[Conflicts of interest: The blogger (Katrina Witt) has previously collaborated on work with two of the authors of this study (KH and NK)].

Suicidology has long struggled with nomenclature, and the field of self-harm research is no different. Although the general public often views self-harm as synonymous with self-injury, and particularly self-cutting, clinical and emergency hospital staff are significantly more likely to encounter persons who poison themselves, either by toxic substances (e.g. pesticides) or medications (e.g. paracetamol), within their day to day work (Bergen H, 2010).

Given that clinical intervention research, service design and provision, and risk assessment and management practices are designed around those who present to clinical and/or hospital-based services, current guidelines for the management of self-harm may not adequately reflect the needs of those who engage in self-injury; and particularly self-cutting (Lilley R, 2008).

Additionally, as many people who engage in self-harm often switch between self-poisoning and self-injury (Lilley R, 2008), risk assessment and management practices which seek to predict suicide risk on the basis of method of self-harm at presentation to clinical services/hospital may be misleading, and potentially dangerous.

For these reasons, knowledge of the frequency with which people switch methods of self-harm between episodes, and the characteristics of those who switch methods, could have important implications for clinical service design and risk management practices.

Methods

A cohort study of all patients presenting to one of six hospitals located in Oxford, Manchester or Derby in the United Kingdom over a five year period (2003-07) was therefore undertaken to understand:

• How frequently patients switch methods of self-harm
• The factors associated with method switching
• The speed at which patients switch methods following discharge from hospital.

Using a standard patient record form, basic information on patient demographics (e.g. age, gender), clinical information (e.g. prior episodes of self-harm and the nature of these), and hospital management information (e.g. referral for outpatient psychiatric care) was collected for each patient. Additionally, information on the method of the self-harm episode resulting in admission was also collected.

To determine whether or not a patient had switched methods of self-harm, the authors developed the following five categories to describe the method of each episode of self-harm:

1. Self-poisoning (including both toxic substances and medications in excess of the recommended dose);
2. Self-cutting;
3. Other self-injury, severe (a heterogeneous category including hanging/asphyxiation, jumping from a height, carbon-monoxide poisoning, drowning, injury from gunshot wounds, and other potentially life-threatening injuries);
4. Other self-injury, less severe (a heterogeneous category including traffic-related injuries, and other less life-threatening injuries);
5. Combination of self-poisoning and self-injury within the same episode.

Results

Over the five year study period a total of 21,225 persons, 12,467 (58.7%) females, 8,755 (41.2%) males, and 33 (0.1%) whose gender was not recorded, presented to one of the six participating hospitals for treatment following an episode of self-harm on at least one occasion.

• A total of 4,721 (22.5%) persons engaged in a subsequent self-harm episode during the five year study period:
  • 2,820 (22.6% of the 12,467) females
  • 1,888 (21.5% of the 8,755) males
• Of these, just over one-third (1,620/4,721; 34.3%) switched methods at some point during the five year study period
  • 58.9% (954/1,620) switched methods one time
  • 24.6% (399/1,620) switched twice
  • 6.7% (108/1,620) switched three times
  • 9.8% (159/1,620) switched four or more times.

Factors found to be significantly associated with method switching on at least one occasion, included:

• Male gender;
• Younger age;
• A history of prior self-harm within the preceding year;
• Currently receiving mental health care as an outpatient.

The strongest predictors of switching, however, were those related to the method of self-harm at the index (i.e. first recorded hospital presentation) episode:

• Self-cutting;
• Oher severe self-injury;
• Other less severe self-injury;
• Combined self-poisoning and self-injury.

Factors significantly associated with switching of methods on multiple occasions were similar to those associated with method switching on at least one occasion, as were predictors of method switching based on the characteristics of the episode immediately prior to the switch in method (rather than on all previous episodes), with the exception of a history of self-harm in the preceding year, alcohol consumption, and gender.

Survival analyses were also undertaken to determine speed of method switching in those persons who repeated self-harm at least once during the five year study period. According to these analyses:

• Half of all method switches occurred within the first six months;
• Persons who engaged in self-poisoning at the index episode were slower to switch methods;
• Persons who engaged in self-cutting at the index episode and did not have a history of self-harm prior to the index episode were quicker to switch methods;
• Persons admitted to hospital on multiple occasions for self-harm were quicker to switch methods.

Conclusions

The authors conclude that self-harm is a fluid, changing, and changeable behaviour. Clinical decisions and risk management practices formed on the basis of method of self-harm at one point in time are therefore misguided, and potentially dangerous.

The authors reiterate the current UK National Institute for Health and Clinical Excellence (NICE) guidelines for the short-term management of self-harm (NICE, 2004), which state that all persons who present to hospital for self-harm should receive a psychosocial risk and needs assessment prior to discharge irrespective of the method of self-harm used.

Strengths and limitations

Although commonly reported in clinical practice, there have been few investigations of method switching between episodes of self-harm to date. This paper therefore addresses an important gap in the literature by assessing the frequency, speed, and characteristics associated with method switching over a relatively long observation period (five years).

Participants in this study were identified following presentation to hospital following an episode of self-harm. However, many more episodes occur in the community and do not come to the attention of clinical and/or hospital-based services . The participants included in this cohort, therefore, likely reflect more acutely ill patients who use more violent, and therefore lethal, methods of self-harm. Results of this study may therefore not generalise to those without significant medical/psychiatric diagnoses who engage in lower level forms of self-harm, such as self-scratching, skin picking, or hair-pulling.

The authors were also unable to account for psychiatric factors, such as diagnoses, substance misuse, and personality typologies, as determinants of method switching as these factors were not routinely collected by the six hospitals.

Summary

This study demonstrates that switching methods of self-harm between episodes is more common than previously thought; just over one-third of the sample (1,620/4,721; 34.3%) switched methods at least once during the five year study period. Clinical decisions and risk management practices formed on the basis of method of self-harm at one point in time are therefore misguided, and potentially dangerous.

All persons, irrespective of method, should be provided with adequate psychosocial risk and needs assessments prior to discharge, in line with NICE guidance for the short-term management of self-harm (NICE, 2004).

Links

Primary paper

Owens D, Kelly R, Munyombwe T, et al. (2015). Switching methods of self-harm at repeat episodes: Findings from a multicenter cohort study. Journal of Affective Disorders, 180, 44-51. [Abstract]

Other references

Staying safe if you’re not sure life’s worth living– to share hope, compassionate advice, practical ideas & links for people in distress http://www.connectingwithpeople.org/StayingSafe

U Can Cope 22m film and online resources – for people in distress and those trying to support them http://www.connectingwithpeople.org/ucancope

Bergen H, Hawton K, Waters K, et al. (2010). Epidemiology and trends in non-fatal self-harm in three centres in England, 2000 to 2007. British Journal of Psychiatry, 197, 493-398.

Lilley R, Owens D, Horrocks J, et al. (2008). Psychosocial assessment following self-harm: Results from the Multi-Centre Monitoring of Self-Harm Project. British Journal of Psychiatry, 192: 440-445.

McMahon EM, Keeley H, Cannon M, et al. (2014). The iceberg of suicide and self-harm in Irish adolescents: A population-based study. Social Psychiatry and Psychiatric Epidemiology, 49: 1929-1935.

National Institute for Health and Clinical Excellence (NICE) (2004). Self-Harm: The Short-Term Physical and Psychological Management and Secondary Prevention of Self-Harm in Primary and Secondary Care. National Clinical Guideline CG16. London, UK. British Psychological Society and Royal College of Psychiatrists.

Photo credits

• Joe Houghton CC BY 2.0
• Surian Soosay CC BY 2.0
• That Guy DouG CC BY 2.0
• Romana Klee CC BY 2.0
• Benjamin Ellis CC BY 2.0