Article Text

other Versions

Sometimes, not always, not never: a response to Pickard and Pearce
  1. Patrick J Sullivan
  1. School of Law, University of Manchester, Manchester, UK
  1. Correspondence to Mr Patrick J Sullivan, School of Law, University of Manchester, Oxford Road, Manchester, UK; Patrick.Sullivan-2{at}


This paper provides a response to Hanna Pickard and Stephen Pearce’s paper ‘Balancing costs and benefits: a clinical perspective does not support a harm minimisation approach for self-injury outside of community settings.’ This paper responded to my article ‘Should healthcare professionals sometimes allow harm? The case of self-injury.’ There is much in the paper that I would agree with, but I feel it is important to respond to a number of the criticisms of my paper in order to clarify my position and to facilitate ongoing debate in relation to this important issue.

  • autonomy
  • paternalism

Statistics from

I would like to thank Hannah Pickard and Stephen Pearce for their informed and clinically focused response to my paper on harm minimisation.1 They argue that harm minimisation is not a viable or responsible approach within an inpatient environment.2

Given some of the media attention attracted by my paper,i it is important to be clear at the outset that I am in no way advocating the use of harm minimisation as a routine measure across all inpatient facilities. This would not be practical or appropriate. Nor am I proposing the use of harm minimisation as an approach for all patients who self-injure. This would ignore the complex nature of a phenomenon that is not fully understood. I follow the NICE (National Institute for Health and Care Excellence) guidance, which suggests there may be a small group of patients, whose self-injury is resistant to change and may benefit from a harm minimisation approach.3 The essence of my argument is that harm minimisation is ‘sometimes’ an option that provides an ethical alternative to more traditional interventions on the basis of a net reduction in harm. There are a small number of inpatients who may benefit from this approach and it cannot be used routinely. When used it recognises self-injury as a means of coping with psychological distress and the clinical aim will always be to develop alternative means of coping.

Although Pickard and Pearce appear to accept that my argument is only confined to a small group of patients, their response, at times, appears to ignore this element of my argument and imply that I advocate harm minimisation as a ‘one size fits all’ approach. This simplifies the argument to an unacceptable level allowing them to develop a counterargument based on the application of the approach to patients who pose a significant threat to themselves or others. They are then able to undermine my argument with ease. Let me illustrate my point.

The authors argue that the provision of sterile cutting instruments cannot form a part of any harm minimisation approach. I remain of the view that if concern regarding sepsis is to be taken seriously then this has to be a consideration in some circumstances. However, I do accept that this is the most contentious and risky area of the approach and it is certainly one of the main reasons that it is rarely considered in inpatient facilities. The legal problems this raises, for example, are complex and would require a detailed analysis in their own right. Pickard and Pearce are right to be concerned but I would suggest they overstate their position.

The provision of sterile cutting instruments is not the only consideration and in some circumstances alternatives are used. For example, snapping elastic bands against the skin or squeezing ice cubes can be used as an alternative to self-injury. In addition, the patient does not necessarily possess the instruments they would use to self-injure, it is the ability to self-injure in a safe and controlled environment that is important. In this respect, there are precedents for dangerous items to be used under supervision in some inpatient facilities. For example, a patient with diabetes may well inject himself or herself with insulin. When seen in this way, it should be clear that I do not propose the routine distribution of sharp instruments to patients in secure units4 ii or to patients who are acutely ill. Obviously, it would be irresponsible to provide potential weapons to patients iwho are dangerous to others, or to provide the means of self-injury to someone who is acutely ill, probably lacking in capacity and potentially suicidal. I do not and have not advocated such a position.

I also accept there are risks in providing education on basic anatomy and physiology to patients who self-injure. Many types of knowledge can be used for both positive and negative purposes and there is no doubt this information can be used to promote rather than reduce harm. However, I am not sure about the alternative. Do we not tell people and just hope they will not make use of the internet to find out? That information should be made available to patients is both a moral and in some circumstances a legal right.5 The fact that it may be used inadvisably is one of the costs of respecting a person’s autonomy. Surely people have the right to understand the damage they are actually doing to themselves while setting this against the damage they may do with increased knowledge. The important issue is how the information is relayed, the assessment of the degree of suicidal intention and the therapeutic context in which the education takes place.

Like Pickard and Pearce, I have, in fact, considered the impact of harm minimisation on both patients and staff and appreciate there are risks to adopting such an approach in an inpatient facility. The problem of patients learning from each other is illustrated by the experience of one of the authors, working in a unit using harm minimisation in the 1990s. This issue of contagion is well documented as the authors note. However, I would like to explore this a little as I also have experience of mental health units in some capacity from the 1970s onwards and the units I have worked on or managed have not adopted a harm minimisation approach.

In some of these units, self-injury was almost endemic, certainly contagious and yet the approach to intervention was based on restricting self-injury and trying to stop it happening. In such environments, the impact on staff is in many ways similar to that described by Pickard and Pearce as a possible consequence of harm minimisation. In these units, patients sometimes die as a result of their own actions. In some cases, this may involve a person who self-injures and when this occurs it is often due to an escalation in risk associated with the use of more dangerous attempts to cause harm involving the use of ligatures for example. In some cases, the patient will have been subject to increased levels of observation or to recent reduction in observations following a clinical review. In these situations, the emotional impact on patients and staff is also significant and the human costs of managing people in this way can be measured in terms of high rates of staff burnout in inpatient units.6 These issues are less about the rights and wrongs of harm minimisation and more about the need to focus on the individuals’ needs, skilled clinical care, detailed risk assessment and the involvement of the person themselves in their care, and then the possibility of suicide may be reduced although it cannot be eradicated.

What I believe Pickard and Pearce fail to do is to address the fundamental problem that provided the impetus for my paper. In considering the costs and benefits of harm minimisation in a hospital setting, no real consideration is given to the costs and benefits of current means of managing people who continually self-injure or to the fact that stopping self-injury is almost impossible to achieve and often trying to do so results in an escalation of the behaviour. Many patients are subject to high levels of observation and often with only minimal if any real psychotherapeutic support.

The patient can be in a Catch 22 position whereby access to specialist treatment depends on a reduction in self-injury, but the environment within a hospital setting contributes to the continuation of their behaviour through the methods used to try and prevent it occurring and the failure to provide appropriate psychotherapeutic support. The pressure on clinical staff that have to contain such a situation can be immense and the effects on the relationship with the patient detrimental and harmful. The dangers of malignant alienationiii in increasing the risk of suicide are well documented.7 In these situations, the patient may also receive the message that no one understands or even cares. As Pickard and Pearce point out, failing to prevent harm may give the message that ‘we won’t stop you from hurting yourself because you are not worth it.’ However, by restricting basic liberties to try and stop the behaviour we may inadvertently give the message that ‘we know you really need to do this but we are going to stop you because we don’t really care what you think.’

In closing, I would like to suggest that I might have more in common with Pickard and Pearce than their analysis sometimes allows. I believe that people who self-injure should expect that the risks must be mitigated by highly trained staff with a high level of psychological skill that provides the relational security and therapeutic containment that is required to support people who self-injure. Only then can any type of intervention be a possible option. The position of many people who self-injure and are admitted to hospital is not acceptable and the nature of acute care in today’s National Health Service may not make harm minimisation a reasonable option. Unfortunately, this lack of a sustained therapeutic ethos (in many but not all units) may be a fact but that does not make it right. People who self-injure deserve something better.


View Abstract


  • i I am particularly concerned that some media coverage associated with my paper totally misrepresented my position, suggesting I was promoting the routine distribution of sterile cutting implements to patients who self-injure.

  • ii However, it is also important to acknowledge that harm minimisation approaches have been used and promoted within secure environments. The types of interventions used are obviously adapted to this environment.

  • iii Malignant alienation refers to a process of deteriorating staff and patient relionships that can precede suicide in some patients.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.