Who runs this, and how

I’m S.R. The site is published by Big Wella, a small Australian research and education company I run. I’m not a clinician and don’t claim to be. People I love have ankylosing spondylitis, and I built this because the research that would have helped them was either paywalled, written for specialists, or buried under wellness marketing.

I don’t publish my full name on the site. Big Wella is a registered Australian business and the trail from there is public for anyone who wants to follow it. If you need to know who you’re dealing with, email contact@inflamed.community and I’ll tell you.

Three sources. A monthly automated search of PubMed, CrossRef and the bioRxiv and medRxiv preprint servers, using eight fixed queries covering the mechanisms of AS, over a rolling 60-day window. Suggestions from readers through the contribute form. And a small curated set chosen by hand when the map launched.

The automated search finds far more than belongs on the map: around 240 new papers in a typical two months. That is why the rest of this page exists.

Every paper is sorted before anyone writes a word about it. Declined on sight: single case reports, errata, comments and letters, conference abstracts, papers where AS is a comparison group for another disease, and anything from a source that doesn’t look like peer-reviewed research. Flagged as low priority: small correlation studies, questionnaire validations and translations, pilot studies, and reviews that add nothing new. Of the first 239 papers the search found, 56 went on the map.

The test for what stays is simple. Did the paper find something, or change how the disease is treated? Trials, long-term cohorts, registries, guideline updates, meta-analyses and genuine mechanistic findings pass. Papers that confirm what was already known, in a smaller sample, mostly don’t.

Declined papers are never deleted. Each one is logged with the reason, so a decision can be revisited and so a reader who thinks a paper was wrongly left off can ask.

Each paper on the map has a two-sentence plain-English summary and a one-line note on why it matters. The rules: say what was found, then what it means for someone with AS; use the drug names people recognise; be honest when a study is small, early or in mice; never hedge for the sake of it and never oversell.

AI is used to draft those summaries from the paper’s abstract. The draft is then read against the abstract and edited or rewritten by the editor before it is published. No summary goes live without a human approving it, and the published text is the editor’s responsibility, not the model’s. AI is never used on the public site itself: it does not answer questions, interpret your data, or talk to visitors. Abstracts are not reproduced on the map; the summaries are original text.

Where the AI draft and the abstract disagree, the abstract wins. Where the abstract and the full paper disagree, we may not catch it: summaries are written from abstracts, which is the honest limit of a one-person operation. If you have read the full paper and think a summary is wrong, say so.

Email contact@inflamed.community with the paper’s title or the page link and what you think is wrong. Corrections to summaries are made in place and noted on the changelog. Clinicians and researchers pointing out an error get the same treatment as anyone else: looked at, answered, and fixed if they’re right.

The site costs a few dollars a month to run and is paid for by the maintainer and by readers who choose to buy a coffee. There are no advertisements, no sponsored content, no affiliate links, and no pharmaceutical or supplement funding. Nothing is sold. No data is sold. The registry data belongs to the community that contributed it and is only ever shown in aggregate.

The site runs no analytics and loads nothing from third-party trackers. The patient registry never asks for a name or email, and the server does not store your IP address or browser details with your answers. The only identifier on a registry row is the removal code you are shown once. Any statistic with fewer than five contributors is hidden. The full write-up is on the privacy page.

It is not medical advice and does not give any. It does not tell you which treatment to choose, interpret your scans or results, or stand in for a rheumatologist. The research map exists so you can walk into your next appointment better informed, not so you can skip it.