ABOUT
What this is
Applied Evidence publishes plain-language summaries of therapy and nursing research. AI writes each one from the full paper, and no clinician reviews it before it goes up.
The problem it exists for
A practising clinician has a few minutes between patients, an abstract behind a paywall, and no realistic way to tell whether a paper should change anything. Most research reaches clinicians as a title and a conclusion, which is the part most likely to overstate what was found.
Applied Evidence reads the whole article instead and writes what a clinician needs from it: what was done, what was found, how much weight it can carry, and the ways it is most likely to be misread.
How a summary is made
Papers come from peer-reviewed journals in physiotherapy, occupational therapy, speech and language therapy and nursing. AI reads each paper in full and writes the summary. Automated checks then test the summary against the paper before it is published. Every number in a summary, for example, must appear in the paper. A summary that fails is held back.
What it cannot do
No clinician reviews a summary before it is published. The checks confirm that a summary’s figures match the paper. They cannot tell whether the right things were said, whether something important was left out, or whether the paper is any good.
It is not medical advice or a clinical decision support system, and it does not replace reading the paper. Every summary links to the original record. Check the source before anything changes for a patient.
The index covers roughly the last ten years of research. It is useful for keeping up with what is being published, and it does not replace a systematic search. The guides explain the labels on every summary, including what the certainty rating means and why nothing here is ever rated high.
Who maintains it
Applied Evidence is built and maintained by a practising occupational therapist, registered and licensed, holding a Master of Occupational Therapy and a certification in accessibility core competencies. The guides are written and reviewed by them and carry a byline. Everything else on the site is generated by software and marked as such.
It is independent. There are no advertisements, no sponsored placement, and no arrangement with any publisher, device manufacturer or professional body. It is funded by subscriptions.
Contact
Corrections, questions and anything else go through the contact form. If a summary has misread a paper, please tell us.
What a summary is checked for, and what it is not, is on the summaries page.