Director Pandemic and Epidemic Intelligence Systems @WHO #WHOPandmicHub #globalhealth #surveillance #riskassessment #healthemergencies #outbreaks
Berlin, Germany
Joined January 2018
- Tweets1.8K
- Following214
- Followers3.6K
- Likes1.3K
AI reads Gram stains with over 95% agreement in controlled settings.
But most infectious disease models rest on retrospective, single-centre data, and few show better prescribing or patient outcomes. The gap is evidence, not algorithms.
Full digest: aipublichealth.substack.com/…
An LLM gave safe exposure limits of 5 to 300 mg/day for one chemical across runs.
Unreliable? Partly. But toxicologists disagree on this substance too.
The authors' answer: LLMs search, extract and summarise; experts make the call.
Full digest: aipublichealth.substack.com/…
1/ An AI eye screening trial in Tanzania measured something most AI studies don't: whether patients actually reached care.
Giving people their diabetic retinopathy result on the spot roughly doubled referral attendance. 🧵
4/ The caveats matter. The AI caught 91.6% of referable cases, but lower specificity meant more false-positive visits to a stretched clinic.
And the trial can't separate the AI from the counselling. The effect belongs to the whole pathway.
5/ Most people with referable disease still didn't attend, and health insurance strongly predicted who did. AI changed the timing of care, not its cost.
Tanzania now plans to expand AI-supported screening.
Full digest: aipublichealth.substack.com/…
#APHIDigest
Can survey data tell us which infants will miss their measles vaccine before they do?
In Uganda and Kenya, ensemble ML reached AUCs of 0.88. Child's age and vitamin A mattered in both. Other drivers differed, so outreach should too.
Full digest: aipublichealth.substack.com/…
R² up to 0.998.
That's the best PM2.5 forecast accuracy across 14 air quality early warning studies in South-East Asia.
Yet most had a high risk of bias. Without external validation, high accuracy is not a reliable warning.
Full digest: aipublichealth.substack.com/…
This week's AI + public health journal digest is out.
An AI eye screening trial in Tanzania, air pollution forecasting, missed measles vaccines, LLMs in toxicology and legal epidemiology, and AI in infectious disease.
Full digest: aipublichealth.substack.com/…
#APHIDigest
By 2030, we want every new health threat identified and understood within 24 hours of emerging, anywhere in the world.
We've already done it many times. We can't yet do it reliably, everywhere.
I spoke with @Devex about how we get there: youtube.com/watch?v=pxoouVSb…
Structure outlasts sequence. That's reshaping how we study viral evolution.
With AlphaFold, 4.4M viral protein structures now span 8 databases, and 60%+ of their folds match nothing known in cellular life.
A new lens for surveillance.
Full digest: aipublichealth.substack.com/…
Pregnant women, new mothers and children are left out of most drug research. So how much do we actually know about the medicines they're given?
A new study mapped the gap across 38 million PubMed abstracts. The answer is sobering. 🧵
3/ The scale of the gap:
22-66% of drugs given to pregnant or postpartum women lacked pharmacokinetic or epidemiology/trial evidence.
8-40% of drugs used in children did too.
Missing pharmacokinetic data was the most common gap of all.
4/ AI used well: not to replace the evidence, but to show us exactly where it's missing.
The results are now the MPRINT-KP Silver knowledge portal, with a call for more pharmacokinetic studies in these groups.
Full digest: aipublichealth.substack.com/…
#APHIDigest
We test medical AI for accuracy. That misses the real problem.
Sex and gender bias doesn't show as wrong answers. It hides in tone, word choice and clinical detail, and it runs both ways, not just against women.
Full digest: aipublichealth.substack.com/…
Can a routine blood count help triage dengue where lab confirmation is slow?
Tree-based ML on 1,523 patients hit 0.709 AUROC, 91% sensitivity, 46% specificity.
Useful for triage in endemic settings, not a replacement for testing.
Full digest: aipublichealth.substack.com/…
79%.
That's the share of predictive AI models for pharmacovigilance in the Asia-Pacific with no external validation, across 64 studies from 14 countries.
A model never tested outside its own data is a prototype, not a safety tool.
Full digest: aipublichealth.substack.com/…
This week's AI + public health journal digest is out.
9 papers: pharmacovigilance, dengue diagnostics, gender bias in clinical AI, maternal drug-evidence gaps, viral evolution and more.
Full digest: aipublichealth.substack.com/…
#APHIDigest
