Applying "user testing" resulted in a largely health-literate summary suitable for people across a range of backgrounds.

Increasing amounts of information are being made available to patients but how do we know Iwe describe how we have employed user testing to test and improve not just information for patients, but also for professionals and others. This is built on the many years of using this technique, first at the University of Leeds and then through the spin-out company Luto Research (www.luto.co.uk). Key words:User testing to test whether the information actually works can they find and understand the information they need? Luto Research is a university spin-out company which develops, refines and tests health information materials. In a recent test at Luto Research, the participants included a retired cleaner, an unemployed person, a stand-up comedian, and a bus driver representative of the real people who have to be able to use and understand the information we produce.Page 5 How is user testing performed?User testing is a defined process originally developed in Australia by Professor David Sless 3 , and the key steps are described in Table 1. An important point to note is that the participants are potential users of the information. If people are already familiar with the topic the document is describing, that would give them prior knowledge which they could draw on, and not just rely on the information in the leaflet to answer questions. User testing mimics the situation when someone first encounters a particular treatment or health issue, and receives information about it.[Insert Table 1 near here] Also important is that, when applied properly, user testing is an iterative process (Figure 2). The document is drafted and then tested, generally with a round of 10 people. The results are then assessed, bearing in mind that not all feedback from participants can be taken forward; indeed sometimes the feedback can be contradictory. After this careful analysis of the data, good practice in information writing and design is applied to amend the document and it is then tested in another round of 10 people. Crucially, the testing itself does not improve documents;it is the application of good practice between rounds which is the key skill. The use of small numbers in the testing often raises the question How can you test something on just 10The answer is that user testing is a form of diagnostic testing finding out where documents do not work, and remedying problems using expert information writing and design practice. Our experience, in over 20,000 individual user testing interviews that Luto has carried out, is that if there is a significant problem with a document, this will become apparent in the Page 6 first two or three D S you do not need a representative sample to find a creak. 4 Remember, though, that in user testing it is the people who are testing the information; we are not testing the people. This has to be stressed at the beginning of each interview we want them to find the weaknesses in the document. How does this relate to health literacy?It could be argued that the people least suited to assessing t...