Developing an algorithmic approach to detecting Enhanced Patient Observations on psychiatric wards and assessing their effectiveness in reducing the risk of self-harm and suicide

Suicide prevention is a national public health priority. In 2019, the National Health Service developed a plan to reduce self-harm and suicide among psychiatric hospital patients. However, there is limited evidence on the effectiveness and cost-efficiency of current interventions. One-to-one nursing, or Enhanced Patient Observation (EPO), is widely used to help patients feel safe and prevent self-harm, but its effectiveness remains unclear and some patients find it invasive. Using anonymized electronic patient records, we will analyze the characteristics of those who have experienced EPO, such as ethnicity or neurodivergence, and compare them to patients not under EPO to evaluate its impact on reducing self-harm and suicide, while accounting for differences between the two groups.