Co-production and validation of an online resource to support the diagnosis of migraine

Most adults with migraine never consult their family doctor or receive a formal diagnosis. It is often these people that turn to the internet for healthcare information. Here we describe the development and testing of an on-line version of a previously validated telephone migraine classification interview.

We co-produced and tested the tool with people with migraine, clinicians, a web-design company and the National Migraine Centre, a UK based charity. We started with an online stakeholder meeting to understand what resources people wanted and any potential obstacles and facilitators to successful implementation. At a consensus meeting we used nominal group technique to gain consensus on the key questions to include. We used a 'Think Aloud technique' with people with migraine to explore the performance and acceptability of the questions in the new tool. To validate it, we asked people with headache disorders, including migraine, to complete the online tool before a doctor specialising in headache diagnosed their headache type. Level of agreement was measured between the outcome of the new migraine tool and the diagnosis by the doctor for 100 participants.

At the consensus meeting it was agreed that the questions for the new tool should be clear and easy to understand and distinguish between migraine, tension type headache, other primary headache disorders and medication overuse headache. For our initial validation exercise the level of agreement between the migraine quiz and the doctor diagnosis for 100 participants was 78% agreement. We made some adjustments to the wording and logic of the tool and repeated the validation exercise with 130 participants; the level of agreement was 80%.

Despite a careful development process our on-line tool did not perform to an adequate standard. There is a risk that misclassification could lead to people receiving inappropriate treatment. Any on-line classification tool for multiple headache disorders should be adequately validated before widespread use to avoid risk of iatrogenic harm.

Not applicable.