What are outcomes?
The effectiveness of an intervention is typically determined by the use of outcome measures, whether this be in research or in clinical practice. In research, outcomes allow for the determination of an intervention’s suitability for wider distribution into communities. In clinical practice, outcomes can contribute to practitioners’ understanding of service users’ wellbeing and whether it has improved or if extra support is required. Clearly, outcome measures can carry a lot of weight, so it is important that the target constructs are meaningful and relevant so that they accurately reflect the experience of ‘recovery’ for each individual.
Why are they problematic?
Within self-harm research, a reduction in the frequency of self-harm episodes is often prioritised as a primary outcome. However, recent research has begun disputing the validity of this. Owens et al [1] illuminated some key issues with this approach; firstly, measuring frequency or repetition of self-harm does not account for the medical severity of each episode of self-harm. If a person is self-harming less often, it may mean that the feelings and emotions that predicate self-harm compound for longer and make each instance of self-harm more medically severe. Consequently, a measure of self-harm frequency or repetition would erroneously signal improvement here. Similarly, focusing on a reduction in self-harm behaviours overlooks any other possible behaviours that may have arisen in its place. For example, Gelinas and Wright [2] found that students were able to self-harm less as they had started taking street drugs instead. In this way, the self-harming had been replaced with other behaviours that were also potentially dangerous and damaging. Again, this would suggest that the person’s experience of self-harm had not necessarily improved, yet a measurement of self-harm frequency would indicate that it had.
Another way in which self-harm outcomes can be problematic is that self-harm repetition is often indicated by attendances at hospital. The issue here lies in that not everyone who self-harms seeks medical attention for it. This is particularly true during the current pandemic, which has seen a significant drop in attendance at primary care for self-harm [3]. Further, as individuals have more experience of self-harm, they may also develop ways of managing their wounds themselves. This would potentially negate the need for medical attention as often. Consequently, relying on hospital admissions as an indication of self-harm repetition is not necessarily accurate and may underrepresent the frequency of self-harm occurring. This therefore further challenges the validity of using self-harm repetition as a primary outcome measure.
What is the solution?
Clearly, outcomes that are commonly used may not be entirely representative of the experience of ‘recovery’ from self-harm and may therefore be misguiding researchers or practitioners. Overall, it is apparent that re-evaluating self-harm outcomes is warranted. An essential step in ensuring that outcome measures are relevant and meaningful is to consult individuals with lived experience of self-harm, through the use of patient and public involvement (PPI) groups. As people with lived experience are the experts, their input and feedback can be helpful in all stages of the research process.
It is also important to remember that people who self-harm are all individuals with different feelings, experiences and needs. The heterogeneous nature of this population means it is difficult to develop a uniform set of outcomes that are relevant for everybody. Because of this, taking a more idiographic and personalised approach could be very beneficial. However, this causes issues in quantitative research as, for example, if different outcomes were measured for different participants, it would be difficult to do any formal, statistical comparisons across individuals or groups. This would limit any findings or conclusions that could be accurately drawn. As a result, qualitative research has a lot of merit with people who self-harm. However, qualitative research alone is not typically considered ‘enough’ to indicate the efficacy of an intervention that could be then generalised and used with a wider population of people who self-harm. Interventional research usually relies on the use of randomised controlled trials (RCTs) (e.g. [4]) for this purpose as they are generally regarded as the highest quality of evidence [5]. However, as discussed above, the scale and design of RCTs does not allow for much of an individualised approach. The value placed on RCTs does limit the use of other study designs that may be more appropriate in self-harm research. However, while it is important to reconsider the reliance on RCTs in self-harm research as a longer-term goal, there are alternative study designs that can be used in the meantime that align with the current frameworks. For example, there is a good case for using mixed or multiple methods, whereby an RCT or similar can be used to draw conclusions about the effectiveness of an intervention and its appropriateness for wider use, while supportive qualitative methods can be employed to gain a more in-depth understanding of whether the individual feels that the measures reflect their experience. There may of course be practical challenges that this would pose for large scale trials, for which it might be more feasible to gather qualitative data from a subset.
Importantly, mixed methods research as above would still be flawed if self-harm repetition / frequency is still used as a primary outcome. So, other outcomes must still be considered. This is something that House [6] discussed in some detail. They suggested that self-harm outcomes typically fall into three categories: symptoms (e.g., mood or hopelessness), functioning (e.g., social) and quality of life. I believe that coping ability should also be considered, perhaps within the function category, as a key outcome in self-harm research. There is a large body of evidence to suggest that self-harm can result from low coping skills, and that helping people to develop alternative coping strategies can lead to recovery from self-harm (e.g. [7]; [2]; [8]; [9]; [10]). Similarly, Dialectical Behavioural Therapy is considered one of the best forms of treatment for self-harm [11], and this has a strong focus on learning better ways of coping with difficult emotions. Finally, research (not yet published) I have recently conducted with university students who have experience of self-harm identified coping ability as a desired outcome from treatment for almost all participants. Following this, it seems that measuring coping skills could indeed be a good indicator of recovery from self-harm. However, it is again important to remember the heterogeneous nature of self-harm, meaning that coping skills wouldn’t necessarily be a relevant or meaningful outcome for everyone. This again emphasises the importance of working with PPI groups and/or employing qualitative methods in self-harm research.
Summary
- Outcomes in self-harm research need to be reconsidered.
- Coping skills may be an outcome worthy of consideration.
- Involving individuals with lived experience of self-harm in research design is essential to ensure that outcomes are relevant and meaningful for participants.
- A variety of study designs should be considered as RCTs may provide less meaningful findings.
References
1. Owens, C., Fox, F., Redwood, S., Davies, R., Foote, L., Salisbury, N., Williams, S., Biddle, L., & Thomas, K. (2020). Measuring outcomes in trials of interventions for people who self-harm: Qualitative study of service users’ views. BJPsych Open, 6(2). https://doi.org/10.1192/bjo.2019.93
2. Gelinas, B. L., & Wright, K. D. (2013). The Cessation of Deliberate Self-Harm in a University Sample: The Reasons, Barriers, and Strategies Involved. Archives of Suicide Research, 17(4), 373–386. https://doi.org/10.1080/13811118.2013.777003
3. Kapur, N., Clements, C., Appleby, L., Hawton, K., Steeg, S., Waters, K., & Webb, R. (2021). Effects of the COVID-19 pandemic on self-harm. The Lancet Psychiatry, 8(2), e4. https://doi.org/10.1016/S2215-0366(20)30528-9
4. Witt, K. G., Hetrick, S. E., Rajaram, G., Hazell, P., Salisbury, T. L. T., Townsend, E., & Hawton, K. (2021). Interventions for self‐harm in children and adolescents. Cochrane Database of Systematic Reviews, 3. https://doi.org/10.1002/14651858.CD013667.pub2
5. Burns, P. B., Rohrich, R. J., & Chung, K. C. (2011). The Levels of Evidence and their role in Evidence-Based Medicine. Plastic and Reconstructive Surgery, 128(1), 305–310. https://doi.org/10.1097/PRS.0b013e318219c171
6. House, A. (2020). Measuring outcomes in self-harm trials: What is important and what is achievable? BJPsych Open, 6(2). https://doi.org/10.1192/bjo.2020.5
7. del Carpio, L., Rasmussen, S., & Paul, S. (2020). A Theory-Based Longitudinal Investigation Examining Predictors of Self-Harm in Adolescents With and Without Bereavement Experiences. Frontiers in Psychology, 11. https://doi.org/10.3389/fpsyg.2020.01153
8. Hambleton, A. L., Hanstock, T. L., Halpin, S., & Dempsey, C. (2020). Initiation, meaning and cessation of self-harm: Australian adults’ retrospective reflections and advice to adolescents who currently self-harm. Counselling Psychology Quarterly, 0(0), 1–24. https://doi.org/10.1080/09515070.2020.1737509
9. McLafferty, M., Armour, C., Bunting, B., Ennis, E., Lapsley, C., Murray, E., & O’Neill, S. (2019). Coping, stress, and negative childhood experiences: The link to psychopathology, self-harm, and suicidal behavior. PsyCh Journal, 8(3), 293–306. https://doi.org/10.1002/pchj.301
10. Wal, W. van der, & George, A. A. (2018). Social support-oriented coping and resilience for self-harm protection among adolescents. Journal of Psychology in Africa, 28(3), 237–241. https://doi.org/10.1080/14330237.2018.1475508
11. NICE. (2013). Self-Harm (Quality Standard). https://www.nice.org.uk/guidance/qs34/resources/selfharm-2098606243525
Bethany Cliffe (@bethanyjcliffe) is a PhD student at the University of Bath (bc731@bath.ac.uk).
Featuring photo by Kate Kalvach on Unsplash.
netecr.org 