In writing this collaborative post between Esri Canada and Open North, we reflect on the data challenges Canada is facing during the COVID-19 pandemic. Large-scale collaborative solutions are required to improve intersectoral collaborations between geospatial data and open data experts, as well as between governments, non-profits and the private sector.
The COVID-19 pandemic is not just a health crisis, it’s also a data crisis. People are asking for better data on the pandemic and the media is echoing that call. In this blog post, we will explore how this cry for data around the pandemic highlights longstanding data issues, data mapping and visualization and how, despite the complexity around sharing data across jurisdictions, there is a path for making it easier to do and of more help to decision makers and the public.
The pandemic has forced governments and organizations to share data as never before; much more data is required and very quickly. It must be national, consistent across jurisdictions and respect privacy. The pressure of both the urgency and the high stakes of public health have exposed issues around collecting and sharing data. Why is healthcare data particularly difficult to share?
Most of the data about our population is based on census data. We have years of consistent census-based data and we can understand a lot about this country with that information. However, sometimes health reporting boundaries aren’t aligned with census boundaries.
Each province has its own health system structure and exposes data in a different way geographically. For example, some provinces/territories report the total number of COVID-19 cases only at the provincial level. Some count them at the health region level, and others report on individual cases with a finer focus, with health units.
Sourcing, comparing, validating and aggregating the data for each province, before mapping out the COVID-19 cases, requires a lot of detailed, manual work. To map out the situation of COVID cases nationally, someone has to comb through websites of all the ministries of health, understand their COVID-19 reporting level, obtain the boundaries of each health system and update the data daily.
Even when data is collected for each province, comparing infection rates in different provinces isn’t always possible. Ontario, BC and Alberta have similar case level information, meaning for each confirmed patient, we know their gender, age group and how they probably contracted the virus. We can compare those three provinces but other provinces such as Québec have different data collection and sharing practices. We can’t draw an accurate national picture because not every province or territory shares data the same way.
These are some of the reasons why it is so difficult to make simple comparisons across provinces in Canada and it’s so hard to ensure that data or statistics released by each province are comparable.
This can be partly explained by Canada’s separation of powers. Separation of powers means that the Public Health Agency of Canada (PHAC), a federal government agency, cannot enforce data sharing from provincial or territorial governments or even the private sector, limiting its ability to enact health surveillance.
Different levels of government also have different capacities to collect data. In the world of GIS and mapping, the differences between municipal-level data infrastructures and provincial and federal governments can be vast. Municipalities have different GIS capacity and practices and for some municipalities, a fully staffed GIS department is a luxury. Divergence in data publication practices can begin in the smallest of areas, such as how we tag data or define fields. This leads to challenges in searching for, combining and aggregating data. Ultimately, this makes it difficult for all governments to create a consistent set of facts and understanding for the public.
Creating national data standards is a good start. Data standards help structure data to make them consistent and reliable, which helps the recipient of the data comprehend, process and analyze them. This builds trust in the dataset and the organization responsible for maintaining it. But the standards landscape for healthcare in Canada is as complex as the healthcare system itself.
Standards can come from general standard-setting bodies, such as the International Standards Organization, or from international associations that represent specific parts of the industry such as the International Council of Nurses. They can come at the national level from Health Canada and the Canadian Institute for Health Information, or provinces, such as eHealth Ontario. In fact, Canada Health Infoway has a list of organizations that either define or recommend data standards for healthcare in each province.
There are so many ways to structure data from the healthcare system, and so many standards setters in Canada alone, that it’s clear that coordination remains a barrier to getting reliable and quality data aggregated from across the country.
Not knowing which data can and which cannot be freely shared has slowed down decision-making and biased it towards privacy protection at the cost of data sharing. This is especially the case for vulnerable populations, such as the homeless, or Indigenous peoples, where they often feel the government breathing down their necks and they want to hold onto any protection of their privacy.
To solve this issue, governments need to make the collaboration and coordination opportunities around health data more visible and accessible, provide resources for practitioners to provide standards. The public, as an end user of data, can also take part by providing our own thoughts and ideas on data visualization and analysis needs to government.
It is important for data providers, including government, to have a clear understanding of the eventual use cases for data. This allows them to share data in the appropriate formats and with the required level of detail and documentation. Understanding needs of end users is not a trivial task, particularly as public end users of data may have little interaction or influence on the original data producer. While public surveys and consultation can provide feedback on data needs, more interactive forms of engagement, such as co-design workshops between government and the public, can lead to more informed feedback on data, visualisation, user interface and analysis needs.
This type of co-design is already being practised as part of service design in the Government of British Columbia. Even without such collaborative workshops, anyone can engage with government on data and analytical needs for health data. Canada’s open data and open government communities provide a forum for discussion on data needs. Canada’s Multi-stakeholder Forum on Open Government is a forum for dialogue between the Government of Canada and civil society, established as part of Canada’s commitment to the Open Government Partnership. At the local level, open data and civic tech meetups are another great place to start getting engaged with developers, data scientists and researchers.
For certain kinds of data, citizen science initiatives and community asset mapping of health resources and programs can engage members of the public in the act of data collection itself. These initiatives are often led by a government, non-profit or academic organization and often provide basic training to ensure that anyone can participate. The act of collecting data in the field or making a map is a learning experience on coordination of multiple data collectors, ensuring accurate and reliable data, and ensuring that data entry is not error prone. These types of experiences can lead to increased public awareness of the challenges of generating good quality data.
Governments around the world are grappling with this. The European Union’s joinup is an example of a data and information sharing platform that allows anyone to share links to information, data and software resources around the world, with a specific repository for their Digital Response to COVID-19.
Because it is an open platform, government officials can share resources with each other, but also learn from those outside government. Communities such as joinup’s Community of Practice on Core Data Models, provides a platform for government officials to share data standards, new tools and applications to implement, with regular lessons learned shared data between government institutions.
In Canada, GCCollab is the Government of Canada’s platform for civil servants to connect with each other and share resources such as analysis workflows, code or papers. More mechanisms may be needed to facilitate sharing across jurisdictions. Organizations such as Canada Health Infoway were established and funded by the Government of Canada to promote interoperability solutions in the health sector.
But more is needed. A strong central standards setting body, strong collaboration and coordination, and stronger institutional mandates can help address our current fragmented health data within provinces, at the national level, and globally.
This post was translated to French and can be viewed here.
Open North is a non-profit organization based in Montréal. Operating since 2011, we work with public, private and research partners and community stakeholders to foster efficient, responsible, and collaborative use of data and technology to solve complex problems.
To learn more about Open North, visit www.opennorth.ca
About the AuthorMore Content by Guan Yue