Monday, January 19, 2015

Deceased donation in Italy: a discussion of the complexities

A thoughtful article on organ donation in Italy, with help from google translate.

Nudge, la spinta gentile tirata per la giacchetta

"Policies inspired to nudge (translated into Italian as 'gentle push') and to the political philosophy of libertarian paternalism have gained increasing attention over the years, representing a tool available to the policy makers to design and implement interventions is increasingly aimed at the citizen, to simplify and, ultimately, the conditions in which the latter makes his choices. What is a great tool, among many of the box, however, is likely to turn into a mantra, sometimes losing sight of the complexity of reality, if you neglect the main teaching of the nudge same: the experimental method used to test every intervention measure and evaluate their effectiveness. The risk of giving the principle of auctoritas, a bit 'as in the "Name of the Rose" by Umberto Eco, is strong: it is just to update, from century to century, the person called upon to play the role of Aristotle.
"The first issue was first addressed in a working paper of the Nobel Prize Alvin Roth and Judd Kessler, published last August on NBER. The results are wavering confidence accorded to the theory of the nudge. The active choice, in fact, it even seems to decrease enrollment rates, not increase them! The research of the two authors is based on two experiments. The first is a natural experiment - the case where the conditions of an experimental set are created without the intervention of an investigator. From July 2011, the state of California has, in fact, introduced the active choice through the modules of the Department of Motor Vehicles. The Californian city is now asked: "Would you like to sign up to become an organ donor?", With the dual option of answer: "Yes, add my name to the donor registry" and "I do not want to sign up now." Comparing the data before and after the introduction of active choice and using the rate of registration of the other 26 states as a control, it is seen as from July 2011 recordings in California have declined. Through this change of context of choice, it is estimated to have been lost almost 3% of potential subscribers compared to not change the rules.

Obviously, one can not impute to the introduction of the single active choice the fall of recordings. To isolate the causal effect, the two authors have thus created a particular experiment. In Computer Lab for Experimental Research of Harvard University, 368 subjects had the opportunity to login to register for organ donors and to change its status as a donor. The design of the experiment - conducted in the laboratory, but with consequences on the real life of the subjects - has allowed the manipulation of two variables. The first was the way in which each subject was asked if he intended to become a donor, asking to check one available ("I want to register in the register of donors"), or one of two boxes ("I want to register in the register of donors" or "I do not want to sign up to the register of donors"), reproducing the method of active choice. The condition of the single box is meant to simulate the principle of consent. In the event that the box had not been crossed, the subject would not have been entered in the register. The second variable consisted of the information provided about organ donation. In this case, for some subjects appeared on the web page the words "It is estimated that a donor can save or enhance the lives of as many as 50 people donating organs and tissues." Others appeared the same sentence, with the addition of a list of the organs can be donated, such as corneas, heart, kidneys, liver, lungs and others. The intersection of these two variables, therefore, has enabled the creation of four treatment groups.

The subjects who showed a higher rate of registration were those included in the treatment containing the explicit consent and the list of the organs can be donated (40% chance of recording at the end of the study), doing better, respectively, of the combination of "active choice list + "," explicit consent not list + "," + active choice not list ". In general, the subjects were more likely to enroll when the request was framed in the form of explicit consent, rather than active choice between two options. The difference between the two types of treatment is also in the order of magnitude of the difference observed in the experiment natural rate registration in California.

The ability to know what were the subjects already enrolled in registers before the experiment has uncovered two encouraging results. Provide a list of the organs can be donated to those who initially were not donors has increased the probability of recording. The 34.9% of non-donors with list joined, against 22.6% of those without the list. The most obvious result, however, was another. Provide individuals the opportunity to change their status of donor increases the number of registered donors. The subjects are in fact as many as 22 times more likely to enroll in the registry rather than unsubscribe. For this reason, the title of the article quotes the motto of Winston Churchill: "Do not accept a 'No' for an answer." Remember after a little time to the citizens the opportunity to enroll in the registers may prompt them to change the decision not to enroll taken in the past.

If this first experiment seems to question from an experimental point of view the effectiveness of the previously assumed active choice, such problems are overlooked by the behavioral sciences in addressing the complexity of the issue of the donations?

First, the regulations vary from country to country, making it impossible attempts to enclose in categories precise recording systems. A study conducted in 2012 by Amanda Rosenblum and colleagues shows how each donor registry has its own peculiar characteristics of membership, including the method of recording, the minimum age for enrollment, the role of family members and the ability to change the choice. How is it possible to identify the causal contribution of each of these characteristics?

Here emerges a second problem. Labels such as those of "explicit consent", "explicit dissent", "presumed consent", "silence", "active choice" (which in English is expressed by formulas as "mandated choice", "prompted choice" and "active choice ") are often missing in representing important nuances present in the regulatory systems. In literature, this leads inevitably difficulties on the use of terms, which affect the very possibility of successfully replicate entire experimental sets. An example is provided by the working paper analyzed, as the "explicit consent" used in the study may seem like a form of "active choice". This proposal is subject to, within a laboratory, while in many states, in the life of every day, is the subject that has to take steps to register.

A third problem is related to the major role that is played by the decisions of the family of a potential donor. In many countries, in fact, family members can change the choice of the deceased donor, not consenting to organ harvesting. Conversely, may consent to the donation, without the deceased has expressed its consent to life. Alvin Roth and Judd Kessler conducted a survey with 803 subjects, asking if they would have consented to organ harvesting of a relative in the event that the latter had written, or not, to the registers via the express consent or active choice. The results show that a family member is more reluctant to agree the removal of organs if the deceased relative has expressed his desire not to be a donor through the active choice, rather than not giving his explicit consent. As another experiment, therefore, the introduction of active choice would seem to lead to worse results than actual donations to the express, influencing the opinion of family members.

Finally, a fourth problem is the meaning that individuals attach to gesture to donate their organs. We are sure that a citizen confers the same value in the donation through different systems of registration? In this case, active choice and explicit consent could lead to similar outcomes. Shai Davidai and colleagues, in a study of 2012, as noted in a context of choice where there is presumed consent, people placed on a scale of values ​​the gesture of organ donation close to passing in front of someone when you are in tail and to devote part of their time to volunteering. Conversely, in a scenario where worth the explicit consent, assign people to donate organs a value similar to that which is given to gestures like to donate half their wealth to charity or do a hunger strike in support of a cause.

The cited literature, abundant and complex, and not containing unequivocal conclusions about the effectiveness of a particular intervention, does not point to the conclusion that certain institutional immobility is the answer: designing policy interventions that seek to achieve a particular goal is not only desirable, but it is necessary for a society of citizens aware and well informed. That said, it is worth reiterating as there is a science of miracles, whatever form it takes the same miracle that, in turn, becomes fashion in the academic community, generating curiosity (legitimate) and trust (often uncontrolled). Study, analyze, publish a scientific result, no certainties. If not that, in fact, not to have.

of Carlo Canepa and Luciano Canova"

No comments: