Discussion
In this national study of the immediate impact of a social media–based intervention on organ donor registration rates, we have shown that online registration rates in the United States increased dramatically after the Facebook organ donor initiative. The effect of this initiative varied by state, with the impact on registration rates ranging from 6.9× to 108.9× on the first day.
The organ shortage crisis in the United States has both medical and social causes. The Institute of Medicine has estimated that an additional 5000–10 000 deceased patients per year would be appropriate for donation but are lost due to failure to obtain consent. Given that each donor can provide up to eight solid organs and additional tissue for transplantation, improved donation rates would have a major impact on the organ shortage. However, lack of consent for deceased donation is prevalent and multifactorial. While the American public, when polled, is overwhelmingly (95%) supportive of donation and organ transplantation, only 30–50% in any given state are ultimately registered by the DMV[15]. This disconnect suggests ongoing barriers to donor registration or at least inefficiencies in our current system for obtaining registrants, which require the impact of any educational effort to be remembered and acted upon in the future, perhaps months to years later during the next trip to the DMV. Our findings highlight the potential for non-DMV based interventions, such as social network based online registrations or other novel approaches, which allow for immediate action when an educational message or a friend's input is received. Additionally, DMV and social media efforts need not be mutually exclusive: if a new registrant at the DMV is then also encouraged to announce that decision on Facebook, the power of the single registration may be multiplied manyfold by the influence of the message as it spreads in the social network.
Changing opinions and behaviors on a sensitive and difficult topic may prove more amenable to discussion in a different environment, that is "among friends," compared with the DMV and other previously targeted areas. A social media approach has been used successfully in other areas ranging from encouraging participation in disaster relief efforts to discouraging bullying among youths, although metrics to accurately quantify the impact of these interventions have been difficult to establish. Not only does organ donor declaration seem uniquely suited to social media dissemination, but it is also an area where the outcomes (namely online donor registration) are immediately measurable.
The Facebook donor initiative also provides an easy portal and intermittent reminder for those who might consider donor registration in the future. In other words, every time a network member registers, an announcement will be sent out to hundreds or thousands of other potential registrants. This "chronic virality" may give the Facebook organ donor initiative a chance of sustained impact that other previous media campaigns have lacked. While the impact of the introduction of the Facebook organ donor initiative decreased over the 2 weeks examined, it is important to note that the effect measured in this study reflects the impact of the intervention on only the 30% of Facebook members currently using the "Timeline" profile. As Facebook rolls out this profile to the rest of its US members, more will have access to the Organ Donor status option, and more are likely to participate. While the organ donor initiative is currently limited to English and to the United States, it will be added to other countries' users and other languages sequentially, and made available on mobile phone platforms as well. As time progresses, the majority of Facebook's nearly one billion members worldwide will thus have immediate access to convenient organ donor registration, educational materials about the process and the ability to communicate their decision to a receptive, friendly audience. Of note, an online effort like the Facebook organ donor initiative does not require the preexistence of a nationwide organ donor registry; in fact, its use may represent the only documented source of an individual's wishes in an area of the world which is still without a registry. However, the initiative's demonstrable efficacy is enhanced when it can be linked to a preexistent online registry, as our study illustrates. There remains great worldwide variability in deceased organ donation registries and Facebook and Donate Life America may be able to participate in the online construction of such resources in countries where they do not exist or are being developed and can be encouraged by the rollout of the donor initiative, that is, Mexico.
There are several limitations of this analysis that merit discussion. First, this report describes only the immediate impact of this social network based intervention; while the immediate impact was evident, its durability remains unknown and was obviously diminishing. Ongoing efforts will still be required to direct people toward organ donation and now additionally toward announcing their decision more directly to their friends. Second, it will take years to see an increase in actual deceased organ donations resulting from this effort, so we are limited to organ donor registration as a surrogate endpoint. Finally, the extent to which the dramatic increase in donor registration was due to social network communication versus the conventional media announcements that accompanied it cannot be separated through an observational approach. However, since the increase in new donor registrations was largely limited to online registrations, it is unlikely that the increase was caused by events which would affect people's overall willingness to donate (e.g. public service announcements aimed at the general population). Also, major national press coverage of other transplantation stories has not translated into this level of increase in donor registry activity. The minimal increase in DMV new donor registrations during May 1 and 2 may reflect what little effect was derived from conventional media coverage of this unique Facebook initiative.
Some have questioned whether this online initiative may raise privacy issues. As information about the decision to donate is voluntarily shared on Facebook either publicly or with only family or friends as specified by the user, privacy issues are not relevant simply because the medium is online. While many issues regarding one's medical information are trending toward restriction, Facebook remains a communication utility that here simply serves to facilitate the voluntary sharing of information which the user has dictated. Healthcare providers and many others will continue to encourage the population to share with their loved ones their wishes regarding donation after their death and updating one's profile on Facebook may serve as a reminder or opportunity to have that discussion verbally with family members as well.
In the 13 days following the Facebook Organ Donor Initiative's rollout, approximately 100 000 Facebook members selected this option, and approximately 33 000 Facebook members who were not previously designated organ donors completed this designation online. This initial "early adopter" cohort represents only <0.1% of Facebook users in the United States and yet the immediate impact on donation registration rates in the United States was greater than that seen with prior media campaigns. To better understand why the Facebook organ donation effort mobilized such a small percentage of potential online participants, it may be helpful to examine a recently published analogous effort attempting to utilize social media to increase voter registration. Fowler et al. constructed a randomized controlled trial of political mobilization messages delivered on election day to 61 million Facebook users and compared voting behavior changes to a control group that received informational messages devoid of "social message". These researchers were able to demonstrate a very small but significant impact of this intervention: users who received the social message were 0.39% more likely to vote than users who received the informational message or no message at all. This effect, like that seen with our organ donation initiative, was indeed small, far <1%. However, in the political mobilization study, because of the large scope of the network utilized (61 million), the researchers were able to demonstrate that the Facebook message resulted in an additional 340 000 votes, certainly enough to alter election results. Similarly with the organ donation initiative, expecting more dramatic initial results may be unrealistic, but multiplying a small impact by a large network may be enough to eventually save lives.
The next challenge for efforts like the organ donor initiative will be utilization of social media applications like Facebook, Twitter, YouTube or Instagram more effectively and more durably. D'Alessandro reported that among college students, social-based communications had the greatest impact for donor registration and described a social media campaign utilizing Facebook and YouTube video submissions that resulted in 9000 documented donor registrations. Integrating the use of novel social media with existing mass media donation campaigns may prove most productive. Stefanone demonstrated in New York via a comparative approach with three different online media formats that while traditional online advertising offered the greatest message exposure, when combined with student seeders promoting donation using social network sites and challenge campaigns the result was an increase in request for organ donor cards and registrations. The donation initiative described here ultimately had a diminishing impact over the 2 weeks examined, a fate longer but similar to other online media campaigns. In order to re-energize this effort and plan others, modifications similar to those employed by Stefanone could be utilized. Experts in the field of behavioral economics have examined organ donation barriers and suggest that a mandated choice approach or modest incentives to register would benefit our Facebook organ donation initiative relaunch as well.
Sociall media has begun to play a role in public health campaigns in areas as diverse as suicide prevention, childhood obesity and attitude toward vaccination, and the CDC has now defined a set of social media tools, guidelines and best practices. The spread of complex social behaviors with relevance to the public health, such as alcohol consumption, smoking or obesity has been tracked through social networks such as the 12 000 person Framingham Heart Study Cohort [36-38]. Studying the spread of registration for organ donation through the billion person Facebook social network will represent a similar opportunity to learn how such medical health related behaviors and decisions are transmitted within small and large interconnected groups. This knowledge will be helpful as we design future interventions to improve the low, static organ donation rates that continue to complicate the organ availability crisis in transplantation or reapproach other refractory public health problems via social media and social networks.