A readable IFU is not necessarily an understandable IFU

Looking beyond readability: What ISO 20417:2026 means for Human Factors.

ISO 20417:2026 has introduced a very interesting requirement for medical devices intended to be used by lay users. The new standard says instructions for use that are meant for lay people should be written at a lower secondary education level, or at another education level if that is more appropriate for the intended user and is specified in the use specification. It also requires that the IFU be delivered in a format that is appropriate and accessible to the user.

At first, this might look like just another labeling requirement. But there is a much bigger Human Factors discussion behind it. One of the ways companies might try to show compliance is by using readability tools. For example, you could run your IFU through something like Flesch-Kincaid, get a reading grade score, write it down, and move on. There is nothing wrong with doing that. Readability tools do help spot long sentences, complicated language, and unnecessarily complex terms.

But readability and understanding are not the same thing. You can have a sentence at an eighth-grade reading level and still have people misunderstand what they are supposed to do.

Imagine an instruction that tells a patient not to use a drug if the solution appears cloudy. The sentence is simple enough. But does the patient know what cloudy actually means? Can they tell the difference between an unacceptable solution and a normal variation? Do they know what to do if they are not sure? Will they even look at the solution before using it? A readability score cannot answer those questions. That is where Human Factors comes in.

Medical device labeling is part of the user interface. The user reads it, interprets it, makes a decision, and acts. At any point, a problem can result in a use error. ISO 20417 connects to this idea directly. The standard says that both the ISO 14971 risk management process and the IEC 62366-1 usability engineering process should be used when figuring out what information needs to be provided, including information that relates to safety. It also says information must be put in terms the intended users can understand, taking into account their education, training, and special needs.

For me, that is the most interesting part of the requirement. If you have experience in Human Factors, you already know a user can understand an instruction and still not follow it. They might never notice the warning. They might find the information too late. They might understand every step but perform them in the wrong order. You could interview a user and they can explain what they are supposed to do, but then behave differently when actually using the device.

That is why I would be careful about turning ISO 20417 into another documentation task. I do not read it as saying that every lay-user IFU now requires a full Human Factors validation study. There is nothing in the wording that prescribes Flesch-Kincaid, SMOG, or any specific readability evaluation.

But it does raise an important question about evidence. If an instruction is only descriptive, the evidence you need might be limited. If that same instruction is being used to prevent a serious use error, the bar is much higher. The more your safety depends on labeling, the more important it is to know if users can actually find, understand, and use that information correctly. ISO 20417 makes that especially clear when reading the IFU is the main way to avoid risk. If you depend on the user reading and understanding the instructions to prevent something unsafe, the effectiveness of that information is critical.

That is where different methods can work together. Readability evaluations can tell you if the language is too complex. Comprehension studies can show if users interpret the information the right way. Formative Human Factors studies can reveal if users can find and use the information during realistic use. And when labeling supports critical tasks or crucial risk controls, Human Factors validation may be needed to show the user interface, including the labeling, is up to the job. The right method depends on the risk and the question you are trying to answer.

The education level of the user should not be an afterthought or something you figure out when the IFU is finished. The standard allows for using another education level if that is what makes sense based on your use specification. That means you need to have this conversation up front. Who are the users. What education can you assume. What health literacy or medical knowledge do they have. Are they familiar with devices like this. What information do they actually need to make safe decisions.

These are Human Factors questions, and they should shape the information from the beginning. Too often, labeling is developed alongside the device, not as part of the user interface itself. The device is mostly finished, Human Factors work goes on, risks are found, and the IFU is left to patch whatever remains. That is not a strong approach.

The ISO 20417 requirements for lay users are a reminder that information must be designed for the people who will actually use it. My hope is that companies do not treat this as just another reason to run the IFU through a calculator and add a report to the file.