 to give this, yes, welcome to this session on DHS2 as data warehouse, contraimplementation stories, and we have four presentations for this hour, which means we will have to be quite condensed. We'll do it as in other sessions that we do all presentations first, and then we have the questions at the end. But you are free, of course, free to post your questions in the chat or in the community of practice webpage throughout the presentations. We have a presentation on Hisp South Africa from Hisp South Africa by Moetim Boso and Tanya Govender, and we have two presentations from Indonesia, very Adrian and Mohamed Afdal. Thank you. Good day and welcome to all of our attendees joining us from all over the world this afternoon. My name is Tanya Govender, and together with my colleague Moetim Boso, we will be taking you through our presentation for the day, which demonstrates how we utilized DHS2 as a data warehouse together with other components through interoperability to provide a human resources for health solution to the Ministry of Health in South Africa. And just three and a half months after the start of the coronavirus pandemic in Wuhan, China, South Africa confirms its first coronavirus case in March, 2020. Shortly after that, our President Cyril Ramaphosa declares a national state of disaster and announces an immediate travel restrictions and closure of all schools. A national coronavirus command council was established to leave the nation's plan to contain the spread and mitigates the negative impact of the coronavirus in South Africa. Immediate questions from the national coronavirus command council were regarding the workforce that would take care of our sick, requiring information on how many healthcare workers we have within each facility, and also identifying where our specialists are. For example, our intensive care health workers. To support the procurement and distribution planning, there was an urgent need to know how much personal protection equipment would be required and where would the stock need to be directed to. As the virus spread throughout the country, the number of infected healthcare workers were rising rapidly. We needed to respond to questions on where the infected healthcare workers are and how can the remaining staff be redistributed to continue attending to the sick. South Africa receives its first batch of vaccines in February 2021, and the rollout commence shortly after with the priority to vaccinate all healthcare workers. The vaccine rollout planning required us to respond to how many vaccines were required for the healthcare workers and where should the vaccines be distributed. To respond efficiently to these requests, I would like to introduce you to South Africa's Human Resources Information System and how through interoperability, we were able to pull data from multiple data sources together to get the answers that we were looking for. But before we get there, it is important to mention that our initial approach was to immediately leverage our existing systems, which include our routine health system in the HIS-2, which is the national platform for health data reporting. In South Africa, we were not starting from scratch. We had long been planning how to manage the health workforce more efficiently and effectively. This process began with the concept of aligning the human resources information system to the National Department of Health's Human Resources Strategy. The strategy can be summarized to these five areas. Effective workforce planning for current and future needs, institutionalized data-driven policies for planning, having a competent and multidisciplinary health workforce, ensuring optimal governance in leadership and management, and building a productive and empowered workforce. The systems that support the human resources strategy lives within an integrated ecosystem with multiple contributing components. Here is an overview of the current components. The first component is our human resources for health registry. This is our authoritative repository of all the healthcare workers in South Africa. It is developed in alignment with the IHE profiles on a fire registry, followed by our second component, which is the human resources warehouse. This is where aggregated data from the registry sets and is used for planning and reporting. This is created in DHIS-2. The third component is created using the DHIS-2 web app. It is the human resources for health portal, which is the web interface for users to engage with the data for management and planning. And our fourth component is a custom app that is currently being developed in Hisp, South Africa. It is our human resources planner module that will be utilized for scenario-based and use case planning and predictions. The current system is still flexible and allows for additional components and further interoperability channels. I would now like to hand over to my colleague Moeti, who is our senior data engineer, and he will take us through a look into the HRIS architecture. Over to you, Moeti. Thank you, Tanya. Can I just quickly get confirmation if my voice is coming through okay? Yes, I can hear you go ahead. All right. Good day all watching. The current slide shows a simplified view of the HRIS architecture. I'll present the architecture by going through the information flow across its different components. We have inflow of data coming in from different primary data sources, which includes a government payment system with records of public sector personnel, statutory councils, and the private sector. The inflow data is basically demographic information about individual health workers. We have a staging platform, which is responsible for two-mayor processes, the processing of inflow data coming in from the primary data sources and data aggregation for importing. The preferred way of receiving data from primary data sources is in HR7 fire format. However, the HRIS system is able to receive data in any digital format. All right, this capability is achieved by establishing the activities required for transforming the source data into fire resources, and then automating those activities using Apache Airflow, which is an open source ETL2. Data sits within the HRIS system at different levels of detail. The HRH registry stores a detailed view of a health work practitioner record, which is consolidated from data coming in from the different primary data sources. The registry is implemented using happy fire and open source fire server. The following is an example of a fire practitioner and practitioner role resources housed within the HRH registry. The resources, lengths and details have been shortened for display purposes here. The HRH data warehouse stores aggregated data for HRH planning and reporting. It is implemented using DHIS2. This slide here is showing aggregated data on the DHIS2 dashboard. We can use any DHIS2 favorite for analysis of our data. Other databases used in SA are also part of the HRIS architecture, such as the national routine data coming from DHIS2, the master facility list, COVID response systems, training information systems, and others. The interoperability component provides access to the HRIS system for interoperability. It is implemented using OpenHIM, also an open source software. For COVID response in answering the questions that Tanya mentioned earlier, we established interoperability to the SA COVID data lake to determine health workers affected by COVID. We also connected and provide a list of all health workers in the country who are eligible for vaccination. The interoperability component is also used to manage the outflow of data from the multiple repositories we've just seen within the HRIS system. This can be for reporting, querying, or application development. Reporting within the HRIS system is done in multiple formats. Firstly, we have DHIS2 data visualizations and dashboards for aggregated reporting. This is an example of HR data analyzed using DHIS2 pivot table. When we have a request coming for analysis, we create the favorite in DHIS2 first in order to finalize the details of the report. We also use BI solutions for more advanced reports, reporting that can be done on both aggregated and granular data. We currently use this, we currently use Power BI for this capability. This report right now is showing a report prepared in Power BI. We have pulled data from the HRIS data warehouse in DHIS2. We can also pull data from other repositories via the interoperability component to analyze it all in a single platform. Once we are happy with this report, we will publish it on the web portal. The web portal is another component. The web portal is a custom DHIS2 app developed using leaflet and custom crafting libraries. It is installed on the HRH data warehouse. The slide you're looking at now is showing data in the web portal. This specific report shows health workers at a facility level, categorized by occupation category, gender, age group and race. This dashboard is interactive. Trilling down or up is possible through the OcUnit hierarchy on the left. A future feature his past started developing is a HR planning module. This would facilitate scenario planning from a current state to a future desired state and stipulate what are the inputs desired required to reach that desired end state. It would facilitate the ministry to establish effective health workforce planning to ensure alignment with current and future needs. It will assist institutionalize data-driven and research-informed health workforce policy, planning, management and investment. This concludes the overview on the HRIS architecture. For ongoing work, we are currently working with primary data sources to improve data quality. We are also still developing the HR planning module. We are looking to develop machine learning and predictive analytic models to facilitate future planning. We plan to work with stakeholders to enhance analytic outputs to facilitate better insights and planning. And we will continue moving forward to enabling the minister of health to reach the HRH strategic goals. This is the end of our presentation. We would like to acknowledge our partners that are working with us on this project. CDC National Department of Health and PEPFAR. If you'd like to connect with us, please reach out to us using this contact details, Shownia. Thank you all for your attention. Back to you, Johan. Thank you very much. And I think we'll then move on to the first presentation from Indonesia. Very Adrian, do you go first? Thank you, Johan. Can I start our presentation now? Yes, please. Okay, please let me share our presentation. I think you should have the authority to do that. Sorry, wait. Maybe I have to share the screen. Okay, first of all, good evening from Jakarta, Indonesia. Thank you for the opportunity that's given to us to share our presentation in this respectful event. I, very Adrian, will try our best to present our presentation. Please let me first introduce about Jakarta. Jakarta is one of 34 province in Indonesia and act as the capital city of Indonesia. It consists of six districts with over 10 million people live in Jakarta. And yesterday, we just celebrated our 494th birthday. In today's digital environment, the importance of data is greater than ever before. And it's critical that organizations are utilizing their data to make informed decisions in real time. For most organizations like Jakarta Department of Health, lacking data is not an issue. But determining how to prioritize data and put it into action is a common challenge. Although Jakarta has shown a lot of progress in implementing digital health services such as teleconsultation, tracing and tracking infectious diseases such as for COVID-19 and others, but it still leaves some basic problems, especially related to data management. There is still a lot of data that is fragmented and only known and or used by certain units. People are still arguing about whose data is correct and other problems. Perhaps like many other organizations, the demand for data integration arise from complex data environment where various multiple system and our manual reports are creating big data. The process of recording and reporting from various sources need to be facilitated so that it can run well and easily. Easy to access, easy to input, easy to store, easy to visualize, and of course, easy to analyze. In this case, the Jakarta Department of Health chooses the HIS-2 as the platform to integrate various reporting system in the form of data warehouse which already has some modules. Last month, we had the opportunity to present the protocol of the Jakarta Department of Health in developing our data warehouse at the IFIP conference. So on this occasion, we will try to present one of the modules in our data warehouse. The Immunization module is one of the priorities because Immunization data is needed in various reports both for the internal Jakarta Department of Health and also for the external that is required by the Ministry of Health and the Jakarta Provincial Government. Immunization data is also one of the indicators used to assessing the performance of the governor and the head of the Jakarta Department of Health. So the provision of Immunization data that is easily accessible by leaders has an impact on making policies related to public health. One of the modules that we build is the Local Area Monitoring Module for Immunization. This module was developed based on a manual report in Excel format that has been standardized by the Indonesian Ministry of Health since approximately 2004. And since its introduction until now, there has never been a significant change. Not only in this reporting format outdated, but it's also burdensome for Immunization officers because the number of the field that must be filled, the report in the form of a very long landscape format, and often the final result does not match the manual data due to the formula error in one or more cells which result in errors in various other cells. So the challenge in developing data warehouse and its modules are for working with the timely data, removing, remove all the tasselos, build a smart architecture, data security issues, and then also we have to keep up with the trends in the common challenge for the organization to match the targets needed to win and also data from newer units or user demands. And some problems in daily practice related to manual immunization data reporting, data validation and verification and difficulty in visualizing data. These are the problems that we used to face before we have the modules in data warehouse platform. So our goals by developing data warehouse and its modules are to speed up the process of reporting and collecting health data, is data visualization and analysis, integration and synchronization of health data. We believe the benefit from these projects is a better collaboration and deployment, availability of real-time integrated data, data from multiple distributed sources can be collected, helps in achieving better partnership and customer relationship, saves time, boost efficiency and reduce error, and of course, making excellent public policy. We would like to thank you to all also the HIS2 team and especially to Professor Yorn that given so much support for us and not to forget to all the committee who already held this event. Thank you from Jakarta. We back to you, Johan. Thank you very much, Bery. Then I think we'll just move to the last presenter, Mohammed Afdal also from Indonesia before we have a round of questions. Thank you very much, Mr. Johan. Let me share my screen first. Can you see my screen? Hello, Mr. Johan. Yes, we can see your screen, yeah, thank you. Thank you very much. Let me introduce myself. I will represent our team, Mohammed Afdal, to present our project into the digital implementation for Health City for study in MacArthur City. First of all, let us introduce our project. This is the project is namely Health Cities project. This is a five-year project starting from 2012 until 2022. This project has been implemented in four cities, indoor in India, MacArthur in Indonesia, then in Vietnam, also Kathmandu in Nepal. The Health Cities project is implementing the system approach for the project implementations that encourage multi-sector collaboration that include health, urban planning and development, information and communication to women and child department, nutrition and food safety, education, water and sanitation, waste management, transport and tourism department. From this system approach, we found that there are three key issues to achieve the healthy urban planning. They are data-driven learning, citizens empowerment and also multi-sectoral engagement. The goals of our project is to support healthy urban planning. The one of MacArthur City, we found that the biggest demand is to support better data integrations from the different sectors. This is to build the decision-making that's coming from the comprehensive data from the different sectors. And how does data integration make the city healthier? First, the data is the key role in multi-sector collaborations. This is to encourage different sectors to have a collaboration to address the social determinants of health. Also, the integrated data actually can allow the cities to work quickly, identify potential health risks, including pandemics, COVID-19, environmental issues and also climate change. Also this integration of the data can also highlight the vulnerable group that have health risks due to disparities in service coverage. So from those analysis, the building healthy cities project is trying to increase the quality of data, increase the accessibility and timeless of data for decision-making. And then we decided to use a consistent and user-friendly data sharing platform. Also, we tried to train the city officers for using technology for good implementations. And also we are improving the quality of data. We improve also the trust between sectors in using the data each other. Also, we are gaining for and timely access to community data. So from this context, we are using the DHS2 platform to fulfill the demand. So we are using the DHS2 platform to integrate data from the different sectors based on the healthy city program in Indonesia. So let us introduce first, the healthy city program has been running starting on 25 with around 170 healthy city indicators. This 170 indicators is coming from the different sectors, including housing and infrastructure, traffic, transportation, mining, tourism, and the sector. So the healthy city program in national level is led by the Ministry of Home Affairs and also Ministry of Health. This is our summary of our process. So starting on the initial meeting with the Ministry of University of Foslo team at the UNIFN annual conference 2019. And then moving to one year later, September 2020, during pandemic the BLC held the webinar to improve awareness of data integrations to the city that attended by 200 stakeholders from the different sectors, including also the mayor itself. And then one month later, November 2020, BLC also initiated the working group for multi-sector data integrations. As you know, this is to integrate data from the different sectors. This team need to have leading sectors where in this context, the BAPDA or the city planning and the club and agency is leading this process. So we can encourage different sectors to share the data. So on December 2020, we had a meeting, a technical meeting with the DHS to a business in Indonesia. So we planned for a free technical planning for this process. In March 2021, the BAC also had two days free talk workshop to define operational definitions, including numerator and denominator for each of those 117 indicators to be integrated into the SHS2 platform. And in April, we had workshop of data collections, also in April 12th because of pandemic. This is a summary of our metadata. So we developed the metadata into the DHS2 platform. There is included 170 indicators that's coming from different sectors, health, environment, threatened industry, education, social, health, transportation, public source and tourism. And then we visualized it into the dashboard of the SHS2. And also from the government side, we are also developing the standard operating procedure. So every sectors under the city can share the data which is the standardized mechanism and also agreements. This is a summary of our 170 indicators. So there are seven main areas we included into the SHS2 platform versus health settlements and facilities infrastructure, traffic and facilities, industry and tourism, health societies, security and nutrition. Also we have had also the minimum health standards from non-communicative industries into the SHS2 platform that include health service for children at elementary school, health service for protective age, elderly, hypertension, diabetes and also mental health. Also there were a request coming from the leading sectors the city planning and development agency to include also the poverty issue and stunting issue into our data integrations. Of course in our process there are some of challenges. So here are our best practice. So due to pandemic there are a lot of restrictions. So we convert with every process into online or virtual base. And also during the process there are some of issues around data security coming from the different sectors. So they are aware on how other sectors using their data. So we initiate something like memorandum of understanding or memorandum of action. So every department can have an agreement and that can also agree on standardized mechanism to share the data. Also the last thing, as there are some of different sectors ego here we initiate something like the leading sectors which we have mentioned in the previous slide that the city planning and development agency lead this process to ensure data sharing from different sectors. I think that's all. Thank you very much for all the attention from the participants. So now we're going to hear to the last presenter Walid Waziri from Afghanistan. We're going to present DHS2 as a data warehouse in Afghanistan. Hi Walid. Hi, can you see my screen? Yes, we can. Yeah, perfect. Yeah, thank you so much for the opportunity and I'm really exciting for this presentation. By the way, I'm really obliged for there's a background noise that's currently we're suffering from the electricity and there's a generator sound. So I hope that wouldn't be that much disturbing. No problem, please go ahead. Okay, thank you. So as you mentioned, so I would like to present DHS2 use as a MOPH or health data warehousing in Afghanistan. My name is Walid Waziri. I'm a senior advisor of data quality. Previously I lead the DHS2 implementation at the country where we are working as a partner with Minister of Public Health through USID projects. During the presentation, we'll talk about the specific objective of this project and also how the DHS2 were selected and what we accomplished from this implementation and the different DHS2 modules we will talk about and also the different data marks and our data sources. I will highlight some data marks and also some statistic from our current instance and also the roadmap or the invasion that we have for the project in the future and also during the implementation work where the challenge is that we face. The main objective of the initiative that we started because that access to health data was very limited in the country. There was offline MS access based system that it has some limitation and also whenever some party or someone would like to access the health data so it takes a lot of time to access that data and also data analysis or data visualization was not that much easy because no one is that much expert in MS access and also there was no routine monitoring of a different health program that we're running through government or through a different donor and also to support evidence base or data-driven decision-making also that was lacking and also on time accessibility of the data was a challenge that's atomized data, the frequency for the reporting for the routine atomized data were on a monthly basis. However, the final version which arriving to the central HQ, it was taking three months and also since there was different data sources while someone trying to triangulate the data and do some triangulation analysis of cross-cutting analysis that was very complicated and was not possible. So therefore we started DHI is true as a data warehousing. The process was at the beginning we had conducted two main assessment at the national level, the country level. First of all was the enterprise different systems or the databases that were at the ministry. We assessed all those and identified the potential sources for the data warehousing and also we assessed the infrastructure or the feasibility of the implementation of having online system as a DHI is true. And then we formed a data warehousing development task force team which we called DWD and this team were representative of different donors different partners like WTO itself the MOPH and the USID and some other donor or supporter. And then this team reviewed different option for data warehousing platform whether we should go for custom development of the data warehousing or having some ready-made solution as a DHI is true. Checking all those pros and cons and then the DHI is true were selected and as a data warehousing platform or national HMI system. And in the first version HMIS which has a different modules in EMI it's stand for explanation management information system and pharmaceutical and also HMIS these were different data sources that at the first version that we would like to have in the data warehouse. And for the implementation process we developed a scale up or we called it an institutionalization plan. So this plan really helped us that was a three years plan and also part of this plan we did some donor other donor engagement as I mentioned and part of this plan was all the development of the different system and capacity building and also some required resources for the implementation like some equipment for infrastructure. And then a part of this plan was also 31 different data sources at overall data warehousing to have 31 different data sources or data marks. Since as I mentioned there was legacy data or MS access based data at the ministry and we developed a custom ETL expect transform and load application that's to transform and load that MS access data to the DHIs too. So that was also quite lengthy process for data from 2003 up to 2020 all were based in MS access and we loaded all that data about like 17 years data to that ETL to the data warehouse. And also to easy the use of the system we localized all the interface and also all the metadata that we had in the system we translated to local language and also to support sustainability of the system different types of documentation will be developed in a close collaboration with the HMIS Department of Minister of Public Health. And also we created another group which we called a data warehouse core technical working group. So that group is the main responsible team which is leading under HMIS at the ministry and these group is responsible for all the system development, system maintenance and capacity building. Since the DHIs too was new in the country we did many DHIs to standard academies some academies were supported at the country as we invited HES India and we conducted three in country information use customization for cracker and aggregate level customization academy in country. And also we supported HMIS other staff to attend different academies at the different countries. And also to make easy data analysis and data use we developed 59 different public dashboards in the DHIs too and also out of the DHIs too. I mean out of the DHIs too is some dashboard that we have without user authentication. So that's a public dashboards. We specifically one design for COVID-19 and also there's a HMIS public portal also currently in the development state that's also DHIs too. The dashboard module is customized and will be bypassed authentication and also enhanced with some additional features to fulfill our requirements. In addition to that we had some custom development as well as we have here in the country for P or based on performance payments mechanism of the health system or for that we have a custom dashboards which analyze all those data and based on that analysis payment is made to the different implementer angel or service delivery angels. And also the COVID-19 dashboard also requires some custom development we had that also as a public dashboard. And also other beside that we also developed our web based master health facility that's which is serving as an administrative boundary for the system usually the organets or implementer or donor information. So that's also interacting with DHIs through EPI. So this is our big picture of the DHIs too or the data we are housing. Our instance is quite complex. So we have these aggregates data sources and we have these tracker or event based programs that's all in a one instance we are using and some of the data that I previously mentioned that we especially the HMI legacy data and also for some other data sources we uploaded or loaded through this ETL process. In the country we are using all those three module of the DHIs too also we're using aggregate event or tracker specifically the tracking for the COVID case tracking we are using the standard application or module which is released by University of Oslo. In addition to that we have some custom development custom program for our COVID-19 case tracking similarly for vaccination as well. Current our instance as I mentioned we have 59 different program or thematic area dashboards. In a total we have 77 data sets in our current version and we have more than 1,600 active users and similarly we have 31 programs but the most important that as in our instance is quite big we have more than 227 millions data values that's representing for the data from all those different data sources. These are the public custom developed dashboards that's specifically for COVID and also we have that we call the program here Sahat Mandir which is a multi donor program that's all those payment process or the verification of the data is also happening through the VHIs too. And also this is our public dashboard or HMI's country level dashboards which all those 59 dashboards that I talk about all publicly available for a user without any user or password. For the data warehousing still we are working and we envision envisioning that's to have some more progress in achievement. So based on this roadmap specifically we are working on a capacity building and also some additional resources that's nowadays coming in that's the current data warehousing will be enhanced and also since there's the legacy data that's imported through the ETL so some data quality issue are over there. So now also we are focusing on the data quality and data cleaning process as well. And also some other custom apps also we have in a plan that's for next year we will have some apps that's which is developed in country and also for interoperability that the master health facility that I mentioned so for that also will be the system will be interoperable with that old unit or since the security situation here is so tough and so lots of changes happening in terms of closing health facility or adding some health facility or some outbreak there's some temporary health facility to be activated for that purpose. We have that's a master health facility that's all those interaction should be happened in that master health facility with database and then through EPI that same will be replicated to the DHI stool. During our implementation process we had quite challenges. One of the main challenges that's even currently we are suffering from that's lack of support of the DHI to instant for solar calendar. So in the country Afghanistan and similarly some other countries like Iran or Tajikistan they are using a solar calendar for their reporting but unfortunately this feature is lacking and however we did some many time advocacy for the core team of the University of Oslo that's this should be integrated part of the core data is to next release that's due to other priorities still it's not schedule so we hope that gets some priority and this feature also will be fulfilled by the next releases. Also IT infrastructure is still a challenge in the country as the data collection or data reporting now in the DHI stool is happening at the provincial level. So that's still an issue and also for the use we local computer knowledge at the field at the some also this will were a challenge at some extent this has been address that we had extensive workshop or mentorship that we conducted somehow it's good but still it's need a lot of effort to be done in this area. So that's conclude my presentation. Thank you if there's any questions so I'll be happy to response. So unless there are any those questions now I'd like to thank all the presenters and all the participants for joining this session and wish you all the rest of the day productive and interesting in the continuation of the annual conference. Thank you.