Health Data Analytics for Managing Pandemics
How We Combine Data Analytics & Service Design Blog Series
The COVID-19 crisis has underlined why digital transformation in the public sector is imperative. Highlighting the value of big data and artificial intelligence, Budi Gunadi Sadikin (Indonesian Minister of Health) recently called for urgent reform and development of the country’s digital healthcare systems in line with the Ministry’s Digital Transformation Blueprint. This transformation requires a holistic approach to ensure that data systems are interoperable, and decision makers can access timely and comprehensive insights. With this vision in mind, UN Global Pulse through Pulse Lab Jakarta teamed up with the Indonesian Ministry of Health and the Ministry of National Development Planning (Bappenas) to explore how COVID-19 related datasets from across different systems within the Ministry of Health might be combined to inform the Government’s response to the pandemic.
In 2020, Pulse Lab Jakarta began working with the West Java Provincial Government to develop a COVID-19 risk assessment map. This interactive map combines traditional and non-traditional datasets to identify areas in the province that might have a higher risk for COVID-19 spread. Building on this experience, the Ministry of Health through its Digital Transformation Office was keen on understanding its varied datasets, and therefore requested data analytics support from the Lab. This presented an opportunity to leverage Pulse Lab Jakarta’s mixed-methods approach, which applies service design to support data analytics for decision-making.
A Mixed-Methods Approach
The collaboration kicked off with user research to better understand the needs of the relevant stakeholders within the Ministry of Health. Taking a thorough look at the end-to-end processes and resources required, the team also examined existing tools and relationships within the context of the Ministry’s data ecosystem. This process brought to the surface a shared interest in improving the utilization of the varied COVID-19 related datasets, and assessing the quality of data and interlinkages needed to support broader data harmonization in the health sector. The insights from this process helped the team to later determine the scope of the research dive.
Conducted virtually over a one-month period, several data experts from both the Ministry of Health and Bappenas came together to analyze pseudonymized COVID-19 data related to testing, tracing and vaccination. To complement the research, the team also utilized secondary data sources, such as population data, population projection data, Google mobility data, global COVID-19 trends and impact survey data, as well as situational data related to the Government’s social restriction measures. The three provinces that were examined included DKI Jakarta, West Java and Banten.
The research team was also joined by four domain experts from the Ministry of Health, who provided expert advisory on public health dynamics. The hackathon style adopted for this research dive helped with building a shared understanding amongst all the research participants on both opportunities and challenges. Following discussions with the Ministry of Health and a preliminary assessment of the datasets available, three research themes were identified:
1) epidemiological trends,
2) vaccine breakthrough cases, and
3) public health and social interventions.
Responsible Public Health Research
Issues related to data privacy and ethics were considered upfront in the research, particularly given the personal and sensitive nature of health data. Before diving into the datasets, we were intent on implementing a three-pronged approach:
1) Data pseudonymization, ensuring that all personal data being used in the research was masked to remove identifying information;
2) All parties involved in the research signed a non-disclosure agreement, detailing appropriate terms on data handling, data transfer and data security conditions to prevent potential breach/misuse; and
3) UNGP data ethics framework was incorporated, assessing potential risks, harms, and benefits related to the research questions. In particular, this framework helped the research team to identify and consider risks and harms that could potentially impact individuals or groups, either directly from the data itself or from the results. These risks and harms were then weighed against the potential benefits.
Greater datafication of citizens’ health data can help ensure that public health policies are evidence-based and inclusive. However, as the Ministry of Health’s Digital Transformation Blueprint emphasizes, this should be done responsibly with proper data protection, security and ethical considerations across the entire data ecosystem. Considering the Ministry of Health’s ongoing plans to develop an integrated Indonesia Health Services (IHS) platform, this collaborative research dive is an initial step to help the Ministry’s multidisciplinary team to concretely translate overarching policies, along with technical and non-technical guidelines for handling each dataset.
Looking ahead
In line with the overall objective of the research, we came away with a better understanding of the Ministry’s testing, tracing and vaccination datasets, particularly with respect to data access, data privacy, data quality and standardization. The research dive also allowed for the development of a common understanding on the applicability of the analysis to specific contexts, with practical knowledge transfers among the domain experts involved.
On the data analytics front, there were instances of incomplete, inconsistent and inaccurate data observed. For instance, some of the details related to certain priority groups (such as pregnant women, people with disability, the elderly, and people with comorbidities) were either i) captured only in specific instances, ii) captured incorrectly, iii) or recorded in an incomplete manner. It is likely that these instances may have occurred due to the lack of integration among the health data systems, and the absence of data standardization and shared protocols.
Improving data integration both within the Ministry of Health and with other national data systems ought to be the way forward. The digital transformation blueprint recognises this importance. In addition to data analytics, the benefits of applying service design are two fold. First, for the government stakeholders who will need to consume this data to be able to derive better data drive policies and actions, service design ensures that the data collection system is fit-for-purpose and a clear pathway can be established from data generation to actual decision making. Second, for the public which will interact with these systems, the user experience can be improved so as to ensure robust data capture and data quality, as well as help develop trust in the overall system.
Further discussions are underway on how to leverage the lessons and preliminary findings to strengthen the health data ecosystem in Indonesia for more effective policy making.
Authors: Agus Rachmanto (Deputy Chief Digital Transformation Office/ DTO, Indonesia Ministry of Health) and Desi Vicianna (Government Partnerships Coordinator, UN Global Pulse)
With editorial support from Dwayne Carruthers (Public Advocacy Manager, UN Global Pulse)
Pulse Lab Jakarta is grateful for the generous support from the Government of Australia