Information Reciprocity and ethical deliberation within marginalized networks

information flow netdraw

Shape of Nodes Legend (click on the embedded picture above to enlarge)

RED CIRCLE Medical professional

“Information Reciprocity and ethical deliberation within marginalized networks”
Michele Battle-Fisher
Unpublished manuscript

Renal disease is more prevalent among African Americans than any other ethnic group. Research has shown that while nearly 40% of people on kidney waiting lists are African American, only 23% of deceased organ donors are African American (Siminoff et al., 2006). The time on waiting lists is more pronounced for African Americans versus whites, on average 1335 days to 734 (Siminoff et al., 2006). For patients with chronic kidney disease (CKD), data shows that between 25 to 40% of referred CKD patients in the U.S. need transplantation within 3 months of this initial referral. It has been shown that fewer than 15% of African Americans with CKD knew that they had CKD (Siminoff et al., 2006). According to the Unites States Renal Data System (USRDS), African Americans over 20 represented nearly 38% of all dialysis patients but only 17% of all kidney transplants but are clear that every person is not able to donate due to a myriad of physical and ethical issues (U.S Renal Data System, 2009; Caplan & Coelho, 1998). Ethics involve controversial, socially uncomfortable issues that many would rather not face. Ethics force individuals to choose a philosophic camp that unlike some metaphysical beliefs may have to be acted upon. Because donation is an event that is time sensitive (limited resources and health status to donate), understanding the effect of the personal networks of subjects before the stress point of confronting having to make such a decision is due.
As a consequence, the overwhelming need of kidneys to sustain African Americans who would otherwise languish on dialysis is much greater than other groups. Altruistic donor donations after death have been insufficient in satisfying the rate of donation required to remove more individuals from waiting lists. This suggests the need for exploring barriers to living donation. The National Kidney Foundation and the American Societies of Transplantation, Transplant Surgeons, and Nephrology released a consensus statement outlining the guidelines in which a live organ donation can be performed (Abecassis et al., 2000). The consensus posed that the benefits to both donor and recipient must ethically outweigh the risk linked to procurement (Abecassis et al., 2000). That leaves potential donors to sift through influences within a social network that have never read or intention to read such edicts. Let us turn to the dynamics of how these ethical positions may crystallize within, what I call, a cloistered “marginalized network”.

Incidence and prevalence of CKD and ESRD among African Americans.
Why target African Americans? Look at the numbers. Chronic kidney disease (CKD) is a common public health issue. Nearly 26.3 million Americans live with the earlier stages of CKD (stages 1-4) (Coresh et al., 2007). Renal disease is more prevalent among African Americans than any other ethnic group. The time on waiting lists for kidneys is more pronounced for African Americans versus whites, on average 1335 days to 734 (Siminoff et al., 2006). None of this is good news. For patients with chronic kidney disease (CKD), between 25 to 40% of referred CKD patients in the U.S. require a transplant within 3 months of this initial referral. Too little plus too late? Yet it has been shown that fewer than 15% of African Americans with CKD knew that they had CKD (Cass et al., 2003; National Kidney Foundation, 2009). The before mentioned 25-40% referred for transplant would have been diagnosed at Stage 5 ESRD noting the detrimental impact of lack of early diagnosis and care on morbidity and mortality. This late referral is a substantial burden for both the patient and health care system. We need a better leverage point. Something is broken here.
According to the United States Renal Data System (USRDS), among adults tracked by the Third National Health and Nutrition Survey (NHAMES III) survey, African Americans over 20 represented nearly 38% of all ESRD patients but only 17% of all kidney transplants (U.S. Renal Data System, 2009). It is about time to use complexity to understand how investigating the patient’s social network might help highlight mechanisms to reduce the burden of scarcity of kidneys. Agreed?

Theory of Information Flow, Marginalization & Structural bias
How is information communicated in a network? First what is information and how might an ethic be a bit of “information”? I love Rapoport’s (1955) definition of “information” based on game theory. Simply put, information is so only if it presents new understanding or a new slant to an idea. It needs to become viral. In this case, information regards a living donation ethic could involve presenting a new, changed stance on organ donation or through sharing new information about ESRD in general. It is not enough to reinforce an old position. While it may be beneficial to know where everyone stands, this does nothing to force innovation and information flow. According to Frenzen and Nakamoto (1993) there are two elements to the theory of information flow: decision and structure. Both of these elements are inherent to the ethical deliberation question of this study. Frenzen and Nakamoto (1993) note tie strength and opportunity cost (such as pro versus con of donation) explains how information sharing happens. Moral hazard (such as ethics) impedes “perfect cooperation” in information sharing (Frenzen & Nakamoto, 1993). There are drawbridges for moral hazards by influential actors which control the flow of information (Frenzen & Nakamoto, 1993). But this flow is not done in a vacuum. In accordance to information cascade we obtain private information from the “environment” (Bikchandani et al., 2002). If it is not revealed into an accessible environment then there would be no deliberation or consensus possible. In other words, if there is some truncation of chance of infusing new and beneficial actors into that environment, this network becomes “marginalized”. I define a marginalized network by both the mathematical finiteness (structural bias assumed by transitivity) as well as socio-political constriction on life chances. There will be no new levering of information infused into the network that could change the tide of decision making. Information is only so if it is novel to the network (Rapoport, 1955). Everyone may not be involved directly in the ethical deliberation. The idea of marginalized networks plays great homage to Rapoport (1953). But he also importantly found in experimental studies that an inner circle of “knowers” may slow innovation. That inner circle may be stalwart in their beliefs and that circle, if it does let in new people, will only continue to percolate the same, tired ideas. So in the end, there must be a naturalization path for outsiders to come in the inner circle. In terms of randomness of distribution, new knowers and old, stalwart knowers intermingle. But once the outsiders are there, they may take on the difficult task of being an innovator or disappearing into the scenery and conform. Now before I go on, I understand that just because new individuals become added into a network that this may or may not be beneficial. We relate and we are drawn into ethical harmony with others that we care about. An ethic is not personal attribute, which it is not. Having a wish list for in the know members a network does one no good.

A kidney narrative- a hypothetical social support network of an ESRD patient
I have complied small relationship vignettes of a hypothetical African American patient named Shirley. UCInet 9.0 was used for the network analysis (Borgatti, Everett & Freeman, 2002 . Network visualization was performed using NetDraw 1.125 (Borgatti, 2002). There are two ways to look at networks as small personal universes surrounding one central person (egocentric) or as complete structures with no central figure (complete or full). Egocentric is great in terms of getting a narrative of the personal experience of “Shirley”. You sacrifice the big picture. With a complete network, you lose the intimacy by broadening your conceptual lens by analyzing the totality of network. For the sake of this example, I am more interested in showing how the dynamics on the whole might work. This is often a rare opportunity for this works not unlike a census. In order to get a sense of the overall embeddedness that is happening in this hypothetical network, each of the vertices (11 people in the network) would contribute relationship ties across all 11 of them. Could you imagine accomplishing this on a larger scale when you interview? With the advent of Big Data and platforms such as Gephi, tweets and Facebook posts can be data mined. These analyses are wonderful for the sexiness of the power innate to network analysis. But for today’s story, we will follow the lives of 11 people, one ill and the others in some way fated in her support network.
The patient
• Shirley-the ESRD patient; age 60; widowed, member of local church and Eastern Star; mother of John, grandmother of Kenneth; player of Euchre; few relatives within 100 miles; moved to city to take good paying job as a preschool teacher; forced retirement after 30 years due to dialysis schedule and worsening health; diabetic for 15 years after diagnosed in the emergency room; on Medicare since ESRD

The concerned others
• John-son of Shirley; age 41; divorced; pre-diabetic; works for local City government as a Sanitarian for 12 years; consistent health insurance utilization with government job; has sole custody of 16 year old, Kenneth; loves Cleveland Browns and travelling with son’s AAU basketball team but missing some games due to helping with Shirley’s medical emergencies; spoke with nephrologist about living donation option
• Dr. C.-family doctor of Shirley; pieced together Shirley’s spotty health history from low utilization of family physicians before her arrival
• Dr. K.-nephrologist of 6 months after Dr. C.’s referral; diagnosed Shirley’s Stage 5 after progressive uncontrolled diabetes and hypertension; approaches Shirley and John about ESRD and the need for transplantation
• Sylvia-best friend of Shirley; aged 66; constant companion of Shirley as she lives three train stops East of Shirley’s townhome; volunteers at nursing home; loves home improvement shows
• Ms. Mary-mother-in-law of Shirley; aged 87; single; Mother of her local Baptist Church; player of Euchre with neighbor, Tom and Shirley; suffers from heart problems
• Elder C.-head minister of Shirley’s church; preached “Womb to the Tomb” sermon last month; visits her monthly as a part of the Love the Sick-In ministry
• Josie-first cousin of Shirley; age 62; partnered; lives in childhood neighborhood of Shirley 100 miles away; few personal health concerns yet though weight is creeping up; raised two girls and enjoys time with grandchildren
• Tom- former neighbor of Shirley; Euchre partner
• Kenneth-16 year old grandson of Shirley and son of John; attends local Marist High School and plays AAU basketball; visits Shirley less for he cannot stand the thought of her on a machine
• Gail-local Eastern Star Masonic president; age 59; works with local churches to establish food pantries; used to see Shirley at meetings; little interaction outside of Masonic activities.

This is not a large network. But the size of the network does not mean that you are better off either if the resources are not optimally utilized. For Shirley, she has found 10 people that in some way, shape or form gets to know her as Shirley, the ESRD patient, the neighbor and friend. Mathematically, Shirley is not doing that badly. Since a network is defined by the “actor” and “relationship”, this mathematical parable becomes one of possibility, possibly intentionality to help. Of all of the possible ties (maximum density which occurs if everyone is connected to everyone else), only 34% of all possible ties are being used in this scenario. On average, each person in the network has an average of 3 ties. This leaves a marked range of the social butterfly that is in everyone’s beeswax to the purposeful isolate such as Gail who serves only one master in Shirley. Just by eyeballing the unusually small network, you could probably surmise who is who. The sum of the connections going out from an actor is called “out-degree” which may demonstrate influence. If you have a large “in-degree” which is defined as ties coming to you (arrow pointed to the node), then that person may be able to exert influence. In this network, Shirley and John, the son, are both influential and may wave their wand of persuasion over the network (demonstrated by the highest out-degree and in-degree centralities). But wait, being relied on so much can have its own stresses, on the person as well on the integrity of the “network”. Elder C., Tom, & Gail mathematically are the least likely characters to call upon in a crisis (by centrality measures). But they may also be the most reliable to the patient. Shirley has found a way in this scenario to leverage her resources, in terms of ties and connections. But most central bears the burden of over-reliance. This is also true for the structural integrity of networks. High centrality is a heightened factor in network failure. Without spreading the wealth, there may be a fissure coming in John’s inability and desire to help this mother. If John is gone, how does Shirley rebound and start anew with this new reality in front of her?

The benefit of exploring “networks”
Health has differing zones of effect on daily lives. The life of a spouse may be affected by the loss of income caused for the great time demand required by dialysis. Neighbors may be affected by an increased desire to extend more energy to check on the welfare of a neighbor. The neighbor and spouse may present weak or strong tie motivations that operationally differ toward the positions toward living organ donation. They all are a member of the patient’s social network linked by a shared connection.
The key assumptions of the social network paradigm are:
1. The aggregate influence of the group is more important than the individual attributes of each subject.
2. The analysis among actors is done at the level of the network not the actor in question though transformation to attribute level data is possible.
3. Patterns of relationship accounts are not merely made at bidirectional level of 2 actors, but across the whole network (Hanneman & Riddle, 2005).
Social network theory and analysis seeks to uncover the influence of linked relationship of members of a connected network (Hanneman & Riddle, 2005). By assessing the impact of network membership among a high risk population such as African Americans, it makes sense to determine the relationship of the potential donor to other actors, or connected persons in their personal social network (Hanneman & Riddle, 2005). Social network analysis allows highlighting the influence of the group dynamics on the subject’s individual relationships.
Individuals become patients in a social environment which requires negotiation of needs and resources with others within a bounded network that may be able to assist. While these networks may differ in appearance and relative importance across individuals, what is shared along all networks is that each patient will be in “relation” to someone with the knowledge or resources that are necessary to navigate an illness. Social network seeks not only to support the existence of attributes such as social support but to graphic demonstrate how this attribute reacts and flows specifically within the context of the donor’s personal network. For example, the existence of ties to an attribute such as social support no longer just assumed but is now used a means to investigate the structure in which social support is embedded (Smith & Christakis, 2008). This gives a far more nuanced exploration of the dynamic nature of the social network.

The benefit of exploring “marginalized networks” in small worlds
Ajrouch et al. (2001) specified found that minorities had less diffuse, denser networks. In addition, Blacks had more family members in their support systems and younger subjects had a larger proportion of kin than comparable whites in the study (Ajrouch et al., 2001). This would lead to possibility more strong ties among African Americans than other groups. This finding highlights the concentrated dense “first zone” connections that are to be explored for each patient as each connection may hold more influence over the decision making of the patient. Cornwell (2009) found that adults were more willing to engage in communication about health based on the quality of the person’s social integration. If a potential donor is embedded in his or her network, the social cohesion is linked to the denseness as actors should be willing to call upon the others in the network to bolster the strength of the network. Fingerman and Birditt (2003) approached the issue of family from a position of whether family would be deemed as close and or worrisome to deal with for older adults. While most ties of family were named as close and beneficial, younger adults were more apt to name fewer relatives and/or name those ties as “problematic”. So every member of the network should hypothetical hold some reason to affect the health of the potential living donor, if not solely for altruistic but also collective reasons.
Why explore health within the context of a network? Or within what I define as a “marginalized” network? The work of Marsden (1987) supports the notion of highly homogeneous support within American society. We associated with people like us. We relate and associate with those with like beliefs. We find safety with our own but become adversely constricted by the forces that separate us. Individuals become patients in a social environment which requires negotiation of needs and resources with others within a bounded network that may be able to assist. However it must be cautioned to assume that all networks have equal chances of adding diverse members. Outside of the digital world, socialization and connections happens where we live. A node must acknowledge that a link or tie exists. There must be an accounting on a political level that the emergence of new alters is not created equal in small worlds. Why does this inequality even matter? On a macro level, the micro structure becomes a motif for the divided realities of crossing paths with people unlike ourselves. Some time ago, the work of Rapoport (1953) noted a finite limitation of alters that may be accessible to any of us as a structural bias. We may only add to our friendships as we have access to a new tie then a declaration of friendship is made in some understood fashion. But often the biases become inexplicably tied to our life chances, those latent tentacles that push and pull on us often without us knowing. It does really become about who we know, both as a structural and political phenomenon. Living in the hollers of the Appalachian foothills will have its own structural boundaries and “structural biases”. A bridge that washes out in Tram keeps community ties intact but how can others come in? A child that lived in the now-demolished Cabrini Green formerly splicing Chicago’s Near North Side sky view spent years co-existing, and surviving with people that loved each other in spite of social constraints they felt powerless against. How does that child with a Cabrini identity to then leverage finding and maintaining beneficial ties outside of that Cabrini space? Marginalized populations are often inundated with hazards. In the case a moral hazard such as living organ donation, this is an interesting question to find out just what kind of additional hazards would be deemed as an acceptable risk.

Key questions that may arise
1. How might density of one’s network affect the nature of ethical deliberation and consensus building in terms of living kidney donation within in a population of African Americans?
2. How might the exchange of information in a network affect the ethical position taken by the subject? Is there an effect of “influential” actors that more strongly influence making and maintaining an ethical position?
3. How might the sharing of a “moral” hazard situation be shaped by the dynamic of the social network?
4. Is there agreement in position (color) within close, dense networks? In what situation does nonconformity to the prevailing network position occur or not occur?
5. In practical terms, to what extent does the subject hold the capability to independently choose an ethical position in terms of living kidney donation and exhibit freedom from controlling action?
6. Is there a difference between “moral” hazards and perceived social hazards?


Abecassis, M., Adams, M., Adams, P. et al. (2001). Consensus statement on the live organ donor. American Journal of the American Medical Association, 284, 2919-2926.

Ajrouch, K., Antonucci, T., & Janevic, M. (2001). Social Networks among Blacks and Whites: The Interaction between Race and Age. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 56(2), S112-S118.

Bikchandani, S, Hirshleifer, D., & Welsh, I. (1992).A theory of fads, fashion, custom, and cultural change as informational cascades.Journal of Political Economy, 100, 992-1026.

Borgatti, S. (2002). Netdraw Network Visualization. Harvard, MA: Analytical Technologies.

Borgatti, S.P., Everett, M.G. and Freeman, L.C. (2002) Ucinet for Windows: Software for Social Network Analysis. Harvard, MA: Analytic Technologies.

Caplan, A. & Coelho, D. (1998). “Introduction”. In Caplan, A. & Coehlo, D. (ed.). The Ethics of Organ Transplants: the current debate. Amherst, NY: Prometheus Books.

Cass, A., Cunningham, J., Snelling, P. & Ayanian, J. (November 2003). Late referral to a nephrologist reduces access to renal transplantation. American Journal of Kidney Diseases, 42(5), 1043-1049.

Coresh, J., Selvin, E., Stevens, L., et al. (2007). Prevalence of chronic kidney disease in the United States. Journal of American Medical Association, 298(17), 2038-2047.

Cornwell, B. (2009). Network bridging potential in later life: life course experiences and social network position. Journal of Aging and Health, 21, 129-154.

Fingerman, K. & Birditt, K. (2003). Do Age Differences in Close and Problematic Family Ties Reflect the Pool of Available Relatives? The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 58(2), P80-P87.

Frenzen, J. & Nakamoto, K. (December 1993). Structure, cooperation and the flow of market information. Journal of Consumer Research, 20, 360-375.

Hanneman, R., & Riddle, M. (2005). Introduction to Social Network Methods. Retrieved on Dec. 29, 2009 from

Marsden, P. (1987). Egocentric and sociocentric measures of network centrality. Social Networks, 24, 407-422.

National Kidney Foundation. (2009).Chronic Kidney Disease. Retrieved on Dec. 6, 2009 from

Rapport, A. (1953). A marginalized network highlights both the mathematical finiteness as well as socio-political constriction on life chances. Bulletin of Mathematical Biophysics. 15, 523-33.

Rapoport, A. (1955). What is information? Synthese, 9, 157-173.

Siminoff, L., Burant, C., & Ibrahim, S. (2006) Racial disparities in preferences and perceptions regarding organ donation. Journal of General Internal Medicine, 21, 995-1000.

Smith, K. & Christakis, N. (2008). Social networks and Health. Annual Review of Sociology, 34, 405-429.

U.S. Renal Data System (2009). USRD Atlas of Chronic Disease and End-Stage Renal Disease in the United States, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Bethesda, MD. Retrieved on December 4, 2009 from

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s