Implementation intentions, cognitive abilities and self-harm

Overview
Suicide is the second leading cause of death among 15-29-year-olds and nearly 800K people die from suicide each year (World Health Organisation, 2017). A history of self-harm is the best predictor of eventual suicide. As is the case for many behaviours (e.g., Elliott et al., 2015), self-harm is a multifaceted problem (e.g., Rasmussen et al., 2012). However, it is well-established that critical situations (e.g., feeling hopelessness) act as triggers for this behaviour (O'Connor et al., 2013). Interventions that provide coping strategies for dealing with these critical situations are therefore required. Implementation intentions (IMPs) have recently shown promise in this regard.

Context
IMPs (Gollwitzer, 1999) are 'IF-THEN' plans that require people to specify critical situations that trigger unwanted behaviours (e.g., self-harm) and mentally link them with coping strategies (e.g., IF I am tempted to self-harm when I feel hopeless, THEN I will think about how my self-harming would upset my loved ones). The specification of the critical situation (IF) serves to encode a representation of that situation to memory. That mental representation is activated when the specified critical situation is encountered. The mental association between the specified critical situation (IF) and the coping strategy (THEN) serves to initiate a suitable response for ensuring that the unwanted behaviour is not performed when the mental representation of the critical situation has been activated. Essentially, the situations that typically prompt unwanted behaviours become triggers, instead, for desirable courses of action (e.g., avoidance of self-harm).

Aims
1. To test the extent to which IMPs reduce self-harm in the longer-term (6-months post-intervention) using objective outcomes (hospital readmission rates) in addition to self-reported outcomes (e.g., suicidal thoughts) and therefore extend the findings of Armitage et al. (2016). It is hypothesised that suicide-outcomes will be less prevalent in patients who have been asked to form IMPs than in patients who have not.

2. To explore the profiles of cognitive abilities within self-harm patients that are likely to help explain why IMPs reduce suicide outcomes for some patients but not others. It is expected that there will be variation in key cognitive abilities that govern attention and memory encoding, as well as long-term storage and retrieval.

3. To identify which cognitive abilities distinguish between patients for whom IMPs reduce suicide-outcomes and those for whom they do not. It is hypothesised that deficiencies in key cognitive abilities will distinguish between these two groups.

4. To develop ways of delivering IMP interventions that can help overcome the cognitive deficits that interfere with their effectiveness.

Methodology
A series of randomised controlled experiments will be conducted. All consecutive adult patients (aged 16 years or over) admitted to the medical wards of Glasgow Royal and Queen Elizabeth Infirmaries following a self-harm episode (intentional self-injury or poisoning; Hawton et al., 2003) will be eligible for participation in the research. The Consultant Psychiatrist at Glasgow Liaison Psychiatry Service has confirmed that access will be provided to the N > 1000 self-harm patients per year in these two hospitals, meaning that the sample sizes specified below can be achieved within the time-scale (power analyses confirm that the target Ns provide sufficient power).