I’m interested in whether LLMs can assist in peer review, and in particular some of the technical aspects of peer review around things like statistical review. So every now and again I throw the following paper at an LLM - Mortality rates among patients successfully treated for hepatitis C in the era of interferon-free antivirals: population based cohort study because it has the reviews published and I can compare what an LLM says to high quality commentary from a statistical reviewer.
So, I had that paper lying around and today I’ve been reading about how good Claude 3.7 is. The headline examples have been about one-shot examples of Claude creating amazingly interactive games, so I decided to ask Claude if it could do something similar - but based on this paper.
It did.
Here are two things that Claude made for me.
1) an engaging visualisation of some of the findings of the paper https://claude.site/artifacts/e16c597a-2bb7–41d7–9499–883485224ad3.
2) an interactive simulation game where you have to keep 5 patients alive over a number of years by choosing the correct treatment plan, year by year. https://claude.site/artifacts/af74e36b-b5ac-4ac0–8d3e-8e7cb0b1de4c
Less than 10 minutes. That’s how much time it took to create both of these. From a highly technical paper. I don’t know if the content is correct. I think the game might be a little buggy. I just wanted a peek to see what this model could do. So there are a lot of caveats, but on the other hand this is nothing short of astonishing.
This was my initial prompt
The following is an academic paper looking at a population study. Read through the paper and then try to generate an intersting and engaging visulaisaiton or interactive game that teaches some of the points that come out through the paper. Be very creative in your approach.
It did the data visualisation immediately.
The game had a few glitches and I had to run the prompt 3 times. This is what Claude came up with in the end:
I apologize for the continued issue. Let me create a simpler version of the game that will work properly:
I’ve completely redesigned the Hepatitis C Patient Journey Game with a simpler, more focused approach. This interactive simulation puts you in the role of a healthcare provider managing patients who have been successfully cured of hepatitis C.
How the Game Works
- Meet Your Patients: You’ll guide 5 diverse patients through their post-HCV treatment journey. Each patient has different characteristics based on the research paper:
- Various liver conditions (no cirrhosis, compensated cirrhosis, or end-stage liver disease)
- Different risk factors (substance misuse, alcohol misuse, comorbidities)
- Personalized mortality risk profiles
- Choose Interventions: For each patient, you can select up to 3 interventions from a menu of evidence-based options:
- Liver Cancer Surveillance
- Addiction Support Services
- Alcohol Treatment Program
- Chronic Disease Management
- Comprehensive Follow-up Care
- See the Outcomes: The game simulates a 5-year follow-up period where your intervention choices directly impact patient survival. The simulation uses realistic probabilities based on the research data.
Educational Value
The game teaches several key findings from the research paper:
- Despite being cured of HCV, patients still face mortality rates 3–14 times higher than the general population
- Different risk factors multiply mortality risk (substance misuse: 2.9x, alcohol misuse: 1.9x, comorbidities: 1.3x)
- Drug-related causes and liver complications are the primary drivers of mortality
- A holistic approach to care that addresses both liver health and behavioral factors is essential
This interactive experience brings the abstract statistical findings of the paper to life, helping players understand the real-world implications of post-HCV treatment care and the importance of addressing all risk factors for these patients.