Maurie Markman, MD
It is perhaps a little unusual that an oncology commentary would begin with a highly provocative discussion about the future of driverless cars1
, but there are similarities in the sharp corners of the debate over this new technology and emerging trends in cancer care.
A particularly thorny issue concerning driver less vehicles relates to design requirements for the automatic safety features. If a crash is about to occur (perhaps caused by reckless driving of a human who has had far too much to drink), an algorithm will “make decisions” as to the optimal course of action. One of the most important decisions in the development of the system’s logic would be the priority given to specific impacted groups. Should the goal be to program the onboard computer to take evasive action designed to minimize the risk of serious injury or death independent of the specific danger to the passengers in the driverless car itself? Or should the highest priority be to minimize the risk to that car’s passengers?
In a research project designed to address the question of how drivers would respond if asked what they believed to be the top priority, most respondents agreed that the ethically correct approach would be to minimize risk for all potentially involved in an unavoidable accident or crash, independent of the specific risk to the individuals in the car in which they were a passenger. However, when asked if they would purchase a car with this particular builtin feature, the answer was, "No." Although they fully recognized that a particular action may be best for the entire population, that action may not be optimal for the individual and therefore would not be acceptable.
Many Oncology Examples
Conflict between the individual and the group has a constant presence in oncology. One example is the ever-present, intense debate about the staggering increases in the costs of medical insurance—which spreads the financial risk across a large population—and the natural desire of individuals with cancer to have their third party payer agree to pay for all interventions and services that may favorably impact their outcome regardless of the cost to the insurance plan.
Another example of the tension between the individual and the group is vaccination for the sexually transmitted human papillomavirus, which has been unequivocally confirmed as a highly effective strategy to prevent cervical cancer and a large percentage of oropharyngeal malignancies. Society will benefit greatly by having a lower burden of these cancers, but only if individuals agree to be vaccinated.
A recently reported comprehensive review of the growing problem of endometrial cancer in the United States is another poignant illustration of this tension.2
It is estimated that the incidence of this malignancy will increase by 55% to 42.13 cases per 100,000 women over a 20-year period (2010 to 2030). The major culprit in this development appears to be the growing obesity epidemic, which is believed to be responsible for approximately 60% of cases. Not only do women who are overweight have an increased risk of developing endometrial cancer, but their prognosis is substantially worse in the presence of severe obesity.
Does an individual woman who is obese have any personal responsibility to attempt to lose weight to reduce her risk of endometrial cancer, recognizing societal obligations to pay for the costs of care for this strongly obesity-related malignancy?
This question is quite similar to that which could be addressed to a heavy cigarette smoker, where the substantial healthcare-related costs associated with the multiple medical conditions related to this habit, such as cardiac and respiratory dysfunction and cancer, may have to be borne by society as a group. In this situation, however, it might be argued that cigarette taxes and potentially higher charges associated with buying medical insurance will help compensate for societal obligations.
Another level of tension regarding an individual versus the group relates to what might be viewed as following potentially acceptable population based outcomes, even if a small number of individuals may be harmed.
An example of this effect can be seen by examining the results of the risk and impact of severe neutropenia in patients undergoing chemotherapy for breast cancer.3
The analysis was conducted to determine the benefits of employing prophylactic granulocyte colony-stimulating factors (GCSFs) with several commonly used chemotherapy regimens in this setting. The major clinical outcomes examined were the development of neutropenic fever or hospitalization related to infection.