Cohort profile of the largest health & demographic surveillance system (Dibrugarh-HDSS) from North-East India

Health and demographic surveillance systems (HDSSs) are used to provide a foundation for the prospective collection of sociodemographic and public health data within a community. Such systems, originally called population laboratories, have been in operation across India1-3. It helps to initiate evidence-based research into diseases that are present in the community. HDSS sites in India such as Ballabgarh, Vadu and Birbhum are contributing immensely to the generation of health-related data and policy testing in the country. Further, many of the global HDSS sites generating quality data on global health have come under the umbrella of INDEPTH Network. These HDSS sites have been instrumental in generating longitudinal information on health and population to guide policymaking. Data from national surveys indicate that both maternal4 and under-five mortality rates are highest in Assam. The National Family Health Survey 4 has shown poor coverage of most of the healthcare interventions in Assam5. Dibrugarh district from upper Assam has the highest maternal and infant mortality in comparison with other parts of the country6. It has been reported that the tea-garden community contributed to a major portion of these deaths possibly due to poor socio-economic status, lack of awareness about ongoing health schemes and poor accessibility to the healthcare provider or facility7. In addition to these, non-communicable diseases (NCDs) including hypertension, presence of lean diabetics and various forms of cancers are increasing in the tea garden and village population of the district8. With this background, the Indian Council of Medical Research (ICMR)-Regional Medical Research Centre (RMRC), North-East (NE) Region, Dibrugarh, Assam, India, had initiated a Health and Demographic Surveillance System named ‘Dibrugarh-HDSS’ in January 2019 with the objective to create a health and demographic database of a population of a defined geographical area and a surveillance system to provide technical assistance for implementation of programmes and formulating interventions for reducing morbidities among different population strata. Dibrugarh-HDSS has created a well-defined population-based cohort to promote and conduct research on demographic processes, population, health and disease epidemiology.

Governance structure of Dibrugarh-Health and Demographic Surveillance System (HDSS)

The study was undertaken after approval from the Institutional Ethics Committee, ICMR-RMRC, NE Region, Dibrugarh.

The data collection process was completely paperless (we used mobile app software as indicated elsewhere) but a format for taking informed consent was incorporated into the mobile application itself. Before starting any interview/data collection, this informed consent was read out to the participants in their local language. Only upon recording the participant’s consent the mobile app would allow for the further data collection process.

Setting up of Dibrugarh-HDSS: Dibrugarh is an eastern district in Assam and the health status of the State of Assam is among the poorest in the country when it comes to maternal and infant mortality. To address these issues, a brainstorming session involving expert members from various fields was held at ICMR-RMRC, NE Region. This meeting was attended by experts from ICMR, New Delhi, ICMR-RMRC, NE Region, Dibrugarh, National Health Mission; existing HDSS sites, statisticians and epidemiologists from All India Institute of Medical Sciences (AIIMS), New Delhi, and United Nations Children’s Fund (UNICEF), faculty and principal from Assam Medical College & Hospital (AMCH), personalities engaged in global health research and innovation experts (Community Empowerment Lab, Lucknow), representatives from the district and State health authorities, tea-garden management authorities, medical officers from tea-garden hospitals and primary health centres in the area. During this meeting, the idea of development of an HDSS in Dibrugarh was conceptualized adopting the panel design method. After discussions, it was decided to establish an HDSS in Dibrugarh. This meeting served as a platform for identification of potential investigators from different organizations including ICMR-RMRC, NE Region, as the leading institution, entrusted with the responsibility of development of concept and project protocol. The responsibility for the design of a paperless method of data collection was also allocated to the potential investigator(s). Initially, it was proposed to include 60 villages and 60 tea gardens. The project was written and submitted and funds were allocated accordingly. The current HDSS is funded and being coordinated by ICMR, New Delhi. However, after the first review of the project, it was suggested to limit the population size to one lakh. Till then, 60 villages and 20 tea gardens were already covered and hence finally 60 villages and 20 tea gardens with a total population of 1,06,769 individuals were finalized for surveillance. The baseline data collection was done using the mobile application named ‘ICMR-HDSS’.

Study site: Dibrugarh-HDSS covered 60 contiguous villages and 20 tea gardens taking ICMR-RMRC, NE Region, as a centre point (Fig. 1). The basis of selecting these villages and tea gardens were logistic issues, operational feasibility and representation of both tea garden and the rural population of Assam. Before setting up the surveillance system, a complete line listing of all households was done, and a map was prepared with a temporary serial number of the households. A panel design was adopted to create data at baseline and follow up prospectively.

Close

Fig. 1

Location of Dibrugarh-HDSS study sites. Source: ArcGIS Version 10.5.1. HDSS, Health and Demographic Surveillance System

Export to PPT

Geographical location of the HDSS study area: The study area of Dibrugarh-HDSS consists of villages and tea gardens of the Dibrugarh district, Assam. Geographically, Dibrugarh district is divided by the river Buridihing (a tributary of Brahmaputra river), from east to west. Buridihing flows through Naharkatia and Khowang, and at a later stage in its course, Buridihing acts as a divider between Dibrugarh and Sivasagar districts. The region is flat with a gradual slope from the East Arunachal hills to the west. The soil of the district is mostly fertile and alluvial soil. It is the gateway to the three tea-producing districts of Tinsukia, Dibrugarh and Sivasagar9. These three areas account for approximately 50 per cent of India’s Assam tea crop and this gives Dibrugarh its rightly earned sobriquet as the ‘Tea City of India’. Oil and timber are the other two big industries in and around Dibrugarh9.

Surveillance population: This being a health and demographic surveillance system, included all the individuals from the selected villages (n=60) and tea gardens (n=20). All 60 villages and 20 tea gardens were contiguous taking ICMR-RMRC, NE Region, as a centre point.

Major portion of the Dibrugarh district is covered by tea plantations and tea factories. According to the 2011 census9, the Dibrugarh district has a total population of 1,326,335, with a population density of 392 inhabitants per square kilometre10. The decadal growth rate of Dibrugarh was 11.92 per cent for 2001-201110. The sex ratio of Dibrugarh is 961 females per 1000 males and the overall literacy rate is 76.05 per cent10. Majority of the populations believed in Hinduism followed by Islam and Christianity. Tea-garden community of Assam had migrated from Madhya Pradesh, Bihar, Odisha, Jharkhand, etc. during the pre-independence era and mainly engaged in tea plantation. Since then, they settled in different tea gardens throughout Assam keeping their sociocultural habits intact11. At present, their population contributes to about 18 per cent of the total population of the State. Assam, a north eastern State of India, has diverse ethnic groups with distinct cultures12 which are reflected in villages. As per the protocol of Dibrugarh-HDSS, the entire population from 60 villages and 20 tea gardens of all age groups were eligible for surveillance. Hence, a complete enumeration for all the populations was carried out. However, participants not providing consent were excluded from the collection of detailed personal information. The village population relied mostly on public health facilities such as sub-centres, PHCs and medical colleges for their health problems. On the other hand, tea-garden population mostly relied on the garden hospitals for minor ailments and medical colleges for major ailments. The tea gardens have a tea-garden hospital within the tea-garden area. The villages have sub-centres in their village and the medical college (tertiary level healthcare facility) is within the same district of the population under surveillance. Both the communities in the surveillance area are quite different from each other in many aspects. However, together they impact the health status of the district, as a whole. Hence, while presenting the results, we have focussed on the overall status along with a break-up of the indicators according to the study cluster.

Data collection: The survey was conducted using a mobile application software installed in mobile tablets of the field staff. During the process of data entry, a unique identification number describing the State, district, area and location coordinates for each participant is auto-generated which is linked with the serial number of households prepared during line listing. Data regarding sociodemographic characteristics, blood pressure, anthropometric measurements and disease morbidity and mortality were collected from all the individuals using the mobile application software named, ‘ICMR-HDSS’. Facility survey was carried out for assessing the available healthcare infrastructure.

Tools for data collection: Data collection was done using pre-designed questionnaires from various HDSS sites of India after modifications as per local needs which were then pre-tested. These pre-tested questionnaires were then incorporated into the mobile application. There were separate questionnaires for household information, individual information (sociodemographic information, anthropometric and clinical measurements), birth event (including immunization history), death event, pregnancy and health facility assessment.

Variables: Baseline data including sociodemographic characteristics, demographic attributes of all individuals and elaborated pregnancy and reproductive history from all women of reproductive age group, maternal and childhood illnesses and information pertaining to birth event, death event, were collected by the surveyors. Blood pressure (in mmHg) of ≥18 yr individuals was measured using digital automatic blood pressure monitor (OMRON Healthcare Co., Ltd., Japan). Hypertension was defined as an average systolic blood pressure (SBP) of ≥140 mmHg and/or average diastolic blood pressure (DBP) of ≥90 mmHg. The anthropometric measurements which include body mass index (BMI) (kg/m²) of all the individuals of ≥5 yr were collected using platform balance (Krups, New Delhi) and stadiometer (Seca, GmbH & Co., Hamburg, Germany). All information was collected following informed consent from the participants.

Verbal autopsy: The death cases reported within Dibrugarh-HDSS were visited by supervisors trained in verbal autopsy for verification of the reported death and conduction of verbal autopsy. The narrative history in the local language, which included illness or events leading to the death of the individuals in the chronological order of occurrence, was collected. The complete history of symptoms, signs, events, investigations and treatment was obtained so that the medical reviewer had sufficient information to assign a probable specific underlying cause of death (COD). The list of deaths for the reference period of one year from the concerned health worker for each cluster was obtained. In addition, the list was updated based on additional deaths reported by community members after visiting the area. Additional sources were also used to ensure complete coverage, such as data from the local church or funeral sites. The family of the deceased was approached and, after obtaining informed consent, the semi-structured verbal autopsy questionnaire was administered in the local language to the family member who was closest to the deceased before death. The COD and International Classification of Diseases (ICD)-10 code was assigned independently by two medical doctors. If there was a disagreement, data were reviewed by the two doctors together to discuss and agree based on the consensus. The data were analyzed and summarized in major categories for the COD based on the ICD, 10th revision.

Features of Dibrugarh-HDSS mobile application package: A unique mobile application software was developed for online and offline data entry in HDSS. The mobile application was conceptualized and designed in-house and the software development was outsourced to a computer firm. The app ‘ICMR-HDSS’ used PostgreSQL, PHP (Laravel framework) as back-end database and Java for android application as front end. This application has the provision for monitoring daily field activities including real-time viewing of mobile tablet location on Google maps, task assignment, data extraction in Excel format, QR code generation for follow up study and audit trailing of login and logout activities of all staff including administrator at any time. The application is scalable for pan-India use. However, cost implications may arise for large-scale use of the application.

Statistical analysis: Data were entered into the mobile application, exported to Microsoft Excel spreadsheets and further to the Statistical Package for the Social Sciences (SPSS Statistics 26.0, IBM Corp., New York, USA) for analysis. Continuous data were presented as mean±standard deviation (SD). Categorical variables were presented as frequencies and percentages.

Results

Baseline profile of Dibrugarh-HDSS: Dibrugarh-HDSS covered a total population of 106,769 [male: 52,934 (49.6%); female: 53,835 (50.4%)]. Of these, 46,762 (43.79%) were from 60 villages representing the native rural population and 60,007 (56.21%) were from 20 tea gardens predominantly representing tea-garden community.

Fig. 2 represents the overall age-sex pyramid of the population. Fig. 3A and B shows the population pyramid of the study population (tea-garden population and village). The age group 20-29 yr contributed a major portion of Dibrugarh-HDSS.

Close

Fig. 2

Population pyramid in overall population.

Export to PPT

Close

Fig. 3

Population pyramid of Dibrugarh-HDSS. (A) The tea-garden community. (B) Rural population. HDSS, Health and Demographic Surveillance System.

Export to PPT

There were 22,536 households distributed across 60 villages (n=10,357) and 20 tea gardens (n=12,179) of Dibrugarh-HDSS. It was found that the average household size (4.9 vs. 4.5) and number of females per thousand males (1042 in tea gardens vs. 985 in villages) were significantly higher in tea garden community than in the rural population. Other backward classes (OBCs) were the major (56.1%) caste category in the rural population, whereas tea gardens were mainly inhabited by the tea community (88.1%).

More than one-third (35.1%) of the tea garden population were illiterate which was higher as compared to the village population (17.1%). Majority of them were Hindus, followed by Muslims in both population groups. Major family norms were nuclear family for both the population groups. Tea garden workers were mainly engaged in the tea industry, whereas the major occupation category in villages were homemakers or daily wage workers. Majority used tube well for drinking water, and electricity was the main source of power supply. The tea garden community was mainly engaged in the tea industry as either permanent or temporary workers. Hence, the number of service holders (16.4%) was higher in tea garden in comparison to the village (9.9%).

Tea garden community in Dibrugarh-HDSS reported lower BMI than their rural counterparts (Table I). There was a gradual increase in both mean SBP and mean DBP with an increase in age in both the communities. The prevalence of hypertension was significantly higher (29.4%) among villagers in comparison with tea-garden community (28.2%). The different blood pressure categories in the overall population (village and tea garden) of Dibrugarh-HDSS were normotensive (31.1%), pre-hypertension (40.1%), stage I hypertension (19.8%) and stage II hypertension (9.0%). Table I shows BMI categories according to age and population groups. It was observed that underweight was more prevalent in tea-garden community across all age categories.

Information on vital statistics: Dibrugarh-HDSS recorded a total of 1437 live births at baseline. Of these, 580 (40.36%) births were from villages and 857 (59.64%) were from tea gardens. In the villages, low birth weight was recorded in 142 (24.48%) live births which was significantly higher, 294 (34.30%) in case of the tea garden community. For both the population groups, place of delivery was medical college (56.8%) followed by PHC (32.2%). Normal delivery was recorded for the majority (85.2%) of the births. A total of 668 (village=265 vs. tea garden=403) death cases have been recorded at baseline putting the age-standardized death rate at 5.6 per 1000 population for villages and 8.3 per 1000 population in tea gardens. Of these death cases, 21 neonatal deaths (tea-garden community=12, village population=9), nine infant deaths (tea garden community=7, village population=2) and six under-five deaths (tea-garden community=3, village population=3) were recorded. Dibrugarh-HDSS has also recorded a total of eight maternal deaths (tea garden community) at baseline.

Prevalence of major diseases (18 yr and above): The Dibrugarh-HDSS recorded the prevalence of major diseases in the study population. For all the conditions enumerated in Table II, the individuals reported of ever been diagnosed with the mentioned condition along with documentary evidence of the same. The prevalence of diabetes was found to be significantly higher among villagers (1.8%) compared to tea garden community (0.36%) (Table II). It was observed that stroke and coronary artery disease were significantly higher among villagers (0.35 and 0.38%) compared to the tea garden community (0.22 and 0.09%) (Table II). The prevalence of tuberculosis was found to be significantly higher in the tea garden community than its village counterpart [208 (0.52%) vs. 111 (0.33%)] which corroborates with earlier study13.

Challenges in setting up the HDSS at Dibrugarh: There were hurdles of liaisoning and obtaining permissions in the surveillance area. This was especially prominent in tea gardens. Various stakeholder meetings were conducted with line-chowkidars (designated as the key person who takes the administrative responsibility of a small group of 25-30 households in tea gardens called lines. Each chowkidars serves as a link between the tea garden management, hospitals and garden labourers who are responsible for administrative as well as health-related-event information), mothers’ club members, medical officers and paramedical staff to ensure stakeholder engagement and participation of individuals. The road conditions of many areas were not good. Especially during the monsoon, some areas were difficult to commute. To overcome this challenge, the field assistants were assigned a different area. Although the surveillance site covers a population of 1,06,769 individuals, due to the low density of population, the surveillance sites were scattered across a large geographical area. Vehicles were provided for field staff who were assigned households from remote areas, to carry out their field activities.

Strengths of Dibrugarh-HDSS: All data were collected by field assistants at ground zero through a mobile tablet using a user-friendly mobile application software tagged with GPS which was developed for this purpose. Field staff were supervised by trained supervisors and trained from time to time to ensure quality data collection. Quality assurance sampling methods were adopted to maintain data quality. Dibrugarh-HDSS has incorporated a number of questionnaires including family information sheet, individual information sheet, pregnancy form, birth event form, immunization form, under-five children form, neonatal death form, child death form, adult death form and maternal death form with security features like geotagging of each individual during the survey. The mobile application enables real-time monitoring of health profile of the study population which will be useful for researchers, academicians and policymakers to develop strategies to alleviate the burden of communicable and NCDs and reduce adverse pregnancy outcomes and under five mortality. Dibrugarh-HDSS will also provide a platform to develop new studies.

Discussion

Dibrugarh-HDSS, like other HDSS sites1-3, has the potential to be utilized for methodological studies to evaluate the quality of data collected by other demographic and census surveys in a country like India. In the long run, Dibrugarh-HDSS could prove to be a pioneer in quality data collection and policy recommendations in the north eastern States of the country. The HDSS site would also be instrumental in mapping and monitoring the changes in the health and demographic status of the State. Dibrugarh-HDSS showed an increasing trend in the sex ratio which was also observed in the 2011 census14 data that showed a sex ratio of 958 females per 1000 males. In the Dibrugarh-HDSS study population, the age group 20-29 yr constituted the major portion. The prevalence of hypertension (adjusted for age) among >18 yr population was found to be significantly higher among the village population (29.4%) as compared to the tea garden population (28.2%). There have been many studies15-17 in the past that have studied hypertension in these two communities and higher blood pressures have been observed in tea garden population. Consumption of salted tea served during their workdays in tea gardens was assumed to be attributed to the raised blood pressure among tea garden workers. The decline in hypertension may be the result of widespread dissemination of the findings of the salt restriction intervention study18.

The Dibrugarh-HDSS has established a complete networking system with various stakeholders (ASHA, ANM, village heads, religious heads, etc.) for real-time assessment of information on vital events. For vital events, linkages were established with key persons from villages and tea gardens and information was collected from them against an monetary incentive. The information obtained from the key persons are then verified on ground by field staff and enumerated in the final database. As the engaged manpower was not enough for coverage of a large geographical area, this networking system enabled us to utilize limited manpower and resources efficiently. For stakeholder engagement, community meetings were organized at regular intervals for dissemination of results and creating awareness regarding several issues and risk factors identified during the surveys.

Way forward: Dibrugarh-HDSS provides an academic platform for the assessment of ground realities of different health problems with a number of variables under surveillance. The baseline findings will also be useful to identify risk factors associated with some diseases and to develop community-based intervention modules for the prevention and control of such diseases. This also has the potential to explore the risk factors working at different phases of the demographic, epidemiological and social transitions across population groups. Such efforts will help us for long-term sustainability of Dibrugarh-HDSS. Going forward, Dibrugarh-HDSS plans to initiate different community-specific interventions based on the community’s needs. Dissemination workshops across the institutions should be conducted to encourage the researchers to carry out studies in the HDSS platform. Further, data generated from Dibrugarh-HDSS are stored in ICMR-RMRC physical server maintaining strict confidentiality. These data may be available to the researchers after obtaining the necessary administrative and ethical permission.

Overall, with the strong support from the ongoing HDSS sites, continuous guidance from expert group members, support from State health authorities, tea garden management authorities, community stakeholders and participants, a Health and Demographic Surveillance System called Dibrugarh-HDSS, was successfully established. The development of mobile applications made this task easier for paperless data entry and real-time data monitoring.