When people search on Trip we record – anonymously:
- Time and date
- Search terms used
- Articles viewed
Given Covid-19 is a hot topic I looked at our logs for Covid-19 related queries and found that we had over 500,000 separate search sessions exploring the topic. I’ve had a quick look through the data and found out a number of things that, to me, appear interesting.
Most popular articles
- COVID-19 (Novel Coronavirus) – DynaMed
- Novel coronavirus (COVID-19) guidance for acute care – Ministry of Health, Ontario
- Infection prevention and control for novel coronavirus (COVID-19): interim guidance for acute healthcare settings – Canadian Government
- Interim guidance: public health management of cases and contacts associated with novel coronavirus (COVID-19) in the community – BC Centre for Disease Control
- COVID-19 rapid guideline: managing the long-term effects of COVID-19 – NICE
Lots of data relating to the first time something was searched for, for instance:
- 14/01/2020 – the first search for a Covid-19 topic, in this case the search was ‘Wuhan coronavirus’
- 22/01/2020 – personal protection
- 25/02/2020 – vertical transmission
- 02/03/2020 – pcr
‘Timeline’ of a topic
Taking the top seven documents for hydroxychloroquine we charted their popularity over time:
After a very quick burst of interest, attention moved on (probably after people understood the evidence). Imagine if we’d waited 6 months for a systematic review!
Top search terms
- PPE masks
The above are a few examples of the sort of analysis that could be explored. Another one, not yet explored as it’s difficult to do manually, is to explore topics and see what types of evidence people look at. So, if a topic gets very few clicks does that indicate an evidence gap? If they click mainly on systematic reviews is that a good thing?
Overall, I could do the analysis and report it as a series of blogs. But I feel that there’s probably enough material to make at least one paper. Clearly, it’s not your typical journal article, but there are lots of different article types – so it might fit in somewhere!?
So, I’d appreciate any help thinking this through – for instance:
- What’s the best way of writing this up? In other words, what’s the story behind the date?
- What sort of journal might like this sort of thing? I was thinking of something like the BMJ (yes, ambitious) or Journal of Clinical Epidemiology.
- Any other analyses that might be interesting for people.
If you can help then please send me an email firstname.lastname@example.org