Intended for healthcare professionals

Education And Debate

Rationing in general practice: The Asbury draft policy on ethical use of resources

BMJ 1996; 312 doi: (Published 15 June 1996) Cite this as: BMJ 1996;312:1528
  1. Roger Crisp, fellow in philosophya,
  2. Tony Hope, lecturer in practice skillsb,
  3. David Ebbs, general practitionerc
  1. a St Anne's College, University of Oxford, Oxford
  2. b Department of Public Health and Primary Care, University of Oxford
  3. c Didcot Health Centre Practice, Didcot, Berkshire
  1. Correspondence to:. Dr T Hope, University of Oxford Medical School, John Radcliffe Hospital, Oxford OX3 9DU.
  • Accepted 12 April 1996

Many doctors find themselves torn between two contradictory principles: to do the best for the individual patient and to be responsible for an overall budget that is insufficient for the best care for each individual patient. Little guidance is available for doctors on how to resolve this conflict. Crisp et al present a draft document that one fundholding general practice has developed to clarify the ethical basis for rationing decisions. We invited three interested professionals to comment on the draft.

The general practice partners invited two medical ethicists (RC and TH) to meet them to develop the document. The partners met RC and TH for one and a half hours on eight occasions over one year and met without them on eight further occasions. The entire general practice also had an all day session to discuss in detail an advanced version of the draft. The developmental process was of great value to the partnership and has led to appreciable change in individuals' views. The draft policy presented here is intended to start the ball rolling, so that proper guidelines will be developed at whatever level in the NHS is most appropriate. Comments and feedback are welcomed.


  • The primary aim of this policy document is to provide a principled basis for the distribution of financial and medical resources within the practice.

  • The practice is assumed to have a responsibility to provide health care within budgetary constraints. It is also recognised that differences of opinion between partners about the use of resources are inevitable. A secondary aim of the document, therefore, is to provide machinery for making decisions in cases of disagreement.

  • We believe it is important to consult widely before producing the final document. This draft, therefore, will be discussed with other members of the practice, with patient forums, and with professionals and the public.

  • The document is intended to be available to any one who requests it.

  • This document is the outcome of many meetings attended by the partners. Advice has been sought when appropriate from others outside the practice.

  • The document is concerned with how the partnership should come to decisions about the distribution of the resources under its control. The resources available to the partnership are fewer than the partnership considers ideal. In producing these guidelines, the partnership does not wish to imply that it endorses as ideal the level of funding available to it.

Ethical background to allocation of scarce resources

In preparing this document, we considered in particular three general theories pertinent to the allocation of medical resources. Each theory focuses on different values. We believe that decisions on the use of resources should not be based on only one value. Several values are at stake, and the judgment of the partners will be needed to balance these values in particular cases.


Quality adjusted life years (QALYs)—The theory of QALYs1 was developed specifically to address the issue of how limited resources for health care should be distributed. It focuses on maximising the welfare of patients. Patient welfare, according to the QALY theory, is the product of length of life and the quality (to the particular person) of that life. Various empirical means have been suggested for measuring the “quality adjustment.”

The fact that a certain treatment will produce greater patient welfare than another is a reason—though not necessarily an overriding reason—for that treatment to be chosen. It will thus be important that the partners have as much information as possible about the impact of various treatments on patient welfare.

Needs theory—Needs theory2 3 is based on the view that some patients have a special claim on resources that rests not on the mere maximising of overall welfare but on their greater need for treatment. The most thoroughly worked out version of needs theory is that of John Rawls, who emphasises the value of fairness.

We believe that medical practice should not aim solely to maximise overall patient welfare, because it matters how this wellbeing is distributed among patients. Consider the following hypothetical case of hernia treatments versus kidney treatment. A doctor could treat either one hundred otherwise healthy people for hernias or one very sick person for severe and debilitating kidney problems. On the assumption that each hernia treatment provides one unit of benefit and the kidney treatment provides fifty units, the total number of units of benefit for the kidney treatment is double that of the hernia treatments.

A principle according to which patient welfare should be maximised suggests that there is no reason to treat the person who is worse off-namely, the patient with kidney disease—because he or she is worse off. But another value, that of fairness, requires that some consideration be given to patients who are worse off, perhaps in terms of meeting basic needs, independently of how much patient welfare will be produced by treating them.

Lottery theory—The lottery theory4 arises from the view that in many health care situations there is no good reason, when a choice exists, for treating person A rather then person B, or vice versa. In such situations the value of procedural fairness suggests giving both an equal chance of treatment.


So far the values of patient welfare and substantive and procedural fairness have been mentioned. A fourth value is the autonomy of patients. We believe that allowing patients as much say as is practically possible in their treatment is a good thing in itself.

Imagine that two treatments are available for a certain condition, each of which has different side effects. One treatment is slightly more expensive than the other. Allowing patients to choose their treatment will respect their autonomy. There are limits to autonomy, however, because of the partners' responsibility to provide health care within a specific budget. Thus if two treatments differed greatly in price patient choice may not be possible.


We believe that a plurality of values does not rule out rational decision making—in fact, such decision making requires that all these values be taken into account. In some cases—for example, when a great gain can be achieved in patient welfare by treating those patients who are not in fact the worst off—it may be rational to decide to produce this gain. But in some cases it may be worth sacrificing overall gain to offer some priority to those worse off.


In the absence of a single overriding theory or an ethical principle that provides for all circumstances, we believe that it is critically important to establish a methodical process for making decisions.

Because the decisions depend on the particular facts of each case, as much clinical and financial information as possible must be available. The decisions will be made at the monthly practice meetings. The values underlying these decisions are those discussed above, and this section of the document provides a constitutional framework for discussion.

We believe that it is important to be open about rationing policy. This document serves as a statement of policy. The partners' monthly meetings will be an important forum in which this policy statement will be interpreted in specific instances. If there is any major conflict between partners in interpreting this policy statement or if a change in policy is proposed then a specific policy meeting will be called.


In the discussions about allocation of resources any information about a patient and his or her situation is considered admissible. Both positive and negative bias from the patient's advocate—for example, the partner most involved in that patient's care—may distort the presentation of the case. We believe, however, that the patient is protected from such bias by the presence of the other members of the group. It is for this reason that both policy statements and decisions about choices between individual patients (necessary because of rationing) must be made by the group and not by individuals. We considered whether we should regard some information about patients—such as whether they have a learning disability—as inadmissible to the discussion. We concluded that if some information were inadmissible this may wrongly affect a partner's judgment without it being clear that the partner's judgment was affected. We therefore decided to allow all patient information to be discussed, but the partners must ensure that the decisions made are informed by this policy document.


Because the partners carry legal responsibility for the use of the resources, they have to make the final decisions about what should become policy. The partners should, however, consult as widely as appropriate in drawing up this policy statement, in interpreting it, and in developing it. Some other members of the practice would probably also be present at most policy meetings. We foresee a time when the composition of the policy group—that is, the group that makes the final decisions about rationing policy—will be multiprofessional and representative of a wide range of different views. We are looking at a model for this process of wider consultation, so that the views of the team are incorporated in the policy.


In addition to preparing and developing this policy document, the partnership undertakes to provide an annual report. This report will summarise the issues raised—both at the partners' monthly meetings and policy meetings—and the ensuing discussion when these concern issues of resource allocation. The annual report will be made public along with the most recent policy statement. It is important, therefore, that both the policy statement and the annual report do not breach any individual patient's confidentiality.

Values held by the partnership: general policy

The central ethical principle that guides the practice of medicine in this partnership is that of “the best interests of each patient.” The partnership will try to provide whatever medical care is in the best interests of each individual patient.

This policy statement has been drawn up, however, because the limitation on funding means that the partnership may not always be able to pursue the best interests of every patient in every circumstance. As our practice is fundholding, a considerable portion of the total medical budget is under our direct control. We therefore wish to clarify for ourselves, our employees, and our patients what principles and working practices should guide us when budgetary constraints prevent us from being able to pursue the best interests of every patient. However, for most patients, most of the time, we envisage that we will be able to provide the care that is in each patient's best interests.

When the best interests of every patient cannot be met, a decision about what should be done will normally be made by the partners after discussion at a specifically convened meeting (the partners' monthly meeting). This meeting will normally be cancelled only when a patient's situation is one for which an agreed policy already exists.

All the values identified by the various ethical theories will be considered before a decision is reached. Partners have different views on the way in which these values are balanced. The mechanism for making a decision will be through consensus, and if necessary through voting, at the partnership meeting. We believe that this mechanism is preferable to individual partners making decisions for two reasons: it ensures clarification and identification of various points of view, and it ensures that individual partners' responses to specific patients do not, illegitimately, affect the choices made.

Values held by the partnership: specific issues AGE OF PATIENTS

We do not wish to deny treatment on the basis of age—that is, the partnership rejects any policy that states that no one over a particular age should ipso facto not be allowed a specific treatment. Hospitals, however, which are not controlled by the partnership, may operate such a policy for some treatments, and the partnership may be unable to find suitable alternative treatment. But the partnership will do what it can to ensure treatment and will not itself deny treatment on the basis of age.

Situations may occur rarely in which the partnership has to choose which of two patients should have priority in receiving some beneficial and expensive treatment. In such cases a decision about how to proceed will be made at the partnership meeting. It will be legitimate to consider any factor as at least potentially relevant to making a decision. This does not imply that any factor that can be considered will be thought relevant in a specific case. For example, the patients' ages may be considered and found relevant. This is because the partnership believes that two ethical values may be relevant: (a) how long each patient is likely to live to enjoy the benefit of the treatment (a value endorsed by the QALY perspective), and (b) “the fair innings argument,” which highlights a value in justice which the partners consider to have some weight. The fair innings argument suggests that to “treat the older person, letting the younger person die, would thus be inherently inequitable in terms of years of life lived: the younger person would get no more years than the relatively few he has already had, whereas the older person…will get several years more.”5


With regard to patients with chronic problems affecting welfare—for example, learning disabilities or chronic physical ill health—the partnership does not endorse the values embedded in the QALY theory. A person's learning disability or rheumatoid arthritis, for example, is not a reason for either lowering or increasing his or her priority over people without those problems, although what is in the best interests of the patient may be affected by his or her chronic problems.


The partnership rejects any general policy that denies specific treatment to a patient on the grounds that he or she has brought the condition on himself or herself. The partnership believes that if the patient would benefit from medical treatment then it should be available, regardless of the cause of the condition. However, if the partnership has to choose which of two patients should have priority in receiving some beneficial and expensive treatment then the issue of a patient's responsibility for having induced the problem might be considered relevant.


In drawing up this policy document, the partnership considered the following, fictional situation to help it to clarify its views. Two patients need the same treatment for the same life threatening condition. Patient A is a brilliant surgeon who saves hundreds of lives a year; she is a single parent with three young children. Patient B is unemployed and has no dependants. Resources are available for treating only one patient.

Should the issue of dependants ever affect in any way the priority of patients for access to scarce resources? In the fictional case above, do the facts that patient A through her work has a beneficial effect on many other people and that she has dependent children provide a reason for giving her a higher priority?

The partners believe that patient A's work is not relevant to decisions on priority; they do not think that they should be making any judgments about the value of patients to society in deciding issues of resource allocation.

However, the partners believe that if, as in this case, a patient has dependent children then this could be a factor in increasing the priority for scarce resources if those resources will affect the patient's ability to care for the dependent children.

Paying for treatment

  • All patients have the right to seek private treatment either by referral from the partnership or independently of it.

  • Patients have a right to NHS treatment as laid down in various statutes.

  • Some treatments—for example, paracetamol syrup for children—are advised to a large number of patients. Under the terms of service with the NHS, patients have the right to an NHS prescription for these treatments. Prescriptions for these treatments, because they are common, are a high expense to the practice's drug budget; the money could be spent on other treatment. The cost of a prescription for any one patient, however, is low. The partners believe that, despite a patient's right to obtain a prescription, it is right to inform most patients (or their parents) that they could buy such drugs themselves, although if the drug would not be purchased then a prescription, when clinically indicated, should be given.

Factors in deciding priority for allocation of scarce resources

May be relevant


Dependency on the patient of people who are close relatives

The patient's responsibility for causing his or her condition

Not relevant

Value of a patient to society

Value of a patient's life to that patient



Dependency on the patient of people who are not close relatives

Relation to district health authority

The practice will normally expect to follow any district health authority guidelines for funding specific procedures. The partners accept that, at this stage, most authority decisions have been made on the basis of good reasons after appropriate consideration of the available evidence. The partners will, however, ask the district health authority for the reasons behind its policy. A distinction exists between purely clinical and ethical reasons for refusing to fund treatments. The partners believe that it could be appropriate to act contrary to the authority's guidelines if they found the reasons behind the policy inadequate. They envisage that this is more likely to be the case with regard to ethical guidelines than with clinical guidelines. If the reason behind an authority guideline is found to be inadequate the practice policy would be determined after discussion at a partnership meeting.

The partners believe that it would be undesirable to create a local two tier service, and this should be avoided whenever possible. To help to prevent a two tier service the partners should be aware of monetary constraints forced on local non-fundholding practices by the district health authority. The partners should adopt policies at variance with those of nonfundholding practices only after careful consideration.

At present no funding will be available from the general practice budget for alternative therapies such as homoeopathy and osteopathy.


The partners do not believe that they should directly profit financially from money intended for patient care.


As with the cost-rent scheme, in which we have an interest, we believe that appropriate resources should be allocated in a regional strategy for programmes in which appreciable expansion in establishment is needed to take account of changing demography and population growth. It is the responsibility of the family health services authority and district health authority to take appropriate action to provide adequate premises for accommodating general medical services. Money intended for patient care should not be used for this purpose.

Future budgets may contain development money that could be used for building costs. Such necessary work should be modest and functional. If the fundholding budget is underspent, and all agreed quality standards relating to clinical care have been achieved, the practice should discuss how best to spend the money, focusing on improvement in patient care.


Before the practice became fundholding all equipment costs were the legitimate expense of the practice, but now planned savings can be used to buy items for patient care. We believe that the practice should be well equipped and that when a piece of equipment will directly benefit patient care in offering new services, improving existing services, or in saving expenditure elsewhere—for example, by buying an audiometer or tympanometer—it is justifiable to spend fundholding money.


The QALY theory is likely to result in few resources being put into terminal care. This is because, however much such care might enhance the quality of the patient's life, there will not be much quantity of life. Terminal care is therefore likely to be expensive when measured in terms of the cost per QALY.

We believe that good care at the end of life is an important aspect of medical care and that the quality of life then and the manner in which a person dies have an importance that is not captured by the idea of welfare in terms of life years. This is true because of the importance of how a person's life ends both to that person and to the person's close relatives. Good terminal care therefore might be considered to be a need.

Informing patients

We believe that patients should be told which policies—relevant to the allocation of resources—the partnership is following. Such policies include not only those specified in this document but also clinical protocols that have been developed at least in part as a response to budgetary constraints. The partnership might develop, for example, a protocol for treating a particular condition that states that treatment A is normally to be preferred in the first instance to treatment B. This protocol might be developed because, although treatment B is slightly better—for example, it may have fewer unwanted side effects—it is much more expensive. Any such protocol developed by the partners will be made public and will be available to patients in the same way as this policy statement.

Standards of clinical care: partners' responsibilities, to whom and for what

Our responsibilities are to individual patients—to provide high quality care. High quality care is care that is appropriate to the patient's needs and expectations based on logical rationale and validated by accepted scientific evidence. It should not be determined by availability or cost when a choice of treatment is being made.

We have an obligation to provide the highest quality of care at the best value for money, which means that we should follow agreed guidelines, procedures, and protocols of care and be aware of the costs of various equivalent options.

We have a responsibility to the patient to choose the most effective treatment that also represents the best value. We also have responsibility to assess the health needs of our local populations and to represent these to those authorities that deal with locality funding.

To fulfil these responsibilities we should:

  • Collect and consider all existing agreed district protocols and guidelines and follow those that have been scientifically validated

  • Review contract negotiations with providers to avoid waste of resources by unnecessary repetition of investigations, outpatient appointments, and follow ups (reports and regular review should be made of these and appropriate purchaser-provider locality meetings set up)

  • Validate in house protocols by research and regularly audit agreement of and adherence to these protocols

  • Make representation to supplement funds when the budget or resources are inadequate to fund high quality care based on acceptable criteria; when funds are not forthcoming, we should share our concerns with our patients, the public, and any influential public figures

  • Agree policy decisions relating to standards of care with all relevant professionals and make these public for scrutiny.

DE writes on behalf of the general practice partners of Didcot Health Centre Practice.


  • Funding No special funding.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.