DHIS 2 Symposium

With heightened scrutiny for storage and use of data containing personally identifiable information (PII) and personal health information (PHI), the public health sector is increasingly dealing with security and compliance requirements.  From General Data Protection Regulation (GDPR), Health Insurance Portability and Accountability Act (HIPAA), and local country requirements, the need for including security in health information management systems (HMIS) is no longer an optional afterthought, but rather a key requirement.  

Experts from BAO and the University of Oslo will discuss security considerations at the the following levels: organization, hosting, and application.  The organization level will look into the high level security considerations that organizations should consider and be aware of. The hosting level will discuss strategies and specifics required for proper storage, auditing, maintenance, and reporting.  Finally, for the application level we will discuss specific features in the current and upcoming DHIS 2 version as they relate to individual level data security, storage, and use.

QuestMark (QM) aims to provide NGOs, DHIS 2 project implementers, research organisations, international consultants and other data warehouse server developers with a new Android platform. QM translates DHIS 2 concepts into intuitive processes covering the following key functionalities: Data Collection (e.g. for Malaria Case Surveillance apps), M&E (e.g. for Quality Assessment Tools) and LMIS (e.g. for per-clinic stock control). This is accomplished relying on the combination of a modular system and a clean architecture to properly separate the responsibilities of the different pieces. In the nutshell, QM is not an app but a framework to build an app.

Achieving universal health coverage requires mobilizing and allocating sufficient financial resources. This can only be done properly if the cost of providing health services is known. Management Sciences for Health (MSH) has assisted governments and NGOs in many countries in modeling the costs of health services. Accurate costing improves advocacy for and supports better planning and allocation of, funds to ensure that providers, medicines, supplies, and other resources are made available as needed. Modeling the costs of health services in hospitals, health centers, and communities rely heavily on accurate information on the numbers and types of services provided, including diagnoses for illnesses. For large-scale (national or provincial) modeling, the only effective way of obtaining information on services is through a national health information system. While experiences have generally been positive, matching health information system data with the actual services to be costed can be challenging, and data are often incomplete and/or inaccurate. 
In 2017, MSH assisted the Uganda Ministry of Health (MOH) in conducting a costing study of government and private health facilities, in order to estimate the cost of providing services and the financial resources needed to establish a national health insurance program. The main source of service data was the MOH’s national Health Management Information System (HMIS), which is built on DHIS 2 and has data for most of the public and private health facilities in the country. These critical data for cost modeling were matched with data on staffing and costs collected from a sample of facilities. By comparing the numbers of services provided by diagnosis with the estimated prevalence and incidence rates, the team was able to ascertain both the utilization levels and the gaps between utilization and need for services. The team was also able to show the additional financial resources required to provide better quality of care and to scale up the numbers of services to match the need. Since the costs in the models are driven by the numbers of services, the team was able to ascertain the cost of improving quality of care and of scaling up utilization.  
Uganda’s HMIS is generally well designed, and it provided much of the information needed for the costing, but the team faced a number of challenges. Some facilities that did not provide data could not be included in the sample of facilities costed, and also were not included in the national totals of services. For example, all facilities in Uganda treat malaria; however, if some facilities do not report their malaria treatment data, then the total number of malaria treatments provided in the country cannot be costed or compared with the need. Some facilities only provided data for some of the months of the year, and there did not seem to be an effective follow-up system to ensure that data gaps were filled. Some facilities submitted figures that were obviously incorrect, but there seemed to be no effective validation system in place. Finally, comparisons with the package of services showed that the HMIS does not provide all the information needed for costing.     

The HMIS provided vital information for the costing, and for determining which services should be scaled up. However, improvements need to made, both in the selection of indicators and in the completeness and validation of data so that the system may reach its full potential. 

The challenge that we face is that in South Africa, which has largest HIV burden globally, with an estimated 7.1 million people living with HIV (PLHIV) and a projected increase to 7.5 million in 2020. While progress is being made towards epidemic control defined as less new infections than the number of AIDS deaths, 90/90/90 targets will not be attained by 2020 without a strong focus on acceleration and additional resources. A new PEPFAR funded initiative, the Treatment and Retention Acceleration Program (TRAP) is to accelerate epidemic control in South Africa by putting 6.1 million people on ART in the public health system and 610,000 in the private sector by December 2020, resulting in a total of 6,710,000 people on treatment by December 2020. The presentation will review steps to create dashboards in the DHIS, to set targets at provincial, district, and facility level, using available denominator and target data. In particular the use of dashboards and the new interpretation feature is seen as a major innovation for bringing various stakeholders together to improve performance and enhance supervision.

Beginning in 2013, the Lao PDR Ministry of Health (MOH), with technical assistance from the World Health Organization (WHO), implemented DHIS 2 as part of the national health sector reform framework. Today, the Laos office of Population ServicesInternational (PSI) actively collaborates with MOH to integrate health data from the private sector into the national health management information system and provides technical assistance to strengthen the use of data in DHIS 2 to drive decision making. Data visualization and evidence-based decision making have now been integrated into the daily routine in the MOH and PSI/Laos, as shown in the recent, government-led response to a malaria outbreak in Southern Laos. Electronic integration of private sector data through DHIS 2 increases the timeliness and quality of data to strengthen decision-making. This presentation will present PSI/Laos’ experience, lessons learned, and its plans to replicate this solution for other health areas while continuing to provide technical assistance for interpreting and using data to take action.

Countries are confronted with multiple siloed routine data systems. National health information systems are often separate from other data systems. Separately, they generate a huge amount of data that could be more powerful if analyzed together. While benefits of data integration are discussed, there is still a need for health information architecture that reduces duplicate efforts and facilitates data use. Now that there is a vast amount of data for use by different programs, the next challenge is to improve the use of data between programs for health-systems strengthening. By presenting different use cases, this session will address that challenge from different angles to demonstrate the potential for centralizing data, how it can be done, and what are the challenges along the way.

Lessons learned from the integration of e-TB Manager and DHIS 2 in Bangladesh are of profound importance: both tools are internationally recognized and operate in many countries, but this is the first time that their interoperability has been fully implemented and proven to exchange TB data successfully. The same approach could be replicated in other countries where e-TB manager and DHIS 2 co-exist.

Come see the complex interactive forms used by PEPFAR to track in DHIS 2 USG-funded AIDS intervention around the world.  These forms push the envelope of data collection in DHIS 2, including two sets of tabbed navigation, dynamic calculation, dynamic mutual exclusion, integrated validation rules, and drafts in static HTML.  The process for creating the forms and integrating them into DHIS 2 will be covered as well.

During this session, the speaker will present the pilot of an added module in the health information system of Haiti using DHIS 2 to capture patient-level data for routine Tb cases. The speaker will emphasize how this system interacts with the current national information system and how it prepares the ground for more efficiency which may lead to improved quality of care provided to Tb patients at the national level.

VecnaCares designs and deploys the CliniPAK, a solution for a low-cost, field-ready server for hosting applications and synching data to the cloud. This session will give an overview of how hardware is adapting to and changing the evolving open source digital health landscape. We will present a few case studies of the CliniPAK, including Kakuma Refugee camp, a health facility deployment in Malawi, and a disaster relief scenario. Multiple users can log on to the same instance of DHIS 2 and update data in real-time. By deploying DHIS 2 on the CliniPAK together, there is a possibility for a completely offline deployment that would enable data capture in the field as well as allow better service delivery, increase access, and improve outcomes. 
We had talked about some of these points last year, but we have a new version of the CliniPAK that has SIM-card connectivity, a diagnostic application to monitor CliniPAK functionality in the field, and a new design, which we'll present. 

 

Join Akros as they work through design principles and use-cases aimed at improving
government performance in understanding disease, education, and economic distribution at sub-national levels, responding to these conditions and calibrating service delivery and surveillance as this cycle progresses.

This session gives an overview of how the USAID funded IHPplus project in the Democratic Republic of Congo (DRC) contributed to improving the availability of routine health data by utilizing the MOH’s DHIS 2 integration in the project supported health zones. It also points out the importance of interoperability between MOH and project in-house DHIS 2 instances in order to stop parallel system and improve health zones capacity to collect and enter high-quality data into the MoH’s DHI2 system. The IHPplus M&E team could, therefore, better continue to focus on data analysis, reinforcing capacity to make evidence-based decisions, and improving the quality of project activity implementation. It also presents the ongoing effort to improve the quality of routine health data, data analysis, and use through DHIS 2.

There are times when DHIS2 users desire to have data visualization and analytic capabilities beyond the native features. For example the Ministry of Health in Zambia and PATH chose to partner with the Tableau Foundation to launch the Visualize No Malaria project and use Tableau’s software to improve decision making with the goal of eliminating malaria in Zambia by 2021. The success of this project has led to Tableau’s use in other global health applications with DHIS2. In this session you’ll learn what is possible for health analytics at a variety of levels and see two open-source integrations between DHIS2 and Tableau: First, a connector developed by BAO Systems that allows Tableau users to login to DHIS2 and get data. Second, an app for DHIS2 developed by Palladium Group, PATH, and StarSchema that embeds Tableau views and dashboards into DHIS2. These two integrations enable Tableau users to pull data from DHIS2, author visualizations, publish them, and then DHIS2 users can interact with those visualizations without ever having to leave DHIS2.
 

With the DHIS2 increasing being used to manage Individual records, and seeing significant deployments in fragile settings such as conflict zones and isolated countries, special attention needs to be placed on architecting systems for robustness and handling sensitive data.    This presentation will draw on experiences implementing DHIS2 as the National HIS in Sudan and Eritrea along with more targeted use cases in Palestine, Annexed Crimea and Northern Syria.  We aim to sensitize and inform on challenges, risks and deployment considerations for the DHIS2 platform, infrastructure and surrounding governance.

Several years ago, MSH began an effort to bring projects and systems across the organization into a single DHIS 2 instance with a hybrid architecture: Data Repository Engine for Analytics and Management, or DREAM@MSH.

This hybrid configuration was used to ensure that data collected would serve our staff and their needs across different levels of MSH:
1. in the field (for projects and in country offices)
2. at headquarters

MSH undertook an extensive discovery and design process to:
- explore and harmonize disparate data systems across the organization,
- configure the system with an emphasis on the ability to expand,
- satisfy project and context-specific needs,
- to provide a consistent experience for end-users.

This session will describe the four primary types of use-cases highlighted by the pilot projects, the broader needs represented by these projects, and the challenges posed by an attempt to create one system for these highly varying needs. 
 

Jhpiego, a global maternal and child health NGO affiliated with Johns Hopkins University, has been using DHIS 2 for almost two years. JADE – Jhpiego Analytic Data Engine is used in 27 countries to collect programmatic data. The primary audience for data analysis in JADE is headquarters. We are now faced with the challenge of also serving the needs of country offices separately. DHIS 2 for All encourages and supports Jhpiego country offices to build customized DHIS 2 instances. This is meant to replace the common practice of staff managing disparate Excel files and paper repositories of achievement data stored and shared from Dropbox, personal computers and in 3-ring binders while ensuring that the data is secure, permanent, accessible and ready to be visualized. One of the goals of DHIS 2 for all is to support country offices to collect in-country PMP data while also strengthening the relationship between the Jhpiego country office and MOH in sharing data.  The Jhpiego team will share its strategies and some early lessons learned in working with Liberia, Lesotho, and Mozambique in setting up DHIS 2 instances.

Programme and line managers at National and Provincial health departments in South Africa have not had exposure to DHIS 2 and are, therefore, not using it to analyze the data for both data quality and performance.  They do not know how to use the information for generating reports and for planning and implementing interventions to improve health.  In this session, Jenni Brown from HISP will talk about a training programme that was developed and implemented to use routine data in DHIS 2 in innovative ways to provide managers with a wealth of information on which to base their management decisions.
The aim of the program is to develop managers’ skills to use existing data and information in DHIS 2 for Evidence-Based Health Results Management (EBHRM) which includes Monitoring, Evaluation, Reporting & Feedback (MERF) to achieve health goals and optimize health program performance aimed at increasing life expectancy.

Through this programme, managers are being trained to apply existing Systems Theory, Logic and Result Frameworks and Epidemiological principles of prevention to categorize existing indicators in the South African routine information database into a results framework to inform EBHRM and MERF at all levels of the South African Health Care Services. Managers are also being trained on DHIS 2  reporting functionality using graphs, maps, and color-coded tables to populate a dashboard that displays the data according to the results framework principles taught in the course.

EBHRM training packages were developed for various focus areas:

• Reducing perinatal mortality
• PMTCT
• HAST
• Hospital services

Managers at national and provincial levels were trained and are being supported to use DHIS 2 and apply the results framework when planning, monitoring and reporting.This approach is being used as the method of choice to promote the use of information amongst programme and line managers in South Africa.  The training will be offered down to the facility level.