To the Editor:
In a recent editorial, Dr. Klintmalm argues that when evaluating potential living liver donors our first responsibility is to do no harm (1). At first glance this time honored advice seems sound, but careful reflection reveals that it is too simplistic to serve as a useful guiding principle. Almost every accepted medical treatment has the potential to harm. Based on this reality, Dr. Moore concluded 20 years ago (2) that the admonition to ‘“first do no harm”… seems to have been mere window dressing for the necessary hurt that has accompanied much of medical practice over the centuries’. Similarly, Beauchamp and Childress assert that the principle of non-maleficence is not absolute and can sometimes be justifiably overridden when the benefit of a maiming procedure outweighs the harm inflicted, as when a gangrenous limb is amputated to save the patient's life (3). One might counter that while other surgeries are performed for the benefit of the patient operated upon, donor partial hepatectomy is performed to benefit someone else. This is true but it is not the whole story. Living organ donation often benefits donors as well as recipients, and if the benefit is large enough, it can justify this risky procedure. Thus, ‘first do no harm’ cannot be the guiding principle for living liver donation; if it were, we would have to abandon this life-saving practice because liver donation is not entirely safe and it always maims the donors.
Despite the theme of his editorial, Dr. Klintmalm agrees that ‘Primum non nocere’ may sometimes be violated, as in the case of kidney donation. But because of the much greater risk involved in adult-to-adult liver donation he concludes that ‘violating the principle of “do no harm” is less justifiable’ here (1). We agree that living liver donation is harder to justify than is living kidney donation. But, as we now discuss, our reasoning and its implications for practice are different.
If we continue to allow people to serve as living organ donors, and if we agree that there are limits regarding the degree of acceptable risk, then we must have a mechanism for deciding how much risk is too much. Dr. Klintmalm suggests (1) that this determination requires balancing ‘the risk to the healthy living donor versus the recipient's risk of dying from disease’. We disagree. Dr. Levinsky pointed out many years ago that physicians should act solely as advocates for their patients (4). Physicians cannot accomplish this goal if, when deciding whether to recommend a procedure for one patient, they are asked to balance the risks for that person against the benefits for another (5). Such an approach would pose a clear conflict of interest, the recognition of which has led to the sensible recommendation that potential donors and recipients be evaluated by separate physicians.
How then should the donor team decide if a volunteer is acceptable? We believe that just as is true in other medical situations, for a physician to recommend her patient as an organ donor she must conclude that there will be benefits for her patient, i.e. the potential donor, that are sufficient to offset the risks (5). This approach avoids conflicts of interest that arise under the donor-recipient balancing plan and directs physician loyalty to where it belongs—solely with her patient. It also explains why we believe it is more difficult to justify adult-to-adult liver versus kidney donation: liver donation is riskier than kidney donation and the greater the risk to the donor the greater the benefit needed to offset that risk.
How can people benefit from donating liver lobes? Donor benefits are primarily psychological and often result from the resurrection of a cherished loved one. Not surprisingly, these gains can be very large. For example, consider the enormous benefit that may be realized by a woman who is able to save her spouse's life by donating a part of her liver to him.
The suggestion that donor benefit is the key to justified living liver donation does not mean that the likelihood of recipient benefit is irrelevant. The probability of recipient benefit is very important, but not as a simple balancer of donor risk; instead it provides critical information that affects the likelihood that the donor will benefit. For example, donating a liver lobe to a recipient with such a low MELD score that her survival would be adversely affected by transplantation does not make sense for the donor (nor the recipient) because there would be little possibility for donor benefit.
Given that donor benefit is psychological rather than medical, it follows that whether and to what degree a donor will benefit depends largely on his or her values (5). Consider two potential liver donors with different life goals but similar levels of medical risk who are contemplating donating to the same recipient: one may benefit greatly from donating, e.g. a woman trying to save her husband's life, while the other may not benefit at all, e.g. a brother who steps forward solely out of guilt. If this is so, then it is not possible, as Dr. Klintmalm suggests (1), to ‘develop a consensus on when… the risk to the donor is justified [and when it] is not justified’. As previously stated, we believe that this determination should be based on an assessment of the ratio of expected donor benefit to risk, and this in turn requires an understanding of each donor's values. Because values are not universally shared, a consensus-generated ‘one size fits all’ approach does not make sense, except for agreeing to exclude volunteers for potential recipients with very low MELD scores who will not benefit from transplantation. For some highly motivated volunteers accepting the real risk of adult-to-adult living liver donation may be reasonable, while for less committed potential donors it may not.