Health literacy tool

Our client (sydney australia) want to build, develop and implement an online tool to automate the application of health literacy design strategies to written health information from various organisations like Local Health Networks. Overall Goal- To allow organisations to quickly and effectively check (before publication) written health information is simple and straight forward enough for the majority of the greater population to understand thus creating more health literate documents for use by the general public. At the moment they dont have a single health literacy editor that can do all the things they need it to do – the closest is the Hemingway Editor ([login to view URL]) which utilises artificial intelligence and natural language processing, but this doesn’t do everything (most notably, it cannot flag words not in common usage and provide alternate recommendations for medical terminology eg. Anxiety vs worry or fear. scope: we are looking to develop an online tool like the Hemmingway app (editing tool) that can automatically assess various criteria in a written document such as : • Readability (using standard readability formulas) • Using everyday language (based on word frequency lists for top 1,000-3,000 words in everyday language) • Sentence length and structure • Minimal use of medical terminology / Consistent use of preferred medical terms ( list of preferred medical terms for some specific conditions eg diabetes). Ideally, the editor would be able to assess the criteria in real-time and have an easy-to-use, intuitive user interface. The client has identified several relevant sources that may help with scoping and design/development: • Open-source language checking / proofreading software (not all of these will easily operate in real time) o [login to view URL] o [login to view URL] o [login to view URL] • someone who’s reverse engineered Hemmingway App: [login to view URL] The client would also ideally like to incorporate an iterative design process with user testing (with health professionals within our existing networks at Western Sydney Local Health District). In summary the main features they would like to include by area are as follows: Criteria Readability Specific formula TBC; ideally this formula would treat bullet points as an independent phrase (i.e. ending in a full stop). formula -TBC. Similar to Hemingway app, this should provide a readability score for the whole document. Use of everyday language Based on word frequency lists – this would flag (highlight) words that are not listed in the supplied word frequency lists (ideally this could be toggled to top 1,000 words or top 3,000 words). No replacement word is required. Sentence length and structure e.g. sentences >20 words flagged (highlighted) If possible, flag sentences with >12 words AND >2 conjunctions If possible, flag paragraphs longer than 3 sentences. Minimal use of medical terminology / Consistent use of preferred medical terms If words in the CDC, (Public health) thesaurus are used these are flagged, with a popup that provides possible alternatives (based on CDC recommendations) Using lists (to be supplied) they have developed for specific conditions, flag when medical terms are used that has suggested alternatives. For something like diabetes this can mean providing simpler alternatives or consistency in use of terms such as physical activity, blood sugar, complications and lipids. We would not expect the prompt to replace the word in text, only to prompt the user to reconsider and use their judgement. If possible, it would be great to flag acronyms (perhaps identified at least 2 capital letters in a row) and prompt the user to either not use an acronym, or make sure it is defined first. Use of passive voice If possible, flag if passive voice used. thanks :)