Informed consent readability checker for plain-language forms

Readability formulas estimate how hard a passage is to read from two surface features: how long the sentences are and how long the words are. The Flesch reading ease score is 206.835 minus 1.015 times the average number of words per sentence, minus 84.6 times the average number of syllables per word. A passage of short sentences with short words scores near 90 or above and reads as easy; a passage of long sentences with polysyllabic medical terms can score below 30.

The Flesch-Kincaid grade level uses the same two inputs with different weights: 0.39 times the average words per sentence, plus 11.8 times the average syllables per word, minus 15.59. The result approximates the US school grade at which the text can be read comfortably, so a grade of 8 means about the level of a 13- or 14-year-old reader. The checker also reports the average sentence length and the share of long words, because those are the two levers you actually control when you edit: split a sentence, swap a long word for a short one.

The tool counts words, sentences and syllables from the text you paste. Sentences are found by terminal punctuation, and syllables are estimated with a simple vowel-group rule. That works well for ordinary English prose and poorly for abbreviations, numbered lists, dosing tables and drug names, which can distort the result. If you paste a list of procedures with no full stops, the tool may see one very long sentence. For the best reading, paste a paragraph or section at a time and rewrite any lists as complete sentences before measuring.

Start with sentence length. Long sentences carrying several conditions are the most common reason a consent paragraph scores badly. Aim for an average near 15 words and break any sentence over 25 words. Put one idea in each sentence and lead with the thing the participant needs to know or do.

Replace long and unfamiliar words with short and common ones where meaning is unchanged: "use" for "utilise", "tests" for "assessments", "side effects" for "adverse reactions". Keep the medical term where the participant needs it, explain it once in plain words, and use it consistently. Use "you" and active voice: "We will ask you to visit the clinic four times" reads more easily than "Participants will be required to attend four clinic visits".

Structure helps even where scores do not change. Use headings that are questions people ask ("What will happen to me?", "What are the risks?"), short bullets for lists of procedures, and a summary at the top. Then test with real readers, using teach-back: ask a few people from your target population to explain the study in their own words. Patients and public contributors are valuable reviewers for exactly this.

When you move from paper to screens, the same rules matter more, because people read on a phone. Capture's eConsent lets you present consent in sections, with signature and countersignature steps and an audit trail; see eConsent software for clinical trials, the guide on how to write an eConsent form, and the consent templates at /templates/consent as a starting point. If you are replacing paper forms, how to digitise clinical trial consent forms covers the process.

Take this sentence from a draft: "Participants will be required to attend the clinic for a total of four visits, during which a number of assessments, including blood sampling and electrocardiography, will be undertaken by study personnel." It has 31 words and scores as difficult. A plainer version reads: "You will visit the clinic four times. At each visit, we will take a blood sample and record your heart rhythm." That is two sentences of 8 and 14 words, mostly short words, and it moves the score by tens of points. Nothing was dropped: the number of visits and the procedures are still there.

Do this for the worst paragraphs first, using the scores to find them. Then read the text aloud. Wherever you stumble, the participant will too.

Readability formulas were designed for ordinary English prose. They treat every polysyllabic word as hard, so a short, common medical word of several syllables can drag a score down, and they treat short but unfamiliar words as easy. They ignore layout, font size, vocabulary familiarity, cultural fit and whether the content is accurate or complete. They are not designed for languages other than English, and the syllable count in this tool is an approximation. Results may differ by a grade or so from other tools, because they count syllables and sentences slightly differently.

The checker is a drafting aid, not validated software, and it cannot replace review by your ethics committee or institutional review board, a plain-language editor, or a patient reviewer. The final consent text, and any translation, must be approved by the body that oversees your study. Do not paste real participant data into any tool; consent templates should contain no personal information.