Cohort Profile: Urban Health and Demographic Surveillance System in slums of Dhaka (North and South) and Gazipur City Corporations, Bangladesh

The Urban Health and Demographic Surveillance System (urban HDSS) in selected slums of Dhaka (North and South) and Gazipur City Corporations is a population-based cohort that initially aimed to monitor the primary health care services provided by non-government organizations and urban local bodies of the Local Government Division of Bangladesh. Now, the urban HDSS is monitoring and evaluating the impact of the comprehensive primary health care services provided by the newer Aalo clinics (model clinics). The baseline population census of 2015 covered about 120 000 people living in slightly more than 30 000 households, irrespective of age and sex. After the baseline population census, every household is visited every 3 months by assigned female fieldworkers, who register pregnancy outcomes (live birth, stillbirth, induced miscarriage and spontaneous miscarriage), deaths, migrations (in-migration, out-migration and internal movement), household splits/head changes, marriage/divorce and safe motherhood practices. Data on childhood immunization, acute respiratory infections, diarrhoea and breastfeeding practices were recently added for children aged under 2 years. Geographical coordinates of households and facilities are also collected. The attrition rate is 12% per year, and the cohort has increased to about 156 000 people. Internal and external collaborators are welcome to access the surveillance data, in accordance with icddr, b policies. The urban HDSS is a member of the INDEPTH Network. Bangladesh has experienced rapid urbanization in recent decades. The urban population was negligible (<5%) until the early 1970s but has since risen to more than 30%, and about one-third of the urban population of Bangladesh lives in slum settlements.1 The population of Bangladesh is projected to increase from 167 million in 2018 to roughly 185 million by 2030; during this period, the urban population will increase from about 61 million to nearly 85 million.2 This rapid urbanization in Bangladesh, largely due to in-migration of the rural poor, has led to overcrowding and the expansion of informal settlements (i.e. slums).3 The slum population is highly mobile and vulnerable to precarious economic and living conditions; all of which can negatively affect their health.4 The government of Bangladesh has been struggling to provide employment, housing, health services, education and other basic services to the urban poor. Traditionally, primary health care services are more structured in rural areas of Bangladesh than they are in urban areas. To improve the health of the poor in urban areas, the Local Government Division of the government of Bangladesh has been implementing consecutive phases of the Urban Primary Health Care Services Delivery Project since 1998. Comprehensive primary health care services are now being provided through the establishment of model clinics (Aalo clinics), supported by UNICEF/Sida. Community-level monitoring and evaluation are necessary to examine the impact of the health care services. In 2015, the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr, b) therefore established the Urban Health and Demographic Surveillance System (urban HDSS) in several urban slums of Dhaka (North and South) and Gazipur City Corporations. The urban HDSS has been collecting maternal and child demographic and health service data. These data help us understand levels and trends, as well as to monitor the impact of primary health care services and other intervention programmes. Future opportunities include applying the urban HDSS to quasi-experimental designs that would allow better monitoring of the prevalence of disease and conditions, as well as testing various intervention models, such as: an urban community clinic; a zero maternal, infant and child death model; and health financing schemes for the poor population. After extensive field visits and reviewing the Census of Slum Areas and Floating Population 2014,5 relatively large and stable slums were selected to comprise the surveillance area, as they were deemed to be more convenient for female fieldworkers in terms of travel time and security. About 50% of the slum population of Dhaka (North and South) and Gazipur City Corporations live in large slums (defined as a cluster of 100 or more households).3 For this reason, large slums were purposively selected for the establishment of the urban HDSS. In total, 13 slums in five locations (Korail, Mirpur, Dhalpur, Shayampur and Tongi) were selected (see map in Figure 1). Most of the surveillance slums had more than 100 households. As the slums are informal settlements, several events (evictions, conflicts and fire) commonly affect the slums’ area and population. Slums of Dhaka (North and South) and Gazipur City Corporations The initial round of urban HDSS data collection followed the completion of the baseline population and socioeconomic census in 2015. After the census, a team of 15 female fieldworkers visited each of the 31 577 households (with an average household size of 3.8 people) within the delineated area. However, a few non-slum households (<2 per 100) also exist in these areas, and these households were excluded. The cohort has since increased and now comprises a population of 155 994 in 40 643 households (Table 1). In total, the urban HDSS covers an area of 1.561 km2, with an average of 99 932 people per km2. The fieldworkers visited each household to interview the household head/spouse or caretakers. The fieldworkers explained the purpose of the surveillance system and guaranteed confidentiality of the information they would provide. Willingness to take part was expressed by signature or thumb impression. Informed consent remains valid until it is withdrawn. Refusal is rare (<1 per 10 000 households), and the community has always been supportive of icddr, b activities, as the people receive free treatment for diarrhoeal diseases from Dhaka hospital. Baseline population census and household socioeconomic data are usually taken from one respondent, but routine surveillance maternal and childhood data (conception, pregnancy outcome, maternity care) are always collected by interviewing the mother. Locations and current characteristics of the slums included in the Urban Health and Demographic Surveillance System For identification, two types of numbers are maintained for each household member: a Registration Identification Number (RID) and a Current Identification Number (CID). The RID is fixed, and the CID changes depending on the person’s current location. For births, the mother’s information is obtained from the preloaded database and a new identification number is assigned to the newborn; no identification number is assigned in the case of non-live birth. For in-migrants, a new identification number is assigned, and all baseline demographic and socioeconomic data are collected. In-migrants are always asked whether they were previously resident in the urban HDSS area, and the fieldworker tries to find the original RID number for those who were previous residents. Internal migrants are assigned their old RID in the new location, and only their CID changes to reflect their new location. If death or out-migration occurs, the person’s data are updated through the identification number. For marriage or divorce, one of the partners (husband or wife) must be a resident of the surveillance area and the relevant data are updated accordingly. In the event of a family split, a new household head is identified for the new partition and a new household identification number is assigned. For death or out-migration of the household head, a new household head is determined and the data are updated accordingly. Deaths are further investigated using the World Health Organization 2016 Verbal Autopsy questionnaire, with open narrative to determine the cause of death. Initially, all the data were collected using portable devices equipped with SQLite in the back end and Android Java in the front end. Later, the information and technology team of icddr, b developed online data collection software. The software for data collection was developed in house and includes consistency checks, including range checks and logical checks; however, some logical checks are performed in the office after loading and merging the data files. A team of field research assistants is responsible for monitoring the quality of the data. As part of the process, they observe and check about 2–3% of all interviews conducted by the fieldworkers. After receiving the data from the field, the computer programmer edits the data and updates the master database. After the baseline population and socioeconomic census, every household within the surveillance area is revisited and followed up 3-monthly. In each round, the female fieldworkers capture in-migration (along with registration of new migrant households), out-migration, internal movement, births, deaths and updated information for individuals. The data collection instruments for routine data remain unchanged during this period. Refusal to participate during follow-up and subsequent interviews is rare. The field research coordinator, in consultation with the three field research assistants, prepares a daily work plan for each of the female fieldworkers. The workload is balanced so that they can complete visits in their respective areas within 3 months. In general, each female fieldworker visits 40–45 households per day. The fieldworkers, with a few exceptions, possess a Bachelor’s degree. One of the criteria for selecting female fieldworkers for a particular area is the close proximity of the worker’s residence to the field site. After the baseline and socioeconomic census, data on births (live birth, induced miscarriage, spontaneous miscarriage and stillbirth), deaths, migrations (in-, out-, and internal movement), marriage/divorce and maternity care are collected from each household every 3 months; data on fertility regulation are collected from a sample of households once a year (Table 2). From 2021, child health-related information (Expanded Programme of Immunization; acute respiratory infection/pneumonia; diarrhoea; and infant and child feeding) was also incorporated into the urban HDSS. Geographical information system data on the households, health facilities and other important elements of the urban HDSS area are collected and annually updated. These data are used to calculate health and demographic indices for monitoring and documenting in the urban HDSS annual report. Variables collected in the Urban Health and Demographic Surveillance System Identification number (slum,abari,b householdc and individual) Name, age and sex of household members Education (age 6+ years), occupation (age 8+ years) and marital status Number of dwellings and dwelling area Construction material of main dwelling Water/sanitation, source of light, garbage disposal Cooking place and source of fuel Place of origin Duration since migration Reason for migration Baseline census and socioeconomic data 2015 Household head’s migration duration and socioeconomic data are also collected during routine data collection (3-monthly) Household possession of durable assets Education Occupation Pregnancy outcome (live birth, stillbirth, miscarriage induced and miscarriage spontaneous) Identification of pregnancy, father, mother Sex (live birth), date and place of pregnancy outcome Duration of pregnancy, mode of delivery, pregnancy outcome (number of live births/stillbirths) Antenatal care of mother by trimester, place of antenatal care, and service provider Routine data collection (3-monthly) Last menstrual period was asked confidentially Of danger signsf (pregnancy, delivery, postpartum period and newborn) About referralsg (for mother during pregnancy and for newborn) Components of antenatal careh Postnatal care (mother and child) Visit duration after birth Service provider Iron and folic acid (mother) Identification of deceased Age, sex, date and place of death Verbal autopsy (WHO VA 2016 with open narrative, sign and symptom that led to death) Routine data collection (3-monthly). The WHO VA 2016 data has been adopted and collected since 2021. Data collection by specially trained fieldworker Coded by trained medical assistant and computer-based algorithm Identification of migrant Age, sex, and date of migration Education, occupation and marital status of in-migrants Place of origin (in-migration) or destination (out-migration) Reason for migration Date of internal movement Creation of new household and household split Cause of internal movement Date of change Reason for change in headship Age and sex of new head New relationship with household members due to change of head Identification of partner(s) Age, sex, education and occupation of the partners Previous marital status of the partners Desire for children and ideal family size Contraceptive method use and source of supply Geo-coordinates of bari and households, water source, toilet, cooking place, health facilities, and roads Data on diarrhoea and acute respiratory infection in children aged under 2 years were collected from the mother with reference to past 14 days before the survey date Early initiation of breastfeeding (within the first hour of life) Exclusive breastfeeding in the first 6 months of life Continued breastfeeding for 2 years or more Introduction of safe, appropriate and adequate complementary foods at 6 months of age Childhood immunization records of children aged 12–23 months were collected from the mother A slum is a cluster of compact settlements of five or more households, which generally grow unsystematically and haphazardly on government and private vacant land, and is prone to adverse environmental hazards and exposures (BBS 2015). A bari is a cluster of households where the owner of the houses is usually the same, and the tenants use a common space or facility. A household consists of one or more people who live together, share meals from a common cooking pot and can identify one member as head of the household. The de jure population are the people who usually live in the households. A household member is defined as a migrant when their place of birth is different from their current place of residence. Danger signs during pregnancy include vaginal bleeding, convulsion, severe headache and blurred vision, fever and weakness, abdominal pain and fast and difficult breathing. Danger signs during delivery include waters breaking earlier, excessive bleeding, prolapse of vagina, convulsion (eclampsia), high fever (sepsis), severe headache and loss of consciousness. Danger signs during postpartum period include excessive vaginal bleeding, convulsion (eclampsia), high fever (sepsis), foul-smelling discharge, severe headache and loss of consciousness. Danger signs for the newborn include inability to suck, baby being too small, fast breathing (>60 breaths/min), convulsion, drowsiness or unconsciousness, not moving, grunting, chest indrawing, high fever, hypothermia, central cyanosis and red discolouration of umbilicus. Reasons for referral of mother during pregnancy include prolonged labour, excessive bleeding, severe headache, obstructed labour, convulsion and late retained placenta. Reasons for referral of newborn include breathing difficulty, difficulty with sucking, cyanosis, temperature too high/too low, convulsion and chest indrawing. Components of antenatal care include maternal weight gain monitoring, blood pressure measurement, urine test, blood test, informing mothers of danger signs during pregnancy, counselling for breastfeeding, maternal nutrition counselling and counselling about birth preparedness. An in-migrant is an individual who was neither recorded in the past census nor born or lived in the area after the census and who has moved permanently into the surveillance area. An out-migrant is a resident who moved out of the surveillance area permanently or for 6 or more months. Internal movement refers to a person who was a resident of the study area and moved to another household or formed a new household or moved into another location in the surveillance area. Due to death or a household split. One person must be from the surveillance area. The urban HDSS households were mainly headed by men (82.3%), with an average of 3.8 people per household. The majority of people (64.1%) were of working age (15–59 years), followed by children aged less than 15 years (31.5%) and people aged 60 years or older (4.4%). Among adults (aged 15 years or older), about 36% were men and 42% women had no scholling. Of school-aged children (aged 6–14 years), 14.1% of boys and 8.9% of girls had no formal education. More male than female participants aged over 8 years were engaged in income generation (71.7% vs 40.3%). Most households were very small, with 81.6% having a single room. Most had tin roofs (94%) and walls (70%) and a brick or cement floor (88%). Nearly 94% of households used piped water for drinking, and 90% had sanitary latrines; however, only 30% of the latrines flushed to sewerage or septic tanks. Most amenities were shared: almost 95% of households shared a piped water source, 60% shared cooking facilities and 90% shared a latrine. Electricity was universal as a source of light but cooking was fuelled via a gas line in 53.5% of households, with the rest relying mainly on wood (for details, see Razzaque et al. 20196). Between 1990 and 2018, child mortality (in children aged 1–4 years) declined more than infant mortality. For a more recent period (2010–18), infant mortality and under-5 mortality were higher for: children of working mothers or mothers with a short birth interval; children who were small in size; where there were delivery complications; and those who were vaginally delivered. Mortality was lower for girls than among comparable groups.7 In the slum areas, rates of and use were levels were also but and was with increase in education for but increased for and to with household with education and of high blood pressure increased with education for but not for was usually higher among and households than among their less and with those born in the migrants in antenatal care, birth, birth, child immunization, birth and use of The of service among migrants but not with that of the population. who were engaged in were less to receive adequate of births and 50% of births were by In private facilities, of births were by followed by in facilities and in non-government with the the of birth by were higher among and women who had or more antenatal care during pregnancy had two higher of delivery than women with no antenatal care data on people aged years or one study that the and health status of recent migrants was better than that of and of with recent the and health status of men was better than that of for more people with less people and for or with those of the the of household members experienced Among some of care, with to or informal health care Nearly of among care and a The average treatment for an was and the average was About of household experienced health-related economic and had to assets to with the Future to with infant and child as well as identification of of childhood maternal mortality and plan to for health care also to for diseases and an intervention to their through treatment and practices. The urban HDSS of study designs that allow for and are more than study can also by the sample and which to the range of health outcomes and Urban HDSS data can also be used in to the for data on health health and other in urban areas relevant to the As is high in urban be followed for to improve of by using mobile and identification in the is the on settlements with 100 or more households. This the of the the settlements included in the urban HDSS are of one-third of the informal of the population. The data are not but can be shared via the and by the icddr, data access with the and community are always can be to Razzaque or can be obtained via the urban HDSS The study was by the and of icddr, b of the and the the first of the in the and the for The had access to all data and had for the to the for This work is by icddr, b with the of for of this research icddr, b is also to the Government of Bangladesh for and to and the The is a member of the INDEPTH

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.