Seclusion & Restraint I
Michael A. Gillette, Ph.D.

Recently, I was invited to provide an educational session for a hospital regarding ethical issues in the use of seclusion and restraint. As I began to prepare for my presentation, I attempted to categorize the types of cases involving seclusion and restraint according to their ethical characteristics. This month I would like to introduce some of the thoughts that I had on this topic, and will in future months continue the discussion with example cases in each of the specific moral categories.

To begin, I claim that cases involving seclusion and restraint fall into four distinct moral paradigms. The four paradigms include cases in which the operative moral concern surrounds direct harm to self, direct harm to others, indirect harm to self and indirect harm to others. Examples may help to elucidate these distinctions.

Direct Harm to Self: A good example of this sort of situation involves a patient in a mental health institute who persistently engages in self-injurious behavior such as repetitive head-banging. Restraints might be used in this situation to stop the patient from harming himself, when such harm seems to be the direct result of the patient's intended actions.

Direct Harm to Others: This type of problem is best exemplified by a client who is violent toward others. Imagine a patient who has been known to attack others. The attacks may range in aggressiveness, but include hitting, kicking and biting. The use of seclusion and restraint in these sorts of cases is based on a concern for the welfare of individuals who may be harmed by the patient's aggressive behavior.

Indirect Harm to Self: This case involves a client who is harmful to herself, but in a less active manner. A client at a geriatric center may continually fall out of bed for instance. Staff are concerned that the elderly patient will suffer fractures and other injuries as a result of falling out of bed, and they desire to place the client in restraints to avoid a fall. The restraints in this situation may be mild, such as the use of bed rails, but if they restrict the client's freedom of action then they are likely to qualify as restraints and fall under our present topic of discussion.

Indirect Harm to Others: This class of cases is the most difficult to deal with conceptually. Like the harm to others situation explained above, the primary motivator for the use of restraints or seclusion is that other clients will be harmed if a particular client is not restrained. Unlike the previous case, however, the harm involved here is much less direct and much more difficult to attribute to a particular offending individual.

Imagine for instance a client with extreme Pica who must be monitored around the clock in order to stop him from ingesting dangerous materials. Further, imagine that staffing is limited, and that in order to provide the appropriate support for this client, care must be withdrawn, or at least reduced, from other clients. Is it fair that a larger number of other clients should suffer a reduced quality of care because one client exhausts available resources? Clearly others are harmed here, but the harm seems indirect.

Now that the four moral paradigms of seclusion and restraint have been mapped out, we must consider the two distinct approaches that are applied to the four paradigms. These two approaches are classified as issues of distributive justice and issues of paternalism.

Distributive justice is the area of ethical inquiry that revolves around inter-personal allocations of goods. In society there are many social goods that are allocated by a variety of means. Social goods include all things of value ranging from obvious goods like food, clothing, shelter and money to less obvious goods like prestige, security, authority, freedom and self-esteem.

In considering distributive justice issues, the moral values at stake surround the methods of allocation and discrepancies in the size of personal shares of the desired social goods. We cannot expect that all people will, or should, have perfectly equal shares of all of the social goods. The point of distributive justice is to determine what a fair distribution of goods might be, even when that distribution is not even.

The best example of distributive justice comes from the classroom. Grades are goods of which we all want a healthy share. Some of us will get A's, and some will get F's. Most of us will probably get B's and C's. The question from a distributive justice point of view is "What is the fairest way to allocate grades in the classroom?" Should effort be rewarded, or should good results generate the best grades?

Similar questions can be asked regarding the above outlined paradigms involving harm to others. When one client threatens harm to others, either directly or indirectly, it is necessary to determine who must pay the price for that harm. The client could be protected from restraint, but only at a cost of harm to others. Alternatively, the others could be protected at a cost to the original individual. In determining how to proceed, competing claims to protection must be balanced, and thus the question is one of distribution.

In cases involving harm to self, a different approach is warranted. Unlike inter-personal disputes which are understood in distributional terms, intra-personal difficulties must be understood in terms of paternalism.

A paternalistic activity is one that is executed in a way that restricts the freedom of a particular person, but does so in order to benefit the subject of the restriction. When I intervene to stop my child from naively jumping off of the roof while thinking that a bed sheet will act as a parachute, I do so for my child's own good and no one else's. This second approach to seclusion and restraint, the paternalistic approach, is involved in cases where I act to stop a client from harming herself.

There are cases in which it seems clear that both distributional and paternalistic interventions involving seclusion or restraint are justified. There are also cases where the justification is not so easy to establish. In future months, we will examine the gray areas.