SLIDE 1 Do different HDSS surveillance systems result in different quality of pregnancy outcome data? Akuze, J; Baschieri, A; Kerber, K; Gordeev, V; Waiswa, P; Blencowe, H; Kwesiga, D; Floyd, S.; Lawn, J. On behalf of the INDEPTH Network–ENAP metrics study team ABSTRACT An approximate of 2.6 million stillbirths, 2.7 million neonatal deaths and 303,000 women die from pregnancy and childbirth related complications every year. Low and Middle income countries in Asia and Sub-Saharan Africa bare the largest burden. The INDEPTH Network has various potential roles to inform the global maternal and newborn agenda, and improving population-based birth outcome measurement but faces many challenges. Initial steps are in place to improve the measurement of pregnancies and their outcomes. Five HDSS sites with a population more than 30000, annual stillbirth rates and neonatal mortality rates greater than 15 per 1000 live births, high quality surveillance for birth outcomes, presence of expertise related to maternal, newborn health and stillbirth and evidence of co-funding were selected to participate in the Every Newborn Action Plan Improvement Measurement Roadmap collaborative study. Different HDSS sites have alternative systems for collecting of pregnancies and outcomes data with different intervals (surveillance rounds), different types of village informants, and different modalities for linking health facility and HDSS data, these components could affect measurement or pregnancies and their outcomes. Two out of five sites were able to link health facility and HDSS data. All sites had records
- f birthweight from either; community measurement, or mother’s health cards and or health facility
- records. Unlike birthweight, gestational age was not easily estimated by all sites.
[220 words]
SLIDE 2
INTRODUCTION An approximate of 2.6 million stillbirths, 2.7 million neonatal deaths and 303,000 women die from pregnancy and childbirth related complications every year (1). Low and Middle income countries in Asia and Sub-Saharan Africa bare the largest burden(2). Whilst substantial strides in halving child and maternal mortality have been made although a slower progress is seen for reduction of stillbirths and neonatal deaths. Between 2000 and 2015 the global annual rate reduction was observed for stillbirths (2.0%), neonatal mortality (3.1%), maternal mortality (3.0%) and under-5-year old mortality (4.5%) (3). As a result, the Every Newborn Action Plan (ENAP) to end preventable death was launched in June 2014 (4). The plan was based on evidence published in The Lancet Every Newborn series (5) and on consultations with member states, organizations and individuals culminating into a World Health Assembly resolution. The ENAP is closely linked to the Ending Preventable Maternal Mortality targets and strategies (6). The ENAP set targets for the Sustainable Development Goal (SDG) era, of 12 or fewer newborn deaths per 1,000 live births and 12 or fewer stillbirths per 1,000 total births in every country by 2030 (4, 7). The neonatal mortality target is a sub-target under SDG 3 (7). If practices remain the same, these targets will not be achieved. The INDEPTH Network, a consortium of Health and Demographic Surveillance System (HDSS) sites in 20 low and middle income countries, spread over 3 continents was founded in 1998, to coordinate and bring together a number of already existing HDSS sites (8). The Network is currently made up of 46 member centres, and 53 HDSS field sites who are full members. HDSS sites that are associate members and non- members are encouraged to join the network as full members (8). All centres collect valuable data through the HDSS sites that provide a picture of the health status of the communities through data collection from entire communities over extended time periods (9). Most of these sites are in Sub-Saharan Africa (SSA) and South Asia. All centres collect data mandatory on their populations (including demographic, socio-economic, births, deaths, migrations, environmental, and verbal autopsy) prospectively at least once a year. Besides the mandatory data, the HDSS sites also conduct other research including; HIV, genetics, immunization, clinical trials among others (9). The INDEPTH Network works through Working Groups, one of which is the Maternal, Newborn and Child Working Group (MNCWG) whose main goal is to support coordinated, multi-site generation of evidence to inform policy and programmes for maternal, newborn and child health and survival in low income countries (10). The INDEPTH MWCNG has various potential roles to inform the global maternal and newborn health agenda (8). Since pregnancies, births, stillbirths, deaths, gestational age and birth weight are monitored routinely within HDSS sites, the sites are well placed to assist in improving the tracking of various ENAP and SDG indicators including stillbirth rate, neonatal mortality rate, low birth weight rate, and preterm birth rate. The Network can improve the measurement of outcomes, test innovations to improve outcomes around the time of birth and engage champions who will steer the course of reproductive, maternal and newborn health as we move towards achieving the SDG 3 and ENAP targets. Although the INDEPTH Network has potential for improving population-based pregnancy outcome measurement, it faces several challenges including:
SLIDE 3 a) Many proposed ENAP birth outcome impact indicators including stillbirth, birth weight and gestational age are not currently included within some HDSS surveillance systems or when available, the quality of the data is highly variable across sites; b) Most of the sites lack experience in conducting maternal and newborn health tracking; c) Some sites also lack the necessary funding to facilitate this nature of specialized research (11). The INDEPTH Network’s HDSS sites follow vital events in a defined population prospectively and they are able to document pregnancy outcome data on a regular basis (8). Different HDSS systems use various methods to record vital events. Some HDSS sites have a target system to follow pregnancies and pregnancy outcomes. Considering this unique platform, a collaborative study “ENAP measurement improvement roadmap” or “ENAP metrics” is being conducted in five HDSS sites for a period of 3 years (2016 – 2018) within the INDEPTH Network. The ENAP metrics collaborative study aims to assess whether the “pregnancy history” approach is more accurate for capture of pregnancies and their outcomes compared to standard Demographic Health Survey version seven (DHS-7) “birth history” and how it affects the length of data collection in standardized surveys. This research paper is part of the wider ENAP metrics study and it aims to compare different HDSS systems through assessing how they (HDSS systems) differ in the quality of pregnancy outcome data DATA AND METHODS A Request for Applications (RFA) was published within the INDEPTH Network to all 53 HDSS sites by the INDEPTH Network secretariat about the ENAP Measurement Improvement Roadmap and fourteen proposals were received. The received proposals underwent an expert review through an internal review process spearheaded by the London School of Hygiene and Tropical Medicine (LSHTM), Makerere University School of Public Health (MakSPH) and INDEPTH Network secretariat. All sites that expressed their interest through returning applications for the RFA received personalised feedback about the status
- f their application (12). We selected five HDSS sites for the ENAP metrics project that had a system in
place to follow pregnancy outcomes. These HDSS sites included; Bandim HDSS, Dabat HDSS, Iganga- Mayuge HDSS, Kintampo HDSS and Matlab HDSS. The five successful sites were selected basing on: Presence of a HDSS total population more than 30000. Annual SBR and NMR greater than 15 per 1000 live births. High quality surveillance for birth outcomes including neonatal deaths and stillbirths. Present expertise related to maternal, newborn health and stillbirths from the team members of the applying HDSS and evidence of co-funding in the HDSS submitted estimated budgets. Within the ENAP metrics study, we reviewed the RFA’s of the five HDSS sites to establish the status of maternal and newborn indicators and HDSS systems. From these reviews, the HDSS sites were encouraged to suggest enhancements that are within the HDSS’s ethical approval and mandate, these improvements were to help steer the HDSS towards improved capture of pregnancy outcomes. Some of the Initial steps taken to improve birth outcome indicators included; encouraging HDSS sites to include additional important questions in their surveillance instruments, conducting refresher trainings for their current staff, introducing pregnancy registration and testing among others.
SLIDE 4 Members from each of the selected HDSS sites were invited to the ENAP metrics protocol design workshop in June 2016 in Kampala, during this workshop formative research about each HDSS system was conducted(12). The formative research during this workshop had the following components. Review of baseline data collected from the HDSS: The ENAP Metrics research team had some prior information obtained from the RFA applications submitted by the five HDSS sites, before the meeting, the HDSS teams received a template for which they were required to update or confirm the numbers pre-filled from the RFA. The HDSS teams received this template prior to the protocol design workshop and summaries of these baseline data were compiled. Review of tools and setup surveillance systems of the HDSS: The HDSS teams provided an
- verview of their surveillance systems during the workshop. The purpose of this was to identify
any gaps in their data collection. Discussion groups with the HDSS teams to propose enhancements: During the protocol development workshop, the teams developed context specific approaches to improve the quality
- f pregnancy and outcome data and data collection. These HDSS enhancements were conducted
for one year from January to December 2016. The results presented in this paper are from the data submitted by the five HDSS sites, discussions with representatives from the HDSS sites and from summaries of review of the HDSS surveillance systems, tools and setup. The statistics presented in these results are averaged absolute numbers across (3-5) years and rates with their 95% confidence intervals (CI). ANALYSIS AND PRELIMINARY RESULTS The five HDSS sites are located in Africa (four) and Asia (one); Bandim HDSS in Guinea Bissau, Dabat HDSS in Ethiopia, Iganga-Mayuge HDSS in Uganda, Kintampo HDSS in Ghana and Matlab HDSS in Bangladesh (see figure 1 and table 1). The HDSS sites started at different points in time with the oldest site as Matlab, which started in 1966, and the youngest HDSS site as Iganga-Mayuge, which started in 2004. All sites are majorly located in the rural areas; some sites like Bandim and Iganga-Mayuge have an urban and peri- urban area respectively. The average population over 3-5 years across all the five sites was 143,034 with a range of 69,468 to 230,185 people. The average number of households across all the five sites was 30,648 households within the last 3-5 years, this ranged from 16,000 to 53,226 households. Iganga- Mayuge and Matlab HDSS sites had the smallest and largest number of households respectively. The number of live births averaged for 3-5 years was 3,789 for all sites, this ranged from 1,320 to 5,790 with Dabat and Bandim HDSS sites with the smallest and largest recorded number of live births within the last 3-5 years (Table 1). The five HDSS sites have different established modalities of capturing pregnancies and their outcomes (see table 2). The quality of the pregnancy tracking is dependent on the frequency of rounds and the type
- f the HDSS population data entry, the characteristics of the key informants, proportion of health facility
births and or whether the HDSS site was able to link health facility data and HDSS surveillance data. The sites with a higher number of frequency of surveillance rounds, electronic data collection and entry, larger proportion of health facility deliveries and ability to link health facility and HDSS data reported higher numbers of livebirths, stillbirths and neonatal deaths. Different HDSS sites have alternative systems for collecting of pregnancy outcomes. These data are recorded at different intervals; some have a quarterly or biannually data collection. In addition, some
SLIDE 5
record these events through different types of village informant. In addition, different sites have different modalities for linking health facility data with community surveillance (Table 2). The sites with more frequent rounds were either using tablets for electronic data collection or piloting the system, unlike their counterparts with fewer rounds that were using paper based data collection and entry. From the discussions groups with the HDSS representatives, each site gave an overview of their HDSS including how pregnancy surveillance is conducted, their data linking system and how they measure births and weights and gestational age. These results are presented in table 3. Pregnancy surveillance All sites except Matlab utilizes either field assistants or field workers and scouts to conduct the routine pregnancy visits. The occurrence or the pregnancy surveillance visits varies from twice a month to every six months, Matlab and Bandim urban area reported the most frequent visits while Kintampo and Dabat reported the list frequent visits. All sites reported that when births are not recorded, they are recorded retrospectively, although in Matlab, the mentioned that this is a very rare occurrence (table 3). Matlab and Bandim also reported that they use electronic data collection and entry systems. Data linkage Except Bandim and Matlab HDSS sites who reported that some of their health facility data is linked to the HDSS, other HDSS sites including; Iganga-Mayuge, Dabat and Kintampo have not yet established any systems in place to link health facility and community data collected from the routine rounds. All sites mentioned that all individuals are assigned a unique identification number which can possibly or is used during data linking (table 3). Measuring birthweight and gestational age All the five HDSS sites reported that birthweight data are recorded within their databases. This data was made available through various means, including; taken at the health facility records, recorded from the mother’s health card reports, and measured in the communities by HDSS staff. Dabat, Iganga-Mayuge and Kintampo HDSS sites reported that the respondents (women) are asked a question about the perceived size of the baby at birth during the surveillance rounds, in addition to the birthweight question. Unlike the birthweight data that is more readily available or easily obtained at the HDSS sites, the gestational age data recorded by only two sites Matlab and Bandim. Matlab and Kintampo HDSS sites reported to utilize the data of delivery and last menstrual period recorded to estimate the gestational age (table 3). Enhancements and challenges During the Kampala protocol development workshop, the site representatives suggested some context specific enhancements in order to improve the capture of all pregnancy outcomes (Table 3). All sites except Matlab HDSS suggested possible enhancements, these included; improve field worker performance, ensuring that definitions for indicators are consistent with international indicators, giving incentives to women that encourage them to register their pregnancies with the HDSS systems like health education, counselling among others and early capture of pregnancies.
SLIDE 6 All sites reported challenges expected to hinder the HDSS enhancements towards harmonization of their data across sites, these included; misclassification of stillbirth and neonatal deaths based on recall data, missing data, different meta-data structures of their databases which makes standardization of tools and systems complex and missing indicators of the core ENAP metrics indicators within the already existing HDSS tools. Discussion In this paper, we set out to establish how the HDSS systems in five countries differ in quality of pregnancy
- utcome data. We found that each HDSS site was a unique system with a number of variations, suggesting
varied capacities for measurement or pregnancies and their outcomes. However, we found some similarities, for instance the retrospective registration of missed pregnancies which was similar across four sites except Matlab HDSS, which stated that this was a rare event (only 2% of all pregnancies were missed by the HDSS system). Additionally, the registration of births if the pregnancy was not previously registered was raised by all the five HDSS sites. This particular practice may explain the theories on the low stillbirth rates across the five HDSS sites, since it is unlikely that you can register what does not exist. This finding was consistent with the findings by Kadobera (2017) that at Iganga Mayuge HDSS it is possible that if an event (still birth and newborn death) occurred in between surveillance rounds, such a pregnancy outcome may never be registered (11). We found that the HDSS sites have varying capacities of pregnancy and outcome surveillance, the sites that had better established systems for tracking and reporting pregnancies and their outcomes reported slightly higher rates and more events (live births, stillbirths and newborn deaths). Having a good pregnancy surveillance system including; notification processes for an event, improved mode of data collection and entry, and woman main respondent and the prospective nature of the HDSS renders an HDSS site superior to others in terms of processing and cleaning of data. However, major challenges to the integrated multi-site tracking for maternal and newborn research still exist, for instance lack of consistency in the definition of indicators, lack of standardized data collection tools and databases. Further work in harmonizing tools used for measuring and reporting birth outcomes is required. The 2014 Every Newborn Lancet series and a study by Maxon et al also emphasize the need to have improved surveillance for pregnancies and outcomes (13, 14) since it is unlikely that in low resource setting we will have self-sustaining civil and vital registration systems in the near future, yet we have public health issues to solve with the available data (15). In our study, we found that three out of the five HDSS sites had no previous experience with electronic data collection and did not have established data linkage between health facilities data and HDSS surveillance data, although this was an item on their priority lists towards improved surveillance. We also found that all sites assign unique identifiers to all individuals who are residents within their catchment
- areas. HDSS and health-facility data linkage can be established or improved where it is existent through
use of unique identifiers that are already available in the HDSS systems. The health facilities in low resource settings do not usually assign unique identifiers but with support from the HDSS sites, this is not a problem. In addition, add-on identifiers like other name, household head names, woman’s age, family name among others can improve data linkage. Our findings are consistent to with Kanudula et al’s methodological paper on record linkage for assessing the uptake of health services in resource- constrained settings. Their findings show that matching statistics improve when additional identifiers like another household member’s first name are included to the available set of identifier variables (16).
SLIDE 7 Research implications include; further work needed to test how the INDEPTH-ENAP HDSS enhancements have improved the measurement of pregnancies and their outcomes, and testing health facility-HDSS data linkage in site that have not yet embarked on this. Fortunately, all the sites are now running their update rounds after a year of implementing the enhancements, so we will know their effects sooner than later. Secondly, two out of the five INDEPTH-ENAP sites have already established health facility-HDSS data linkage systems and lessons from them are invaluable for their counterparts that have not. This work will lead to improved measurement of pregnancies and their outcomes, throughout the HDSS surveillance systems. Strengths of this study are that it includes five sites from different contexts and settings, with variant
- experiences. Secondly, in addition to reviewing the data and tools submitted, we also had consultative
meetings with the HDSS representatives to consolidate our findings. However, the study is not immune to limitations including; the sites provided only aggregated data. For us to be able to standardize the computations across sites we needed to run the same algothrims on the HDSS raw data which we did not have access to. Therefore, we do not include any background variables that could give a surround picture
- f the sites from the analysis. However, we asked the sites to update and validate these results. Finally,
the sites have different methods of collecting data and different variables and there is currently no standard definition of indicators across sites. Finally in this Conclusion While all the HDSS sites have the ability to track pregnancies and their outcomes, room for improvements in terms of innovations around early identification of pregnancies, data linkage and electronic data collection or entry systems are needed. With these enhancements in place, a general improvement on the maternal and newborn indicators will be noticed across all the INDEPTH Network HDSS sites. Furthermore, funding towards innovations with the surveillance systems, capacity building and developing analytical skills of the HDSS scientist in areas of maternal and newborn death, will bring us a step closer towards establishing uniformity and standardization of tools and indicators across the INDEPTH Network. We recommend more frequent surveillance rounds within the HDSS sites and HDSS- health facility data linkage as these have proved that the more events are captured by the HDSS. Further research to test the structural effects of the INDEPTH-ENAP metrics study towards improved capture of pregnancies and their outcomes are required, these will be possible once the HDSS surveillance data is available Acknowledgment The authors acknowledge the support of the representatives from the five HDSS sites that are participated in the INDEPTH-ENAP study protocol design workshop in Kampala Uganda at Hotel African in June 2016. The INDEPTH Network operational secretariat, the MakSPH that hosts the technical secretariat of the MNCHWG and the core funder The Children’s Investment Fund Foundation (CIFF) through the LSHTM.
SLIDE 8 REFERENCES 1. Measurement and Accountability for Results in Health. The Roadmap for Health Measurement and Accountability. World Bank Group, USAID, World Health Organization,, 2015. 2. Organization WH. Every Newborn: An action plan to end preventable deaths: World Health Organization; 2014. 3. Lawn JE, Blencowe H, Waiswa P, Amouzou A, Mathers C, Hogan D, et al. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet (London, England). 2016;387(10018):587-603. 4. UNICEF, The World Health Organization. Every Newborn: An action plan to end preventable newborn deaths www.everynewborn.org. 2014. 5. Lawn JE, Blencowe H, Oza S, You D, Lee AC, Waiswa P, et al. Every Newborn: progress, priorities, and potential beyond survival. Lancet (London, England). 2014;384(9938):189-205. 6. World Health Organization. Strategies towards ending preventable maternal mortality (EPMM)
- 2015. Available from: http://who.int/reproductivehealth/topics/maternal_perinatal/epmm/en/.
7.
- UN. Sustainable Development Goal 3: Ensure healthy lives and promote well-being for all at all
ages: United Nations; 2015 [5/2/2016]. Available from: http://www.un.org/sustainabledevelopment/health/. 8. Sankoh O, Byass P. The INDEPTH Network: filling vital gaps in global epidemiology. International journal of epidemiology. 2012;41(3):579-88. 9.
- INDEPTH. INDEPTH Network 2016. Available from: http://www.indepth-network.org/about-us.
10.
- INDEPTH. Maternal Newborn Health Working Group 2010. Available from: http://www.indepth-
network.org/groups/working-groups/maternal-and-newborn-health. 11. Kadobera D, Waiswa P, Peterson S, Blencowe H, Lawn J, Kerber K, et al. Comparing performance
- f methods used to identify pregnant women, pregnancy outcomes, and child mortality in the Iganga-
Mayuge Health and Demographic Surveillance Site, Uganda. Global health action. 2017;10(1):1356641. 12. metrics E. EVERY NEWBORN ACTION PLAN METRICS: ENAP & INDEPTH Research Protocol Design
- Workshop. London School of Hygiene and Tropical Medicine, INDEPTH Network, Makerere University
School of Public Health, 2016 15th - 17th June, 2016. Report No. 13. Lancet T. Every Newborn. The Lancet; 2014. 14. Moxon SG, Ruysen H, Kerber KJ, Amouzou A, Fournier S, Grove J, et al. Count every newborn; a measurement improvement roadmap for coverage data. BMC pregnancy and childbirth. 2015;15 Suppl 2:S8. 15. Fottrell E. Dying to count: mortality surveillance in resource-poor settings. Global health action. 2009;2. 16. Kabudula CW, Clark BD, Gomez-Olive FX, Tollman S, Menken J, Reniers G. The promise of record linkage for assessing the uptake of health services in resource constrained settings: a pilot study from South Africa. BMC medical research methodology. 2014;14:71.
SLIDE 9 Table 1 Demographic characteristics (data based on 3-5 years average)
Iganga-Mayuge Bandim Matlab Kintampo Dabat Country Uganda Guine Bissau Bangladesh Ghana Ethiopia HDSS start year 2004 1978 1966 1994 1996 Area – km2 155 184 7,162 Location Iganga and Mayuge districts, approximately, 120km east
capital, Kampala along Kenya- Uganda highway Guinea-Bissau, covering rural and urban Matlab Upazila, in Chandpur district, 55km southeast
capital, Dhaka Brong Ahafo region Gondor, Amhara regional state, 821 km northwest
Addis Ababa and 75km north Population 83,000 180,000 230,185 152,519 69,468 Households 16,000 36,000 53,226 32,000 16,016 Live births 2264 5790 4863 4710 1320 Stillbirths 45 297 92 86 34 SBR 19 95%CI(14.7 – 24.2) 49 95%CI(43.6 – 54.9) 19 95%CI(15.3 – 23.2) 18 95%CI(14.4 – 22.3) 25 95%CI(17.3 – 34.9) Neonatal deaths 118 204 104 95 53 NMR 52 95%CI(43.3 – 62.1) 35 95%CI(30.5 – 40.1) 21 95%CI(17.1 – 25.4) 20 95%CI(16.2 – 24.4) 40 95%CI(30.2 – 52.2) Total Fertility Rate 4.3 2.6 4.1 3.8 Note: stillbirth rate calculated on live births not pregnancies, expected number of stillbirths should be similar to number of neonatal deaths.
Table 2: Overview of the HDSS system
Dabat Kintampo Matlab Bandim Iganga-Mayuge Frequency of rounds 2/year 1/year 6/year Urban: Monthly Rural: 2/year 2/year Informants/ scouts Local guides report within 48 hours Community key informants Recently started using community key informants 64 Community based “scouts” and VHTs Incentives for reporting 83% female enumerators Each woman asked pregnancy status Urine test; 100% female enumerators Each woman asked pregnancy status Age of informants 15+ if married 10+ in some surveys 15+ if married 15+ 15+ Frequency of re- census the area Every 7 years Last census 2003 8 years or more Every 2 years Each update round
SLIDE 10 Facility births 17% 61% 69% (Intervention area: 87%; Comparison Area: 50%) Urban: 65% Rural: 39% 64% Links to facility Pilot study ongoing Not currently Matlab hospital only (17% of births) In national hospital, not in rural Not currently Data entry Piloting tablet Paper-based Galaxy Tab Tablet planned in rural in 2017 Paper-based
Table 3: Overview of the HDSS surveillance.
SITE Bandim (Guinea Bissau) Dabat (Ethiopia) Iganga-Mayuge (Uganda) Kintampo (Ghana) Matlab (Bangladesh) Pregnancy surveillance Who carries out routine surveillance visits? Field assistants Data collected by trained and full-time field workers who have completed high school and are living in the district. Supervisors present Uses field assistants, Community “Scouts” & recently Village Health Team (VHT) Field workers / supervisors Female Community Health Research Workers Who is allowed to be the primary respondent during visits? The individuals, parents, neighbours Mothers Must be adult usual resident (18yrs +yrs) Household heads or any responsible adults i.e. > 15 years (including pregnant women) within the household visited Adult female members (usually wife of household head) How often are routine surveillance visits made? Urban: monthly Rural: Every six months (in 3 northern regions- 2 months) Update routine surveillance every six months for pregnancy
2 rounds per year to record births and deaths From March 2016,
Bi-monthly
SLIDE 11 SITE Bandim (Guinea Bissau) Dabat (Ethiopia) Iganga-Mayuge (Uganda) Kintampo (Ghana) Matlab (Bangladesh) migration, death, and marital status How are vital events notified and followed up? Through the routine visits (in three regions – +community key informants) Vital events for births and deaths are notified within 24/48 hours Community health workers and scouts Routine HDSS surveillance Community Key Informants (CKIs) record pregnancies and births in their communities. This information is extracted by HDSS supervisors who visit every 2 weeks to check for new events. Household visit What happens when a birth is captured when the pregnancy was not previously recorded? For all children registered– a birth form is filled out Pregnancy
form will be filled retrospectively and pregnancy outcome will also be registered The pregnancy is retrospectively registered, after which a pregnancy termination form will be completed together with a birth form if the woman is a resident
Birth is recorded (rare event) Data Linking System How can pregnancy, surveillance and Through Individual ID Individual ID from residency table and mother ID from Using location ID and individual unique ID Linked by identifiers such as individual ID, compound ID By mother’s ID
SLIDE 12 SITE Bandim (Guinea Bissau) Dabat (Ethiopia) Iganga-Mayuge (Uganda) Kintampo (Ghana) Matlab (Bangladesh)
events) data be linked? pregnancy
and household ID and date of event in related data tables What percentage of births is in a facility? Urban: 65% Rural: 39% 17% 64% 61.1% 69% [MCH-FP service area: 87%, Comparison area: 50%] Is facility birth data linked to HDSS? National maternity ward and Bandim health center in Bissau (Women are identified) Facility birth data were not linked to HDSS previously but now in the pilot stage No No, but this can be done if required Only Matlab hospital data Measuring Birth weight Any data on birth weight? Collected in urban area if available for those born at a health facility Currently have started maternal and child health surveillance Yes Yes, from January 2015 Yes if available Is the data from facility records only or is weight taken at home by HDSS staff? Collected by HDSS staff from vaccination card/ANC card/other document It is planned to record weight data both from facility and at home by HDSS staff Done at facility. Since 2013 we rely on health card Field workers record weight information from children’s weighing cards Facility records only. In a few cases it is from the mother’s report Is it captured in routine surveillance visits? Yes (birth form or specific studies) Plan to capture in routine surveillance Yes Yes Are women asked about perceived size of baby at birth? No Starting from 2014 women are asked Yes, once during pregnancy history survey Yes No Is the data available as a subset of the routine surveillance or sub studies? Routine data Data available as subset of the routine surveillance Yes, part of routine surveillance Not applicable
SLIDE 13 SITE Bandim (Guinea Bissau) Dabat (Ethiopia) Iganga-Mayuge (Uganda) Kintampo (Ghana) Matlab (Bangladesh) Measuring Gestational Age Is there any data on gestational age at delivery? Yes, but of questionable reliability In recent years pregnancy outcome tool addressed gestational age of mothers at delivery No No variable for this, but can be estimated by other variables (date of delivery and LMP) Yes If so, what data exist? Response provided by the mother/parents on month of pregnancy termination (urban) Gestational age of mothers at delivery Date of delivery and LMP LMP date and the date of delivery Is it captured in routine surveillance visits or linked to facility data? Captured in routine visits (urban) Captured in routine surveillance and linked to facility data No (a pilot using 1 or 2 facilities is being discussed) Routine surveillance Routine surveillance visits Is the data available as a subset of the routine surveillance or sub studies? Both (sub studies in Bissau) Data available as subset of the routine surveillance Part of surveillance Yes, part of routine surveillance Routine surveillance visits Enhancements and chanllenges Specific thoughts/questions on enhancing surveillance and making sure we capture all outcomes? Urban area: improve field workers’ performance in terms of pregnancy registration; linking to HF data (gestational age) Rural: Calculation of gestational age based
Definitions and consistency of data especially still births Measurement tools be used to measure GA Gaps in coverage definitions, Provision of services to pregnant women e.g. health education, counselling and checking blood pressure at home Early pregnancy [first trimester ultrasonography for dating of pregnancy] to augment LMP – is that practised in any of the
yes, how do
SLIDE 14 SITE Bandim (Guinea Bissau) Dabat (Ethiopia) Iganga-Mayuge (Uganda) Kintampo (Ghana) Matlab (Bangladesh) registration of pregnancy and
validation & feasibility testing for HMIS use Improve how to capture birth
without pregnancy
they go about it? Early capture of pregnancies could enhance capture of still- births, as we could trace each woman who has a record of pregnancy for its outcome. Overcoming socio-cultural resistance to stillbirth registration Challenges expected in harmonisation of data across sites Misclassification of stillbirth or neonatal death based on recall (dates, born alive or not, etc) Using the same kind
measure parameters Data missingness Different metadata structures across
need to provide standardised data descriptions across HDSS sites. Information not available for core ENAP indicators Information not available for additional indicator
SLIDE 15
Figure 1: Map of Africa and Asia continents showing the locations of the INDEPTH-ENAP HDSS sites
