-
Notifications
You must be signed in to change notification settings - Fork 18
meeting 2021 04 28
Location: MS Teams
Attendees - Doug Manuel (Chair, uOttawa and Statistics Canada), Carolina Perez-Iracheta (Statistics Canada), Niels Nicholaï (uLaval), David Champerdon (PHAC), Chand Magnat (NML/PHAC), Audra Nagasawa (Statistic Canada), Bofu Li (Dalhousie U), Lev Kearney (NML/PHAC), Peter Vanrolleghem (uLaval), Kerry McPhedran (u Saskatchewan).
NML - National Microbiology Laboratory PHAC - Public Health Agency of Canada
Regrets - Howard Swerdfeger, Benjamin Trueman, Wiley Jennings, Claire Duvallet, Markus Brinkmann, Vince Pileggi, Aboubakar Mounchili, Heather Hannah
| Item No. | Item | Purpose | Speaker | Time (min) |
|---|---|---|---|---|
| 1 | Introductions | Information | All | 10 |
| 2 |
Status of ODM 1. What is in V1.1? 2. Who is using the ODM? 3. Who is not using the ODM? 4. Current development and collaboration process 5. Work in progress 6. International collaborations 7. Funding and resources |
Background | Doug and Niels | 20 |
| 3 |
Roadmap and priorities What ias been working well. What are the next priorities? |
Additional help is possible to take on tasks | All | 45 |
| 4 |
Next steps 1. Guidance on international collaboration 2. Terms of reference - coming! 3. Membership |
Discussion | All | 15 |
References
Scoping document: Ottawa Wastewater Data Model
Scoping document: Canadian Wastewater Surveillance Database
The meeting focused on the Ottawa Wastewater Data Model. Discussions are underway for the development of the Canadian Wastewater Surveillance Database. This meeting will focus on the ODM, but stay tuned for further discussions.
Doug noted that there are discussions regarding additions to membership of the Steering Committee to reflect suggestions from the last meeting. Stay tuned.
2.1. There are now stable versions of the ODM since version 1.0. Introduced in Version 1.1 was a range of small corrections for consistent naming convention. Version 1.2. is a lower priority, focusing on working groups to support the use of the ODM for generating dashboards, etc.
2.2. and 2.3. Organizations that use the ODM include Ontario and Québec wastewater-based surveillance programs that comprise approximately 150 sites of approximately 200 sites in Canada. CETO is a Québec commercial company that provides an information system for municipal wastewater systems. They use the ODM in their new application to support wastewater information and reporting. CETO supports both ODM import and export.
National Microbiology Lab is not fully using the ODM. Still, Chand and Lev (from the NML) indicated that this was planned. Further discussions are underway with the core ODM development team to integrate ODM into NML's laboratory information management system (LIMS). Bufo described wastewater data in the Maritime provinces and the potential benefit from ODM use. Still, those labs would require an additional introduction to the ODM and could benefit from the approach of "connectors" developed for Québec labs. Kerry described a similar situation in Saskatchewan where data is being shared to modellers but currently not in ODM format. However, there is interest in ensuring the Saskatchewan modellers can robustly collaborate with their interprovincial and national counterparts.
2.4. and 2.5. Work-in-progress by the core development team has recently focused on supporting the development of dashboards in Ontario and Québec (where most support for ODM development has occurred). Both of these provinces have databases in ODM format. Ontario MECP has shared the additional documentation, validation and data cleaning steps. These resources will be incorporated into the ODM repository in an open software format (Python) for general use. Québec "connectors" will be used for additional laboratory sites outside Québec - likely first piloted with Ottawa. Consideration is being given to a general application that any laboratory can use without Python programming. The general application will follow the recodeflow library developed by the uOttawa/Ottawa Hospital Research Institute or other applications such as Maelstrom.
2.6. Niel's recently organized and chaired a meeting of international wastewater data models, including representatives from the European Community, the United States Centre for Disease Control and Prevention National Wastewater Surveillance System (NWSS), NORMAN SCORE, and CovidPoops. To the knowledge of the attendees, there are dictionaries worldwide: ODM, NWSS and NORMAN SCORE. ODM and NWSS are similar in many regards - a reflection of similar objectives, program structure and collaboration. ODM is a fully open approach, and some common issues for both ODM and NWSS are discussed on the ODM GitHub site. NORMAN SCORE is a simpler approach with one Excel table and a process where sites can report and have their data added to the NORMAN SCORE data repository. Internationally, the Gates Foundation PATH program is developing an international database.
Internationally, there was a recognition of the value of collaboration. The next steps include: A comparison of the three main data dictionaries (by Niels). Ongoing discussions. Potential collaboration to generate a compatible minimal dictionary.
2.7. Additional resources for ODM are hopefully coming. This week, we've brought on additional help from one post-doc and two Ph.D. students. Additional funding may be available to us. The scoping document attached to the agenda is being used to guide discussions on priorities for ODM development — to be discussed at this meeting.
Discussion followed several themes:
3.1. Provide support for people who wish to use the data model but currently have not implemented it.
- Documentation. Including videos, meetings, how to.
- Templates and how to use them.
- Extra information such as validation.
3.2. Support labs that report findings. The onus on providing data rests on the labs. Make their job easy.
- Continue to develop and support Excel templates.
- Provide support for labs to create "connectors" between their information systems (often Excel spreadsheets) and the ODM.
- Create data validation and cleaning that the labs can use. Where possible, give lab warnings, not errors.
- Data flow to the ODM. Consider support upstream from ODM data templates.
- Consider adding definitions and metadata for assay methods, such as calibration curves. These definitions and metadata can then spur the development of a common data management system that would more robustly integrate with the ODM.
- Templates for generating calibration curves, etc.
3.3. Create added value to the ODM to make it "irresistible to use." Collaborate to create a common approach for tasks that are repeatedly performed by labs, database managers, epidemiologists and other analysts.
Examples include:
- Data cleaning
- Data transformation. i.e. how to create wide tables.
- Utilities such as methods for creating moving averages, visualization, simple trend analyses.
3.4. Create minimal data elements.
- Meetings about every three months.
- Terms of reference next meeting.
- Stay tuned for additional members.