Choosing a treatment path involves trade-offs
■ A column that answers questions on ethical issues in medical practice
The Ethics Group provides discussions on questions of ethics and professionalism in medical practice. Readers are encouraged to submit questions and comments to [email protected], or to Ethics Group, AMA, 515 N. State St., Chicago, IL 60654. Opinions in Ethics Forum reflect the views of the authors and do not constitute official policy of the AMA. Posted Sept. 5, 2005.
- WITH THIS STORY:
- » Related content
Every patient does not have access to all medically indicated treatments, sometimes for economic reasons, sometimes because the needed intervention is scarce or not available. Is there an ethical approach to making decisions about individual patient care when all patients cannot receive all needed treatment?
In making decisions at the bedside, we recognize that the core mission of medicine is to heal or at least to alleviate the patient's disease burden and that the patient's autonomy in choosing among the courses of action available for the management of his or her condition must be protected and supported.
In the formulation of policy, it is necessary to take into account a third ethical consideration -- the trade-offs/opportunity costs that the policies entail.
The trade-offs are brought starkly into focus in organ transplantation, for example, by the fact that an organ received by one patient is denied to another who is also on the wait list. Therefore, our objective is to minimize the harm that the policies may inflict on the total population that is affected by them.
The interactions among health care professional, payer, patient and society that have evolved over the past few decades have created tensions that often are cast as ethical conflicts but are, at their root, different assessments of relative costs and advantages to the participants in the medical enterprise.
Our approach here is intended to shine the light of objective data on these issues and, by providing a shared, well-understood basis for informed decision-making, to alleviate perceived conflicts.
We think that the ethical ambiguities in treatment decision-making can be minimized by such a sound, shared model. Our objective is to accurately project what each patient's disease burden would be under each treatment option; share these projections with the patient and allow him or her to weight the dimensions of well-being according to his or her own values and preferences; and then, where allocation of scarce resources is involved, make a decision that minimizes the harm that will be experienced overall, that is, among all members of the cohort who need the resource.
Much of the difficulty in decision-making in medicine results from the complexity (multiplicity and interrelatedness) of the consequences of care (outcomes), and the differing weights that patients and other health care actors have assigned to those outcomes.
A second difficult aspect of medical decision-making is the tension between the need for measures that reflect long-term outcomes and the ease and convenience of dealing with short-term outcomes. We must go, in a meaningful way, beyond the quip that the operation was a success but the patient died.
The patient's prognosis is the projected course over time of the cumulative probability of mortality, morbidity, disability, psychological distress and resource use. Our approach makes projections of the individual and joint probable outcomes on these five measures and how these projected probable outcomes would differ under each of the treatment alternatives available in the management of the patient's condition.
To develop the approach and conduct initial analyses, we chose renal transplantation, where longitudinal data on the cited dimensions of outcome are the most abundant. Even there, unfortunately, we entirely lack data on psychological distress, and the data on disability do not have the detail needed to support useful analyses. Thus, one immediate benefit of our approach is that it identifies the current gaps in the data needed to properly support decision-making, and, indeed, we have already taken measures to correct the deficiency in the data on disability.
What we must do is define the necessary data and data processing tools so that the prognoses for the alternative courses of action under consideration can be made objectively at the time treatment course decisions must be made. In other words, the data system and the analytic tools must be integral components of the patient care system. As is always the case when such detailed information must be instantly accessible, the balance of access and privacy is a concern, but this can be dealt with by technical means.
How does our proposed approach address the ethical issues cited above?
First, the data produced by our analyses provide a clear specification of what may be expected by every individual patient, given that patient's biological characteristics (that is, the underlying pathophysiology that has produced the deterioration of organ function, the severity of the condition, the patient's physiological reserve, as represented, for example, by his or her co-morbidities, and, for the transplanted patient, the viability of the donated organ and its degree of compatibility with the recipient), in terms of the probability of dying, having additional biologically harmful events (morbidity), limitations in the ability to carry out the normal biological and social functions of life (disability), dissatisfaction with the current state and fear for the future (psychological distress) and the amount of effort that will be required to sustain the patient's life and function (resource use).
The predictions are specific for each course of action under consideration. Thus, we can tell what the probability is that the patient will die within any period of time from the present, and we can tell how probable it is that the patient will have a minor or major degree of disability over the selected period of time if he or she accepts the offered organ and undergoes transplantation now, or, alternatively, chooses not to accept the currently offered organ and remains on the list longer.
The latter course may be preferred if the prospects with the current organ, given the condition of the patient, are less desirable than they would be at a later time when the patient's condition had deteriorated further so that the risks of staying on the wait list exceeded the risks associated with the transplant.
Our analyses demonstrate that this is a not a rare occurrence at all. We estimate that as many as 25% of the kidney transplant recipients in our cohort were likely to have had a lower burden of disease had they forgone the organ that they received and waited until a later time.
In the actual decision of whether to accept an offered organ, the patient's personal weighting of the individual components of the disease burden must be decisive.
One patient may be intolerant of disability and may accept a higher risk of early death if the transplant promises to reduce his functional impairment. Another may deem living longer so important that considerable disability is a price he or she will gladly pay.
An informed decision by each patient is contingent on having credible estimates of the probability of dying and the probability of incurring substantial disability under the competing courses of action available at the time the decision needs to be made.
Given the personal nature of the weights attached to the different dimensions of health, we can inform and assist the patient in choosing the course of action that will best alleviate his burden of disease, but we cannot make the decision for the patient about what constitutes optimal well-being; we must respect the patient's autonomy.
While the patient's personal weighting of disease burdens guides his or her decision to accept or reject an offered organ at a given time, we must separate the patient's decision from our decision about who should have priority in receiving an organ that has just become available. Because we must be as even-handed as possible, our objective in determining the priorities must be solely the minimization of the harm that will be experienced by the cohort as a whole.
We must identify the patient who will be harmed most by being denied the organ; that is, the patient whose burden of disease will be increased the most by being denied the organ and having to wait longer for another organ, and give that patient the highest priority, and so on down the list of patients awaiting transplantation.
Our methods equip us to do that also, and without having to explicitly weight the multiple components of the burden of disease. Our estimates of the aggregate burden of disease simply reflect patient preferences expressed by the cohort as a whole (the revealed preferences, in the language of economists).
Thus, our approach goes beyond what is often considered the dominant ethical problem in health, the need to assure access. We ask the more fundamental question -- is the intervention under consideration of help to the patient? The assumption that any available medical service is worthwhile and should be offered when desired has been demonstrated often, including by us, to be incorrect. The assumption, in fact, represents a moral failing of medical practice. Providing a medical intervention to a patient that will harm that patient is not, let us admit, a good thing to do. In addition, ignoring the trade-offs and opportunity costs associated with the intervention, for example the levels of morbidity and disability that may have to be accepted to gain a prolongation of life, is also a profound disservice to both the particular patient and to the other members of the community.
R. Clifton Bailey, PhD, McLean, Va.
Acknowledgment: The concepts and principles enunciated herein evolved in the course of the work my collaborators, Henry Krakauer, MD, PhD, and Monica J-Y Lin, PhD, and I carried out over many years and in various settings. This work is summarized in the paper "Assessing and Using the Multiple Correlated Components of the Burden of Disease in Decision-Making in Health Care," in the May "Health Care Management Science."
The Ethics Group provides discussions on questions of ethics and professionalism in medical practice. Readers are encouraged to submit questions and comments to [email protected], or to Ethics Group, AMA, 515 N. State St., Chicago, IL 60654. Opinions in Ethics Forum reflect the views of the authors and do not constitute official policy of the AMA.