PS37 - Old and New Approaches in Preventing Suicide and Self-Harm

Smartphone-Based Monitoring to Guide Crisis Interventions
August, 30 | 14:00 - 15:30

In recent years, there has been a significant increase in studies using ecological momentary assessment (EMA) to study the digital phenotype of people with suicidal behaviour. EMA can be carried out in a direct or passive way; direct EMA involves asking questions about people's feelings, while passive EMA involves monitoring the person's behaviour using the smartphone's native sensors. Passive monitoring is of particular interest because it is non-invasive, does not interfere with the individual's daily routine and avoids the burden that active EMA can sometimes cause, which has high drop-out rates. In a recent publication (Barrigon et al), our group used this technology in a sample of 225 patients to predict short-term suicide risk by identifying changes in behavioural patterns characterised by real-time smartphone monitoring in a cohort of suicidal patients. Over a 6-month follow-up period, 8% of participants attempted suicide and 6.2% presented to the emergency department for psychiatric care. Behavioural changes identified by the algorithm predicted suicide risk within one week with an area under the curve of 0.78, indicating good accuracy. As a next step, we have designed a clinical trial in which the automatic detection of the crisis triggers a smartphone-based safety plan (previously designed by the patient). Preliminary results of this work will be presented at this symposium.

Speakers