Self-harm in a clinical sample.
Research type
Research Study
Full title
Testing the effectiveness of an implementation intentions intervention to reduce self-harm in a clinical sample and testing the moderating effects of cognitive abilities
IRAS ID
261036
Contact name
Susan Rasmussen
Contact email
Sponsor organisation
University of Strathclyde
Duration of Study in the UK
1 years, 6 months, 0 days
Research summary
Research Summary\n\nThis study aims to (i) test the extent to which making implementation intentions reduce self-harm in the longer-term (3 months post-intervention) using objective outcomes (hospital readmission rates) in addition to self-reported outcomes (e.g., situation-specific self-harm behaviour) and (ii) explore the profiles of cognitive abilities within self-harm patients and identify which are likely to help explain why making implementation intentions reduce self-harm and suicide-outcomes for some patients but not others.\n\nThe intervention being used in this study is based on implementation intentions (Gollwitzer, 1999). Implementation intentions are “IF-THEN” plans which require people to link critical situations which tempt them to perform a behaviour (e.g. self-harm) and mentally link them with an effective behaviour strategy to overcome the temptation. In this study, participants will be invited to complete one of the following three interventions: a standard implementation intentions intervention, implementation intentions plus reinforcement or a control intervention. Previous research has demonstrated promising results for this intervention reporting a reduction of suicidal ideation and behaviour (Armitage et al., 2016). All adults aged 18+ who are admitted to the Forth Valley Royal Hospital for a self-harm episode (i.e., self-injury or self-poisoning regardless of the purpose of the act; Hawton et al., 2003) will be invited to take part. Participants will be invited to complete a baseline questionnaire followed by random allocation to one of the three interventions. This will be accompanied by the administration of cognitive tests. On completion participants will be thanked for their time and their contact details will be requested to permit the 3-month follow-up telephone interview. At follow-up the participants will be asked to complete the baseline questionnaire again and afterwards will receive a debrief.\n\nSummary of Results\n\nStudy summary \nSelf-harm is a major public health concern and one of the key risk factors for future suicide (Hawton et al., 2020). Interventions designed to reduce self-harm are therefore needed. The aims of this PhD study were to test whether an intervention encouraging the formation of implementation intentions could reduce self-harm in a clinical sample, and to identify if particular cognitive abilities (e.g., short-term - ‘working’ - memory) explained for whom the intervention was and was not successful. Implementation intentions (Gollwitzer, 1999) are “IF-THEN” plans that require people to identify critical situations (“IF” statements) which tempt them to perform a behaviour (e.g., “IF I am tempted to self-harm when I want relief from a terrible state of mind”) and mentally link those situations with goal-directed responses, or behaviour-change strategies (“THEN” statements), to help them overcome the temptation (e.g., “THEN I will remind myself about the benefits of not self-harming).\n\nAfter a mental health assessment conducted by the psychiatric liaison team, in the Mental Health Unit in Forth Valley Royal Hospital, the participants in this study were administered a questionnaire to obtain measures of demography (e.g., age and gender) and pre-intervention measures of self-reported self-harm behaviour and goal intention. Cognitive ability tests were also administered to obtain measures of working memory, attention (‘executive function’), long term (‘episodic’ and ‘retrospective’) memory, processing speed, and prospective memory (memory for future intended actions). After completing the pre-intervention measures, the participants were randomised to one of the following three conditions:\n\n(1)\ta standard implementation intention condition, where the participants were asked to form “IF-THEN” plans by selecting critical situations and goal-directed responses from pre-specified lists of IF and THEN statements, and linking them together\n(2) an implementation intention plus reinforcement condition, where the participants were asked to form “IF-THEN” plans, as above, and later re-state them or \n(3)\ta control condition, where the participants selected critical situations only from pre-specified “IF” statements \n\nA 3-month follow-up telephone interview was planned to obtain post-intervention measures of self-reported self-harm behaviour and goal intention, and hospital admission records were accessed to obtain objective data about subsequent self-harm. \n\nData collection began in January 2020 but had to be suspended six weeks later (March 2020) due to COVID-19. The study was finally terminated in January 2021 due to the continuing impact of the pandemic and the funding and time limitations associated with the PhD. We planned to collect complete (both pre- and post-intervention) data from a total of N = 318 participants. However, it was possible to collect pre-intervention data from 13 participants only before the study was terminated. Although it was possible to collect objective post-intervention data from these participants hospital records, none of the participants could be followed-up with telephone interviews. The data collected, therefore, did not allow the study aims to be addressed. For these reasons, this summary report presents only the key demographic information of the sample and reflections on the methodology.\n\nParticipant information\nThe majority of the sample were female (n = 9) with the remaining participants identifying as male (n = 4). The participants were in the following age categories: 18-25 (n = 4); 26-35 (n = 1); 36-45 (n = 5); and 46 years and above (n = 3). The participants were either married (n = 5), in a relationship but not married (n = 3), or single (n = 5). The three most commonly selected levels of employment status chosen by the participants were unemployed (n = 5), in full-time work (n = 3), or in full-time education (n = 2). Over half of the participants had children (n = 7). Two participants spent 0-10 years in education, 10 participants spent 11-20 years in education and one participant spent over 20 years in education. The highest level of education achieved by the majority of the participants was a high school degree or equivalent (n = 10). None of the participants reported experiencing any neurological disorders (e.g., history of stroke). Nine participants reported having a current/past medical diagnosis of a mental health condition, and included anxiety, depression and emotionally unstable personality disorder. \n\nInformation was collected about participants’ self-harm history including their current episode of self-harm which had resulted in the hospital admission. Four participants reported no previous hospital admissions for self-harm, seven participants reported between 1-10 previous hospital admissions for self-harm and two participants 30+ previous hospital admissions for self-harm. Eight participants reported they had previously self-harmed with the intention of taking their life. For the episode of self-harm that resulted in admission to hospital the most common method used by the participants was self-poisoning (n = 9). The other methods were self-injury (n = 2) and a combination of self-injury and self-poisoning (n = 2). For the self-harm episode which resulted in hospital admission, the majority of the participants reported the method undertaken was a way to end their life (n = 10). However, this was not the sole motive as half of the participants acknowledged other reasons for their episode of self-harm which included relieving stress, to feel better, difficulty managing emotions, to get help, and that they felt no pain when they cut themselves. Two of the participants had engaged in non-suicidal self-harm. The majority of participants did not leave a suicide note (n = 2) for the self-harm episode which resulted in hospital admission.\n\nFor all other measures from the questionnaire (e.g., self-reported self-harm behaviour and goal intention) and the cognitive measures of working memory, attention (‘executive function’), long term (‘episodic’ and ‘retrospective’) memory, processing speed, and prospective memory (memory for future intended actions) no inferences could be drawn. This is due to the extremely small number of participants recruited which, unsurprisingly, resulted in wide 95% confidence intervals being obtained.\n\nHospital re-admission data for the 13 participants who were recruited at pre-intervention was obtained; however, for the reasons provided above, no statistical analyses to identify differences between these groups could be conducted. Four participants were re-admitted to hospital for self-harm during the three months between pre-and post-intervention. Each of these participants were re-admitted to hospital on one occasion. Two of these participants were from the implementation intentions plus reinforcement condition, one was from the implementation intentions condition, and one was from the control condition. \n\nReflections on methodology\nWhilst the covid-19 pandemic terminated the study before any meaningful level of information could be collected, it did allow a test of the study protocols. Whilst the study was not explicitly designed for this purpose, two key points did emerge, concerning the recruitment procedures and the participants’ experience of completing the measures.\n\nWith regards to the recruitment procedures, a total of 36 potential participants were in Forth Valley Royal Hospital at the time the researcher was present. However, a final number of 13 participants were recruited. Common reasons for this were that the potential participants could not be approached by the researcher (e.g., not medically fit) or they did not fulfil inclusion criteria (e.g., English was not their first language).\n\nThe participant recruitment was originally designed to run for 15 months. Based on the level of recruitment prior to the suspension of the study (13 participants in 6 weeks), the target of 318 participants would not have been achievable. Approximately, 130 participants would have been recruited. One possible way to rectify this, which was being considered prior to the study suspension, would be to test participants at night-time (8pm – 6am) in addition to day-time (between 8am and 6pm). This would help ensure the recruitment of potential participants admitted during the night and discharged by the morning. Additionally, prior to any attempt to resume this clinical trial in the future, consideration could be given to recruiting participants from a hospital with a larger number of self-harm admissions, or use more than one hospital, in order to boost recruitment rates. Consideration should also be given to the use of more than one researcher, rotating between day and night shifts, to allow participant recruitment and testing to take place over 24 hours. This would mean that all eligible patients admitted to the hospital would potentially be captured by the recruitment procedure.\n\nRegarding the issue of how participants coped with the study measures, there was little evidence of problems with the completion of the questionnaire or intervention. That said, six of the participants asked “how much longer the study would last” or commented that they were “getting tired” whilst they were completing the cognitive tests. In the future, if this study were to be resumed, consideration might be given to reducing the number of cognitive tests or a break could be built into the study procedure, prior to the cognitive tests being administered.
REC name
North of Scotland Research Ethics Committee 2
REC reference
19/NS/0151
Date of REC Opinion
11 Nov 2019
REC opinion
Further Information Favourable Opinion