Math Epi, infectious diseases at National Taiwan University College of Public Health
Andrei R. Akhmetzhanov (吳亞克)
At the end of last year, we published our report on food poisoning outbreak of rare Bongkrekic acid (BA) that occurred in a restaurant in Taipei in March 2024, affecting 33 diners; 9 of whom were hospitalized and 6 died.
@robin.n.thompson
With long unmitigated SI, symptom-based isolation alone is unlikely to be sufficient. Rapid contact tracing and preemptive testing of close contacts are indeed important for controlling pertussis outbreaks.
Code & data: github.com/aakhmetz/Per...
This outbreak CFR was b/w historical pooled estimates from Indonesia (14%) and mainland China (40%). Only 1 of 7 ICU-admitted patients survived - no antidote for BA, only supportive care. All 33 cases consumed flat rice noodles from one batch left unrefrigerated >63 h.
SM github.com/aakhmetz/Foo...
Using Bayesian inference on linelist + restaurant order records, we estimated:
1) Attack rate: 60.3% (95% CrI 48.0–71.3)
2) Case-fatality ratio (CFR): 20.0% (8.6–34.3)
3) Time from dining to seeking healthcare: 3.4 d (hospitalized) vs 5.6 d (non-hospitalized)
A really nice paper with a very interestingly complicated dataset. Lots of modelling details to dig into in the SI.
Recently, Taiwanese government sent a scary book asking their citizens to prepare for war. A lot of money has been spent on that book and there QR codes in it, which would not be working in case of war. Anyway, let's wish the peace will remain
But when Hakuho won his tournament as a Shin-Ozeki, he was Mongolian
Key findings:
1) Unmitigated mean SI: 14.7 d (95% CrI 9.0–27.4) — close to the classic Dutch household estimate (te Beest et al., 2014)
2) Observed (post-intervention) SI: 10.4 d
3) Pre-intervention Reff: 1.4 → post-intervention: 0.45
4) Intervention efficiency: 65% (26%–88%)
Some pictures
In our recent EID letter: Serial Interval and Intervention Efficiency in Pertussis Outbreak, South Korea, 2024, we re-analyzed Cho et al.'s outbreak data to recover the unmitigated serial interval — what would be expected without case isolation.
wwwnc.cdc.gov/eid/article/...
@jdushoff.bsky.social
