During a three-month period, the study subjects will be using an electronic diary app that they can download and install on their smartphones to register the occurrence of seizures and also to register the potential alerting behaviour of their dogs. This information will increase the knowledge of the accuracy of these alerts.

This is a international study by Ghent University that will initially involve people from Belgium, Italy and the UK.

Previous studies of alerting dog accuracy are either case studies or retrospective studies that rely on memory recall of information about alerting behaviour, with the risk of memory bias that reduces reliability.

The use of electronic diaries has been described as more accurate and has better compliance than paper ones. Nowadays, electronic diaries offer the possibility of receiving live information and/or timed answers, thereby increasing compliance and decreasing the incidence of memory bias. This would be the first time that "real time" information about dog behaviour and seizures in the same patients is collected systematically. This will facilitate the analysis of the specificity and sensitivity of alerting behaviour as perceived by the owner.

During a three-month period, systematic data on the occurrence of seizures in the patient and on alerting behaviour by the dog will be collected. This information will increase the knowledge of the type of alerting behaviours that the dogs display and will allow to calculate the specificity and sensitivity of dog behaviour interpreted by the owner as predictive of a seizure. Depending on availability of trained and untrained dogs, possible differences in the alerting behaviour between both types of dogs and in their alerting specificity/sensitivity will be studied.To avoid recall bias the information will be received live through a back-end.

To be able to participate, the volunteers will grant the investigators permission, via the informed consent, to contact their neurologist to confirm their epilepsy diagnosis and to send monthly reports of the electronic diary (see below) to their neurologist.

It will be stated clearly in the instructions and the informed consent that the purpose of the app is to collect data to investigate dog behaviour and that the app is not intended as a monitoring tool for the patient. In the instructions and informed consent the investigators will be insisting that the volunteers only use the app when they are in a safe position or/and have recovered fully from the seizure.

The documents will also include a recommendation to contact their neurologist should a sudden change in the frequency or duration of the seizures occur. In addition, the research team will provide monthly reports of the volunteers' electronic diary to their treating neurologists. Before the start of the study, Prof. De Herdt will contact the treating neurologist to inform them and to confirm the epilepsy diagnosis.

Electronic Diary

A smartphone app for Android and iOS will be developed with a main focus on user-friendliness. The App's main screen will display two buttons. The first button is to add alert events by the dog (= behaviour interpreted by the owner as an alert), i.e., to indicate that the dog has alerted. The second button is to add a seizure event (= to indicate a seizure is occurring or has just occurred). A calendar will show the alert and seizure entries on each day. When one of the two buttons is pressed, the time at which this occurred will be registered. It will not be possible to change the entry time, but the patient will have the option to indicate if the seizure happened at a time different from the time of the entry by accessing the seizure entry on the calendar. Once 24 hours have passed, the seizure entry can no longer be modified.

When the patient does not enter any seizure for a whole day, a notification will appear the morning after to ask to confirm that there were no seizures during the previous day. Extra information about the presence/absence of auras and/or prodromi can be entered during the 24 hours after a seizure occurred, by accessing the seizure entry on the calendar. Data introduction and storage will be possible even without internet connection. The data will be uploaded to a database as soon as the phone is connected to a Wi-Fi signal. Compliance will be monitored and patients that are not filling in information regularly (at least once a week, but depending on the frequency with which they experience seizures) will be contacted to verify whether there are any problems using the App and to resolve potential problems.

Data management and informed consent

The collected data will be treated in accordance with the General Data Protection Regulation (EU) 2016/679. The back-end will be SSL certificated and the data will be processed only via the Athena platform and will never be stored or processed locally. The data will be anonymized for analysis, and the database of patients will be coded and stored also in the UGent central IT environment. The database will only be accessed by the supervising team from Ghent University.

The patients will have the right to request information about the state and usage of their data and to request to modify or remove their data from the database at any point. All personal information that could allow us to identify the patient will be destroyed at the end of the study.

Inclusion Criteria:

• The volunteer has been diagnosed with epilepsy by a neurologist and lives with a dog that anticipates the occurrence of seizures (trained or not)
• Volunteers should be able to operate an electronic diary app on a smartphone by themselves or will have an assistant/carer with them 24/24 that will operate the device for them.
• In the case of children, the parents/tutors will fill in the diary for them.
• In the case of people with learning disabilities, carers or familiars will fill in the diary for them.

Exclusion Criteria:

- People with non-epileptic seizures, that are not able to use a smartphone's app or that can't provide us with a confirmation of their diagnosis.