craigjh4.bsky.social
Work: Healthcare Epidemiologist / Study: MPH Part-time / Hobbies: Taylor Swift & Reading
🧠 Interested in: communicable disease epi / outbreak response / global health / inequalities / hospital infection prevention
🏴🏴🫶🏻
21 posts
611 followers
1,266 following
Regular Contributor
Conversation Starter
comment in response to
post
And here is my full list of books I read in 2024 with ratings, please feel free to send me recommendations you think I may like!
I do update Goodreads but mainly use Storygraph these days so feel free to follow/add me on there
app.thestorygraph.com/profile/crai...
comment in response to
post
I rated 9 books five stars, although my average rating was 4.29🌟 so high scores across the board (I think I’ve just gotten to know which books I’ll probably enjoy)
Going into 2025 reading Jurassic Park and listening to And The Band Played On 📚🎧
comment in response to
post
December 1st! Same as Christmas music 🎅🏻 you’ll use up all the Christmas spirit too early otherwise!
comment in response to
post
Hopefully this work provides more context to clinicians and those utilising LFTs to make better decisions and interpretations regarding their use and results
🧵8/8
comment in response to
post
My final thought, and not one that we picked up on in the paper but from my experience in hospital epi - think about how you are using them, is it really cost effective? Will you PCR anyway? What is the context - is there an outbreak? Could moving a possible case present risk to others?
🧵7/8
comment in response to
post
The study highlights the importance of prevalence in interpretation. High specificity can still result in low PPV when prevalence is low.These nuances are tricky but vital for interpreting results in different settings.
🧵6/8
comment in response to
post
Bottom line: Use LFTs to identify highly infectious cases during periods of high prevalence, but not as a standalone diagnostic tool. Negative results require careful interpretation, especially in asymptomatic or low-prevalence settings. Use them wisely, not blindly!
🧵5/8
comment in response to
post
Negative LFT results are reassuring but not definitive. Sensitivity is only 25.65%, so LFTs miss many PCR positive cases. For high-risk or symptomatic patients, a negative LFT shouldn’t rule out infection.
🧵4/8
comment in response to
post
High specificity (99.91%) means if you get a positive LFT, it’s likely to be a true positive, however low prevalence (0.43%) pulls the PPV down to ~56%, so only about half of positive LFTs are true positives in this dataset. A reminder: prevalence matters for PPV (and NPV)!
🧵3/8
comment in response to
post
So what did we find? Firstly, LFTs are highly specific (99.91%) but have low sensitivity (25.65%). So, LFTs rarely give false positives, but they miss many true positives.
🧵2/8