Lessons learned from setting up the Nahuche health and demographic surveillance system in the resourceconstrained context of northern Nigeria

Background: The present time reflects a period of intense effort to get the most out of public health interventions, with an emphasis on health systems reform and implementation research. Population health approaches to determine which combinations are better at achieving the goals of improved health and well-being are needed to provide a ready response to the need for timely and real-world piloting of promising interventions. Objective: This paper describes the steps needed to establish a population health surveillance site in order to share the lessons learned from our experience launching the Nahuche Health and Demographic Surveillance System (HDSS) in a relatively isolated, rural district in Zamfara, northern Nigeria, where strict Muslim observance of gender separation and seclusion of women must be respected by any survey operation. Discussion: Key to the successful launch of the Nahuche HDSS was the leadership’s determination, stakeholder participation, support from state and local government areas authorities, technical support from the INDEPTH Network, and international academic partners. Solid funding from our partner health systems development programme during the launch period was also essential, and provided a base from which to secure long-term sustainable funding. Perhaps the most difficult challenges were the adaptations needed in order to conduct the requisite routine population surveillance in the communities, where strict Muslim observance of gender separation and seclusion of women, especially young women, required recruitment of female interviewers, which was in turn difficult due to low female literacy levels. Local community leaders were key in overcoming the population’s apprehension of the fieldwork and modern medicine, in general. Continuous engagement and sensitisation of all stakeholders was a critical step in ensuring sustainability. While the experiences of setting up a new HDSS site may vary globally, the experiences in northern Nigeria offer some strategies that may be replicated in other settings with similar challenges. © 2014 Olatunji Alabi et al.

The first step in establishing an HDSS was to select the site. Two state governments were interested in housing the site: Jigawa and Zamfara. To determine which state (if any) and which site within the state met the criteria for an HDSS, PRRINN-MNCH hosted a scoping visit by a team of international researchers with experience in developing and collaborating with INDEPTH sites. The purpose of the visit was to assess their understanding and commitment of each state’s stakeholders to the establishment of an HDSS, and then through discussions and site visits assessing the extent to which the potential sites met the requisite criteria for an HDSS site. The criteria for an HDSS site include: minimum of 40,000 population size; uninterrupted internet service; stable electricity supply (which may be achieved through alternative backups like standby generators and inverters); access to skilled labour force, e.g. computer specialists, programmers, interviewers; access to a diverse population who can provide different responses and a population that wants answers and is willing to answer questions to help researchers find solutions to problems; ability to interface with state and federal planners to identify health system research combinations to study and ability to mobilise resources to sustain the operation of the site among others. In each state, the international team first met with several stakeholders, including PRRINN-MNCH Programme officials, members of the State Health and Planning Ministries, local government and community leaders in potential sites. They conducted group discussions with community, local government, and state officials and assessed the acceptability of the initiative, ongoing efforts in population registration and commitment to population-based research on health outcomes, among others. The availability of supportive infrastructure such as suitable housing and offices was also discussed with the state government officials. The scoping mission next surveyed the physical and social characteristics of the proposed HDSS community sites to ensure that they had sufficient population size and accessibility to interviewers. They also identified existing health systems structures and initiatives in the states, and in the local governments proposed for the HDSS sites in each state. While sites in both Zamfara and Jigawa met the criteria in terms of population size and supportive infrastructure, the state and local government support among stakeholders was found to be much stronger in Zamfara. Zamfara, particularly the Bungudu Local Government Area and the district of Nahuche, offered housing and infrastructure support. Further, the Nahuche site included a functional PHC where various PRRINN-MNCH Programme interventions would be based. Therefore, the team recommended that Nahuche District in Zamfara be selected for the development of the Nahuche site. To ensure their understanding and commitment, key PRRINN-MNCH implementation research leaders and Nahuche stakeholders were invited to visit the HDSS site of Navrongo Health Research Centre in northern Ghana, a fairly comparable setting. The tour included field visits and presentations on the role of research in health systems strengthening and service delivery and translation of research findings into action. Throughout this visit, the PRRINN-MNCH team learned first-hand from the Navrongo experience and was able to solicit technical support and collaboration on setting up the Nahuche HDSS. The Nahuche HDSS study site with coordinates 12°36′ N, 6°54′ E, and 435 m above sea level is located in Zamfara State of north-western Nigeria (Fig. 1). Nahuche has a typical Sahelian climate with temperatures reaching a high of 38°C from March to May and a rainy season from May to late September, while the cold season, the Harmattan, starts from December until February. Situated in Bungudu LGA, the site is 32 km from the state capital, Gusau, and constitutes six districts of Bella, Gada, Karakai, Nahuche Keku, Nahuche Ubandawaki, and Rawayya. Gusau is a key commercial centre with a heterogeneous population from all over Nigeria. (a) Map of Nigeria showing Zamfara State and (b) Zamfara State showing Bungudu Local Government Area. The Nahuche HDSS study site is made up of 120 villages under the leadership of six district heads. Almost all residents in the study area are Hausa. Farming is the most common economic and subsistence activity of the people and is consistent with the slogan of the state ‘farming is our pride’ (9). High unemployment is a catalyst for temporary labour migration among men. The standard of living is quite low, with hardly any families owning cars or televisions, standard signs of improved living standards. Infrastructure remains deficient with no access to an electricity grid and dependence on local generators. There is no community sanitation system, and most families are depending on latrines and wells or boreholes for their water. All villages have at least a primary school, and some have a junior secondary school. One primary health centre is located adjacent to the HDSS site offices while the general hospital is located on the way to Gusau. Most people access healthcare from both Traditional healers and faith based healers apart from public health facilities. On 12 October 2009, a steering committee was set up to fast track the activities that would ultimately lead to a successful establishment of the Nahuche HDSS. Membership of the steering committee was drawn from relevant government agencies including the Ministries of Health and Planning and Budget, Primary Health Care Development Agency, the Primary Health Care (PHC) directorate of the Bungudu LGA, State Ministry of Local Government (MOLG), and the Zamfara State Team Leader for PRRINN-MNCH. The committee’s mandate was to mobilise resources for the HDSS set-up and to facilitate collaboration with communities in the catchment area and with the Bungudu LGA. The state and LGA were asked to recruit junior cadre staff to facilitate the launch of the site, nominate members to the operations research advocacy committee (ORAC), which would serve as the ethics review board for the HDSS. The ethics sub-committee of the ORAC works as an Internal Review Board, and would review all research proposals implemented on the HDSS platform, as well as any in the state. All work of the committee was voluntary. Community mobilisation and collaboration is key in longitudinal surveillance of a population. The Nahuche HDSS Steering Committee began the community mobilisation by ensuring high-level commitment by officials from the State Ministry of Health and other state government departments. The committee then led the community mobilisation through repeated visits and briefings with the district heads and other community leaders. Continuing as a part of the ongoing activities of the HDSS, the community mobilisation meetings are organised in each district by two Imams from the two Islamic groups and two other community members appointed by the chief and the officer-in-charge of the PHC from each district. During the community meetings, issues discussed included: the purpose of the HDSS establishment, and support and commitment of the people and the traditional leaders were solicited. A key part of establishing the HDSS was the memorandum of understanding (MoU) signed between PRRINN-MNCH and the Zamfara State government. The MoU clearly spelt out the resource mobilisation guidelines for the HDSS centre. Three major partners were identified and each with its financial obligations: 1) Bungudu LGA provided three blocks of flats for use as office accommodation for the centre and was responsible for the perimeter fencing of the centre. 2) PRRINN-MNCH Programme provided technical and financial assistance for the successful implementation of the project. In addition, PRRINN-MNCH provided a four-wheel drive vehicle, office furniture, electric generator, motorcycles, computers with battery backups, internet service, and payment of salaries. 3) The Zamfara State government was asked to pass a law designating the HDSS as an agency of state with a budgetary allocation to ensure secured sustainable funding of the centre. The State Ministry of Health also was expected to provide senior staff (a demographer, a statistician, a public health expert, and an administrator) to work alongside the three long-term consultants employed by PRRINN-MNCH Programme for the HDSS; however, they were unable to provide this senior technical staff. As a result, the HDSS steering committee forged a strategic alliance between Zamfara State government and Usman Danfodiyo University, Sokoto (UDUS), which had several faculties able and willing to provide the necessary technical expertise. Therefore, an MoU was subsequently signed between the state government and UDUS in order to harness the technical resources available in the university and to ensure sustainability of the centre beyond the programme. With guidance from the INDEPTH Network and PRRINN-MNCH senior technical advisors on job specification and qualifications, advertisements were placed in the public media and persons were recruited for the positions of the HDSS manager, computer scientist, and field data managers. The INDEPTH Network helped train the PRRINN-MNCH technical advisor and the successful candidates at the Navrongo HDSS site in November 2009 to prepare them for set-up and the pilot census. Through contractual arrangements that mapped specific areas of need for technical assistance, INDEPTH provided a database management expert and an HDSS field management expert to give intermittent follow-on guidance and training to the data manager, HDSS manager, and the enumerators for the pilot census, baseline census and subsequent four rounds of data collection. The quality of HDSS data depends to a large extent on the quality and commitment of the field staff. Getting adequately qualified persons in the surrounding communities was an uphill task. Literacy levels were generally low in these deprived rural communities but even lower among women. Being predominantly Muslim communities, men are not allowed to interview females who are socio-culturally subjugated, so it was imperative that female interviewers were recruited. This meant that the recruitment of field staff had to do extraordinary outreach to find and recruit suitable female interviewers. Even though the screening and recruitment of fieldworkers and data entry clerks was an open selection process through advertisement of vacant positions for qualified candidates to apply, particular attention was given to any women who applied. In fact, few women from the LGA and local communities applied for the job. Further, most of them failed the screening test due to their lower literacy level. Therefore, in order to recruit females, females who demonstrated potential to succeed even if their scores were lower than those of men were selected. Fieldwork and data entry training manuals were developed using samples from other HDSS sites. All those who passed the initial screening for literacy, residency in the surveillance zone, aptitude for interviewing (or data entry) were trained. The fieldworkers and data entry clerks were trained together to ensure that the latter understood the nature of the data to be collected. In this way, fieldworkers would understand the structure of the data entry processing software and appreciate the built-in functionality of the interview to provide for consistency checks. The interviewers could see how any mistakes or inconsistencies contained in the completed questionnaire could be detected. Those with the highest scores in the post-test training were hired as field workers and data entry clerks, with the exceptions noted above to give priority to promising female interviewers who did not do as well on the written test. While 250 candidates came for the recruitment screening interviews, 50 applicants barely met the minimum criteria for selection. Out of the 50 applicants who were trained, only 25 were finally selected (20 fieldworkers and five data entry clerks). Out of the 20 fieldworkers, only four were females whereas out of the five data entry clerks, only one was female. The first step in establishing the database that is to become the surveillance system is to conduct a baseline census, in which the surveillance area is mapped, all communities and households or compounds are mapped, the area is subdivided or chunked into manageable surveillance units (to include about 500 persons each), and then the team conducts a census to establish the number and characteristics of the surveillance population at the initial baseline. Our first step in this process was to send a team of HDSS staff from Nahuche to the Navrongo HDSS site from 22 November to 6 December 2009 to learn all of the critical processes for rolling out an HDSS. Researchers, policymakers, and programme managers have long recognised geographic location as an important factor in population and health outcomes. Knowing how the health of individuals differs by place of residence can lead to a better understanding of where and why events occur and how interventions can be implemented effectively. The Nahuche HDSS therefore included geographic positional coordinates in its mapping of communities and households within the Nahuche surveillance site. Mapping, compound numbering, and recording of global positioning system (GPS) coordinates were done in collaboration with the Nigeria’s National Population Commission (NPopC), which has the mandate and resources to lead demarcation and other activities for census taking in Nigeria. The regional NPopC staff directed the mapping of the surveillance area and its demarcation into smaller units (clusters) for easy enumeration, by re-demarcating the current Enumeration Area (EA) census maps from the NPopC into smaller units referred to as clusters. Each cluster was demarcated to include about 500 people in approximately 72 compounds. A compound was defined as a dwelling unit or structures comprising of one or more households. The staff were grouped into four teams with NPopC staff and some community leaders to identify the boundaries of the clusters and note key features on the cluster maps. As is common throughout INDEPTH sites, it is important that the interviewers know exactly which household they are visiting, so that the interview results can be entered for the correct household, round after round. Therefore, compounds have a publicly visible number painted on their exterior compound wall. These numbers were designed to match up with the identification code of the household in the database, making it easier for the interviewer to ensure that they are speaking with the correct household. The compound codes identify the Nahuche EA or district, cluster, and compound. On the interviewer’s census form, this code would identify the compound, and the interviewer would add codes for household and member within the household. For example, NKA001001 denotes the first household in the first compound of the A cluster of the Nahuche Keku District. When the first round of GPS coordinates were mapped, the NPopC staff went with the field team to re-take the coordinates of a sample of the compounds in each cluster as a quality assurance measure. Re-taking coordinates not only assured the quality of the data but also provided an opportunity to retrain some staff who were still struggling with the appropriate use of the GPS equipment. With approximately 120,000 residents, the baseline census would be an enormous undertaking. Therefore, to ensure that the team was adequately prepared and all details managed, we conducted a pilot census in the 39 villages of Nahuche Keku District (EA), May–June 2010. The HDSS field management team employed a team approach to fieldwork, which generally consisted of one team leader with five or six field staff. For convenience and based on the available team leaders, a total of 25 field staff were divided into four groups, one per cluster. At the pre-census refresher training, fieldworkers were trained in completing the questionnaires and in particular how to ask the questions so that they obtained accurate and complete answers. If an interview was not completed on the first visit, further attempts were made with the household or respondent, up to three times and over three different days, before classifying the case as non-contact. The subsequent contacts were scheduled at times when the respondent was more likely to be at home. The team leader or supervisor reviewed all completed interviews daily, checking thoroughly for empty response fields, illegible responses, and inconsistent responses, using the internal consistency checks built into the questionnaire. Fieldworkers were provided with feedback before proceeding to collect data on the following day. Between May 25 to June 2, 2010, 1,456 individuals were enumerated from 197 households in the four clusters. The results and findings from the pilot study have been documented elsewhere (9) and informed the roll-out of the baseline census. The HDSS programme manager provided additional refresher training on the completion of the items, and in particular on the complete listing of all members of the compound, even those absent on the day of the interview. The number of households per interviewer per day was reduced from five to four. In addition, the interviewers were given better guidelines for responding to community requests for medical help or other benefits in appreciation for participating in the census. Finally, the supervisory process was better organised, as well as the daily logistics of moving the teams. The full baseline census was conducted in the entire surveillance area after the initial pilot. The census was conducted between September and December 2010. The full baseline population of the surveillance area was 125,149 from 19,193 households. The methodology, results, and findings from the Nahuche HDSS full baseline have been reported elsewhere (10). After the baseline census, six-monthly cycles of data collection beginning in January 2011 were conducted in all households identified during the baseline census and new households were also registered. Trained interviewers visited the communities under surveillance, record events (pregnancies, births, deaths, migration, marriages, and vaccination coverage) in registers, and report data to the Nahuche HDSS Computer Centre for processing. As of December 2012, the population under surveillance was 144,496 residing in 20,371 households (average of seven individuals per household). For each round, additional elements of the surveillance system have been put into place. At Round 1, we began the listing of movements in and out of the households, so that migrations could be tracked into and out of the surveillance community. With Round 2, we designated in-migrants and out-migrants depending on their change of status from Round 1 to Round 2. We also added a module on the maternal and child health, including more details about immunisations. At Round 4 and not Round 3, we added verbal autopsy methods, joining the other INDEPTH sites which are trained to inquire about the circumstances just before and at the time of death for any deaths occurring between the rounds. These verbal autopsies are submitted to a panel of three trained physicians who code the presumed cause of death according to the signs and symptoms noted. Stakeholders developed a strategic planning framework for Nahuche HDSS, which includes the mission, vision, and objectives for the 5-year period beginning from 2013. The context within which the strategies were selected for each priority area was based on the SWOT analysis. The priority areas of the framework are: strengthening institutional management; human resource development; sustainable funding; infrastructure development; data quality and INDEPTH Network Membership registration (obtained in November 2012); research agenda; community engagement, support, and benefit; dissemination and communication of findings; and partnership and networking. The strategic plan has since been implemented and institutional management arrangements are being put in place for the centre. The budget for the 5-year plan defines the expected sources of funding of the various components of the plan. Collaborators from UDUS, our international advisors, and the state government are leading this process.

N/A

• Dima Health