneous abortions) at the surveillance hospitals as the denominator. An infant or fetus with multiple birth defects was counted as a separate case for each defect [17]. Prevalence was expressed per 10,000 informative births (live births, stillbirths and spontaneous abortions) using the following formula during a specific time period: The 95% confidence interval for each prevalence estimate was calculated using Wilson bounds [18].
N/A