It is natural for me to wonder how I can best contribute. As I “retired” from clinical practice some time ago, my skills as a clinician are probably not up to the task. However, there are probably some skills I can contribute, and I will consider those options going forward.
Fortunately there are some non-directly clinical contributions I can make, and these are keeping me busy here and now. One is teaching. While society is first and foremost dealing with the crisis at hand, we cannot put all education on hold. The situation is particularly challenging for medical students. One might think that the current crisis gives them the opportunity to learn on the front lines. The reality, however, is that there is not enough personal protective equipment (PPE) to protect them. As such, we need to find other ways to maintain their learning trajectory.
A number of medical educators have come up with innovative approaches, and I have thrown my own contribution into the mix. As one who teaches a well-known virtual course that is an introduction to biomedical and health informatics, we are packaging up an offering that we intend to make available as a medical school elective. Because the course is mostly asynchronous, we can scale it up pretty quickly. I don’t just want to throw the materials out there, and still maintain some sort of interaction and connection with learners, but we can offer the course to many students (including those beyond medical students). We plan to launch the first offering to Oregon Health & Science University (OHSU) medical students next week.
I also have an opportunity to advance research related to SARS-CoV-2 in the form of organizing an information retrieval (IR) challenge evaluation. The goal of this retrieval challenge is both to help develop systems capable of identifying relevant information for the current pandemic, but also to scientifically study how retrieval methods can be quickly developed for such situations in the future. The task will follow the "Cranfield" evaluation procedures that are used in the Text Retrieval Conference (TREC) and related challenge evaluations.
This effort is made possibly by work of the Allen Institute for AI and some collaborators who have assembled an open dataset, the COVID-19 Open Research Dataset (CORD-19). This collection of biomedical literature articles currently contains over 40,000 articles and will be updated weekly. Some colleagues and I will be organizing an IR challenge for search engines that find relevant COVID-related articles within this collection. This challenge will provide:
- A benchmark set of important COVID-related queries (e.g., coronavirus risk factors, COVID-19 ibuprofen)
- A set of manual judgments for CORD-19 articles on these queries
- An ongoing leaderboard for comparison of IR systems
We are even collecting candidate queries in a crowdsourcing manner by asking people to suggest them on Twitter using the hashtag, #COVIDSearch.
The current plan is to run the competition in weekly batches, where that week's snapshot of CORD-19 is used as the corpus and the results of systems participating in that batch are pooled for manual assessment. We will likely use the Kaggle platform to create a “leaderboard” of those whose methods are most effective. The challenge may in the future expand to more detailed tasks such as information-filtering, question-answering, fact-checking, and argument mining.
I make no pretensions that the work I am doing is in any way comparable to front-line healthcare and other essential workers, but I am glad that I can make these contributions that will keep education and research functioning during this tremendous worldwide crisis.