- Email: [email protected]
Socioeconomic factors as predictors of organ donation
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Socioeconomic factors as predictors of organ donation Malay B. Shah, MD,a,* Valery Vilchez, MD,a Adam Goble, BS,a Michael F. Daily, MD,a Jonathan C. Berger, MD, MHS,a Roberto Gedaly, MD,a and Derek A. DuBay, MDb a
Division of Abdominal Transplantation, College of Medicine, University of Kentucky, Lexington, Kentucky Division of Abdominal Transplantation, College of Medicine, Medical University of South Carolina, Charleston, South Carolina
b
article info
abstract
Article history:
Background: Despite numerous initiatives to increase solid organs for transplant, the gap
Received 9 March 2017
between donors and recipients widens. There is little in the literature identifying socio-
Received in revised form
economic predictors for donation. We evaluate the correlation between socioeconomic
28 July 2017
factors and familial authorization for donation.
Accepted 10 August 2017
Methods: A retrospective analysis of adult potential donor referrals between 2007 and 2012
Available online xxx
to our organ procurement organization (OPO) was performed. Potential donor information was obtained from the OPO database, death certificates, and the US Census Report. Data on
Keywords:
demographics, education, residence, income, registry status, cause and manner of death,
Organ donation
as well as OPO assessments and approach for donation were collected. End point was fa-
Socioeconomic
milial authorization for donation.
Familial authorization
Results: A total of 1059 potential donors were included, with an overall authorization rate of
Transplantation
47%. The majority was not on the donor registry (73%). Younger donors (18-39 y: odds
Disparities
ratio [OR] ¼ 4.9, P < 0.001; 40-60 y: OR ¼ 2.1, P < 0.001), higher levels of education (college: OR ¼ 2.5, P ¼ 0.005; graduate studies: OR ¼ 3.9, P ¼ 0.002), prior listing on the donor registry (OR ¼ 10.3, P < 0.001), and residence in counties with lower poverty rates than the US rates (OR ¼ 1.7, P ¼ 0.02) were independently associated with higher authorization rates. Decoupling (OR ¼ 3.1, P < 0.001) and donation first mentioned by the local health care provider (OR ¼ 1.8, P ¼ 0.01) were also independently associated with higher authorization rates. Conclusions: Donor registration correlated most strongly with the highest authorization rates. These results indicate that public educational efforts in populations with unfavorable socioeconomic considerations may be beneficial in improving donor registration. Collaborations with local providers as well as OPO in-hospital assessments and approach techniques can help with improving authorization rates. ª 2017 Elsevier Inc. All rights reserved.
Image is from the Donor Memorial Wall at the University of Kentucky AB Chandler Hospital. * Corresponding author. Liver Transplantation University of Kentucky 740 South Limestone, K301 Lexington, Kentucky 40536-0284. Tel.: þ1 859 323-4661; fax: þ1 859 257-3644. E-mail address: [email protected] (M.B. Shah). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.08.020
shah et al socioeconomic predictors of donation
Introduction The need for transplantable organs continues to increase faster than the supply of organs. As of April 27, 2016, there are 120,963 patients in need of solid organ transplant in the United States.1 There has unfortunately been a relative stagnation of potential organ donors nationally since 2007.2 The organ shortage continues to be a worsening public health crisis. Numerous programs such as the Organ Donation Breakthrough Collaborative and Transplant Growth and Management Collaborative have been implemented to identify best practices associated with increases in organ donation.3 In addition, aggressive management of potential organ donors, as well as the use of resuscitative techniques, has helped maintain the donor pool and salvage of transplantable organs.4,5 Despite these initiatives and changes in the medical management of donors, the gap between donors and recipients continues to widen. This leads to several questions regarding the organ donation. Have organ procurement organizations (OPOs) maximized the number of potential donors? Have OPOs and transplant providers maximized collaborations to educate the public about organ donation? While these are questions that are examined elsewhere, we attempted to answer the question of whether transplant providers have identified at-risk populations for poor familial authorization rates for organ donation. Most decedents eligible for organ donation are not registered organ donors, and their families are approached for authorization for organ donation.6,7 Studies have attempted to identify specific barriers that may negatively impact rates of familial authorization for donation. Some studies have indicated that the technique, timing, sequence, and who approaches the family of the decedent for consent significantly impact the ability to obtain authorization.8,9 Other studies have examined decedent demographics and social characteristics that are associated with familial authorization for donation. Non-Caucasian race has repeatedly been associated with nondonation, whereas medical causes of death have also been shown to be associated with nondonation.3,10-12 Unfortunately, there is a gap in the literature regarding other socioeconomic predictors of donation. It is therefore important to identify the characteristics of eligible decedents who are unlikely to donate. Identification of these characteristics can allow sufficient resource allocation to prehumous phase and an optimal familial approach in the posthumous phase. The purpose of this investigation is to examine a large OPO decedent referral database. The goal is to measure the association between socioeconomic factors and familial authorization, while controlling for variables known to be associated with familial authorization (e.g., race and approach factors). We hypothesize that lower socioeconomic status and reduced education are associated with decreased familial authorization.
Methods A retrospective analysis of referrals to our local OPO, the Kentucky Organ Donor Affiliates (KODA), between May 2007
89
and December 2012 was performed. All potential organ donor referrals, aged 18 y or older, were included. Candidate donor characteristics were extracted from the KODA database. Variables collected included age, gender, ethnicity, registry status, hospital where the donor died, cause of death, manner of death (medical or trauma), which family member(s) was approached for donation, who first mentioned donation (family, local provider, or OPO), whether or not the family had an understanding of the hopelessness of their loved one’s situation (nonsurvivable injury), and whether or not decoupling of the pronouncement of death from the approach for donation occurred.13 Family understanding of hopelessness was assessed by the OPO by asking the decedent’s family what they understood about the condition of their loved one. If it was deemed that the family did not comprehend the nonsurvivable condition, then the process for authorization was halted, and the family was approached later. Additional information about the potential donor was obtained from the death certificates from the Kentucky Office of Vital Statistics and West Virginia Health Statistics Center. These documents provided candidate donors’ home address, marital status, education level, leading cause of death, and manner of death. Home address was used to determine county of residence, geographic area of residence (Appalachia versus non-Appalachia), and mileage from donor residence to the hospital of death. Additional data were collected from the US Census Report, which included median county household income and percentage of individuals below the poverty line. The median county household income was used as a surrogate for decedent family income in the given year of the donor death. The percentage of individuals below the poverty line was obtained in the given year of the donor death. The primary outcome of the analysis was familial authorization for organ donation. A case-control comparison was performed between potential donors who had authorization for donation and those whose authorization was declined. A univariable analysis for all variables was performed using the Student’s t-test for continuous variables and Pearson’s c2 test with Yates correction for categorical variables. Variables with a P value < 0.05 were entered into the multivariable logistic regression analysis. Statistical significance was set at P < 0.05. The institutional review board was consulted, and it was determined that the study qualified for a waiver of informed consent. This was due to the fact that a retrospective analysis was performed on decedents, and no family members were contacted to obtain any additional information beyond data already collected by KODA and available in death certificates.
Results During the 6-y study period, there were 1059 potential organ donor referrals. Table 1 summarizes the demographics of our study population. There were a higher percentage of males in our study. Our study population was predominantly Caucasian. The average age of the potential donor was 48 15 y. Most of the patients in our population had at least a high school level of education, followed by college level. The median household income was $39,915 8667, with the majority
90
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
Table 1 e Demographics (n [ 1059). Demographics
n (%)
Gender Male
613 (58)
Female
452 (42)
Race Caucasian African-American
946 (89) 94 (9)
Age category 18-39
305 (29)
40-60
502 (47)
>60
252 (24)
Education 8th grade 9-12th grade
90 (9) 90 (9)
High school grad/General Educational Development
430 (41)
College
219 (21)
Grad school
45 (4)
Marital status Married
508 (48)
Divorced
219 (21)
Single
250 (24)
Widowed
81 (7)
Appalachia residence Yes
445 (42)
No
613 (58)
Income category <$35,000
323 (31)
$35,000-$45,000
434 (41)
>$45,000
301 (28)
County/state poverty Below
512 (48)
Above
547 (52)
County/US poverty Below
151 (14)
Above
908 (86)
United Network for Organ Sharing cause of death Cerebrovascular accident
513 (48)
Head trauma
284 (27)
Anoxia
220 (21)
Other
40 (4)
Manner of death Medical
661 (62)
Trauma
389 (37)
Registry status Yes
163 (15)
No
767 (73)
Unknown status
129 (12)
Overall authorization Yes
501 (47)
No
558 (53)
of potential donors falling below $45,000. The average poverty rate of potential donors was 19.5% 6.3, compared with the overall state poverty rate of 17.8% 1.5 and the US poverty rate of 14.6% 1.2. The majority of potential donors resided in counties with poverty rates higher than the US rates, whereas a considerable number lived in Appalachia. Only 15% of potential donors were on the donor registry. Familial authorization was obtained in 47% of the study population. A univariable analysis was performed between decedents for which familial authorization was obtained (“authorization” group) and compared with those where familial authorization was declined (“decline” group). Table 2 represents decedent characteristics. There were no significant differences between gender and race. However, significantly higher authorization rates were observed in decedents aged between 18-39 y (P < 0.001), those who were on the donor registry (P < 0.001), and had a head injury as a cause of death (P < 0.001) or traumatic manner of death (P < 0.001). In addition, significantly higher authorization rates were observed in decedents with the following socioeconomic factors: single (P ¼ 0.001), college or higher levels of education (P < 0.001), residents of non-Appalachian geographic areas (P ¼ 0.001), hospital located in non-Appalachian regions (P ¼ 0.02), family median household incomes >$45,000 (P < 0.001), and residents of counties with poverty rates lower than the state (P < 0.001) and national (P < 0.001) poverty rates. Table 3 represents characteristics of family members approached for donation and shows no differences in familial authorization based on gender of the family member approached for donation nor in the distance traveled from the potential donor residence to the hospital of death. Approaching the parent of a potential donor was associated with significantly higher rates of authorization compared with approaching a spouse, sibling, or child (P < 0.001). Significantly higher rates of authorization were seen when the approach for donation was decoupled from the declaration of death (P < 0.001) and when the family understood the hopelessness of the situation of their loved one (P ¼ 0.002). Higher rates of authorization were also seen in potential donors when the family initially inquired about the possibility of donation (P < 0.001). A multivariable logistical regression was performed to identify factors independently associated with increased conversion rates (Table 4). Again, younger donors (18-39 y: odds ratio [OR] ¼ 4.9, confidence interval [CI] ¼ 3.1-7.7, P < 0.001; 40-60 y: OR ¼ 2.1, CI ¼ 1.4-3.1, P < 0.001), as well as candidates with higher levels of education (college: OR ¼ 2.5, CI ¼ 1.3-4.6, P ¼ 0.005; graduate studies: OR ¼ 3.9, CI ¼ 1.6-9.4, P ¼ 0.002), had significantly higher rates of authorization. Prior listing on the donor registry was also associated with higher rates of authorization (OR ¼ 10.3, CI ¼ 6.0-17.7, P < 0.001), as did residing in counties with overall poverty rates lower than the US rates (OR ¼ 1.7, CI ¼ 1.1-2.6, P ¼ 0.02). Decoupling the approach for donation from the declaration of death was associated with higher rates of authorization (OR ¼ 3.1, CI ¼ 1.8-5.4, P < 0.001). Donation first mentioned by a local health care provider was also associated with higher authorization rates compared with the first approach by the OPO
91
shah et al socioeconomic predictors of donation
Table 2 e Univariable analysis of potential donor characteristics between decline and authorization groups. Donor factors impacting authorization
Decline Authorization (n ¼ 558), (n ¼ 501), n (%) n (%)
Table 2 e (continued ) Donor factors impacting authorization P Other
Male
325 (53)
288 (47)
Female
233 (52)
213 (48)
493 (52)
453 (48)
56 (60)
38 (40)
18-39
102 (33)
203 (67)
40-60
272 (54)
230 (46)
>60
184 (73)
68 (27)
8th grade
64 (71)
26 (29)
9-12th grade
49 (54)
41 (46)
235 (55)
195 (45)
College
79 (36)
140 (64)
Grad school
18 (40)
27 (60)
Married
281 (55)
227 (45)
Divorced
120 (55)
99 (45)
Single
105 (42)
145 (58)
51 (63)
30 (37)
Yes
262 (59)
183 (41)
No
295 (48)
318 (52)
0.8
Race
African-American
Medical
399 (60)
262 (40)
Trauma
156 (40)
233 (60)
<0.001
Discussion
<0.001
Marital status 0.001
Appalachia residence 0.001
Registry status Yes
19 (12)
144 (88)
No
456 (60)
311 (40)
<$35,000
194 (60)
129 (40)
$35,000-$45,000
235 (54)
199 (46)
>$45,000
128 (42)
173 (58)
Below
239 (47)
273 (53)
Above
319 (58)
228 (42)
<0.001
<0.001
<0.001
Education
Widowed
13 (32)
(OR ¼ 1.8, CI ¼ 1.1-2.9, P ¼ 0.01). And, although it did not reach statistical significance, there was a trend toward head trauma (P ¼ 0.06) and potential donor residence outside of Appalachia (P ¼ 0.07) being associated with higher rates of authorization. To eliminate a potential selection bias of registered donors confounding our results, a subgroup analysis of 767 potential donors was performed analyzing only potential donors who were not on the donor registry. The univariable analysis is not included in this summary, but a multivariable analysis was performed using only factors from the univariable analysis that achieved statistical significance. Table 5 illustrates the multivariable analysis for factors associated with increased rates of authorization in potential donors not on the registry. Again, younger donors (18-39 y: OR ¼ 5.4, CI 3.3-8.6, P < 0.001; 40-60 y: OR ¼ 2.2, CI ¼ 1.4-3.3, P < 0.001), as well as candidates with higher levels of education (college: OR ¼ 2.9, CI ¼ 1.4-5.9, P ¼ 0.003; graduate studies: OR ¼ 3.4, CI ¼ 1.2-9.4, P ¼ 0.02), had significantly higher rates of authorization. Decoupling was again associated with higher rates of authorization (OR ¼ 3.2, CI ¼ 1.7-5.9, P < 0.001). Donation first mentioned by a local health care provider was again associated with higher authorization rates compared with that mentioned by the OPO (OR ¼ 1.8, CI ¼ 1.1-3.1, P ¼ 0.02). Although not quite reaching statistical significance, there was a trend observed toward higher authorization rates in potential donors residing in counties with poverty rates lower than the US rates.
0.3
Age category
High school grad/General Educational Development
27 (68)
P
Manner of death
Gender
Caucasian
Decline Authorization (n ¼ 558), (n ¼ 501), n (%) n (%)
Income category
County/state poverty <0.001
County/US poverty Below
59 (39)
92 (61)
Above
499 (55)
409 (45)
Yes
151 (59)
105 (41)
No
407 (51)
396 (49)
Cerebrovascular accident
305 (60)
208 (40)
Head trauma
106 (37)
178 (63)
Anoxia
120 (55)
100 (45)
<0.001
After controlling for known factors associated with familial authorization, we were able to demonstrate several new socioeconomic factors not previously recognized as being associated with familial authorization. Multiple factors clearly contribute to disparities in donation and authorization rates. To begin to identify at-risk groups for lower conversion rates, a two-phase approach examining prehospital and in-hospital assessments as separate entities is prudent.
Hospital region in Appalachia 0.02
United Network for Organ Sharing cause of death <0.001
(continued)
Prehospital assessment It is well established that deceased donation rates are improved when the potential donor has made a previous decision to donate prehumously.14,15 Prior designation of donor registration likely leads to easing the burden of the decisionmaking process for donation of the potential donor’s family.3 We observed that prior donor designation was associated with an 11-fold increased odds of authorization (P < 0.001).
92
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
Table 3 e Univariable analysis of family member characteristics between decline and authorization groups. Factors involved with family decision making
Decline Authorization (n ¼ 558), (n ¼ 501), n (%) n (%)
Table 4 e Multivariable analysis of predictors for increased rates of authorization. Predictors P
373 (57)
283 (43)
Local health care provider
85 (54)
73 (46)
Family
61 (31)
137 (69)
185 (54)
161 (46)
Female
304 (51)
288 (49)
Spouse
236 (56)
188 (44)
Parent
72 (33)
147 (67)
Sibling
42 (58)
30 (42)
<0.001
139 (62)
84 (38)
0.5
<0.001
467 (51)
455 (49)
No
24 (63)
14 (37)
Yes
409 (49)
427 (51)
No
72 (67)
36 (33)
<25 miles
254 (52)
237 (48)
>25 miles
303 (54)
263 (46)
1.08
0.52-2.24
0.8
High school grad/General Educational Development
1.42
0.80-2.52
0.2
College
2.46
1.31-4.60
0.005
Graduate studies
3.91
1.62-9.44
0.002
Donation first mentioned by
Local health care provider
1.80
1.13-2.86
0.01
Family
2.55
1.71-3.80
<0.001
1.33
0.97-1.82
0.07
Anoxia
0.91
0.62-1.34
0.6
Head trauma
1.43
0.98-2.08
0.06
10.28
5.97-17.69
<0.001
3.11
1.80-5.36
<0.001
1.70
1.11-2.62
0.02
40-60
2.11
1.44-3.10
<0.001
18-39
4.88
3.08-7.74
<0.001
Residence in Appalachia Yes (referent) No
Family understanding of hopelessness Yes
9-12th grade
OPO (referent)
Family member approached
Children
P value
<8th grade (referent)
Gender of family member approached Male
95% CI
Education
Donation first mentioned by OPO
Odds ratio
0.002
United Network for Organ Sharing cause of death CVA/stroke (referent)
Decoupled <0.001
Registry status
Miles from home
No (referent) 0.8
Yes Decoupling No (referent) Yes County/US poverty
However, there are a multitude of other potential factors that, once identified and targeted, can potentially increase donor authorization rates. We also demonstrate that increasing education levels are significantly associated with increased rates of authorization for organ donation. This was significant in potential donors with college or higher levels of education. It has been suggested that individuals with higher education levels have more positive perceptions of organ donation.14 Our data demonstrate a stepwise decrease in familial authorization with increasing decedent age. The highest frequency of familial authorization was observed in the youngest decedent group, 18-39 y of age. Our data also demonstrate that parental approach was significantly associated with increased authorization rates, which may account for the increased familial authorization in the youngest decedents. This particular observation is likely multifactorial. Although parental approach has been examined in the pediatric donor population to some degree, factors that influence parental consent warrant additional study.16 In potential donors older than 60 y, there is likely to be more spousal, child, or sibling involvement who potentially have their own preconceived notions regarding donation and the fact that their loved one is simply “too old” to be a suitable donor.3 To further identify at-risk populations for poor authorization rates, we examined poverty levels in potential donors. We looked at poverty rates in counties where the potential donor
Above (referent) Below Age (y) >60 (referent)
CVA ¼ cerebrovascular accident.
resided in and compared that with the US and state poverty rates. Potential donors who resided in counties with higher poverty rates than the US rates were significantly less likely to donate. It is not surprising that counties with higher poverty levels potentially correlate with fewer years of formal education and fewer jobs with higher financial benefit. However, even when controlling for education and other socioeconomic factors, county poverty rate higher than the US rate remained independently associated with poor authorization rates. Donation rates were also examined in potential donors who resided in Appalachia. Appalachia is known as an area with high poverty rates, high unemployment, and health care disparities. Although residence in Appalachia was strongly correlated with poor authorization rates in a univariable analysis, this did not reach statistical significance in the multivariable analysis. This suggests that residence in
shah et al socioeconomic predictors of donation
Table 5 e Multivariable analysis of predictors for increased rates of authorization in donors not registered for donation. Predictors
Odds ratio
95% CI
P value
Education <8th grade (referent) 9-12th grade
1.18
0.53-2.65
0.6
High school grad/General Educational Development
1.62
0.84-3.11
0.1
College
2.90
1.44-5.85
0.003
Graduate studies
3.35
1.20-9.36
0.02
Local health care provider
1.83
1.10-3.05
0.02
Family
2.96
1.89-4.63
<0.001
3.16
1.69-5.91
<0.001
1.56
0.98-2.48
0.06
40-60
2.15
1.40-3.30
<0.001
18-39
5.35
3.32-8.60
<0.001
Donation first mentioned by OPO (referent)
Decoupling No (referent) Yes County poverty rates Higher than US rates (referent) Lower than US rates Age (y) >60 (referent)
Appalachia is a surrogate marker for poverty, which is associated with poor authorization rates. Our data suggest that targeting economically disparate areas may be beneficial in increasing donor registry.
In-hospital assessment Numerous studies have examined factors impacting authorization rates when potential donors are in the hospital. The OPO techniques for approach, involvement of an OPO family support representative, cause and manner of death, and donor hospital education programs have been examined by others.3 Similar to other OPOs, KODA uses a structured approach when contacted by the local health care team about a potential donor. This approach includes (1) assessment of the family; (2) assessment with the local health care team regarding family visitation with their loved one, timing of the approach, privacy for discussions, and introduction of KODA staff to the donor family; (3) assessing family understanding of the condition of their loved one; (4) providing education and allowing for questions regarding organ donation; (5) allowing family time to make a decision and providing informed consent for organ donation. Similar to other studies, our study also demonstrated that cerebrovascular accidents and anoxia as the causes of death were significantly associated with poor authorization rates.3 This may be explained in that these causes of death are typically seen in older individuals, which is an independent negative predictor for donation. This suggests that medical reasons
93
for death are surrogate markers for increased age, which is an important independent negative risk factor for donation. Decoupling is a strategy employed by OPOs as a way to approach grieving family members. Conceptually, this involves separating the declaration of death to the family from the approach for donation. Our study clearly demonstrates that when the technique of decoupling is employed, there is a significantly higher rate of authorization for donation. It has been suggested that the success of using this technique may come from comforting and letting the family members cope the process of grieving their loved one. Also interesting are the rates of familial authorization when examining who made the first approach for donation. It is widely reported that the first approach for donation by an OPO representative is strongly correlated with increased rates of authorization compared with a local health care provider’s approach.8,9 It has been suggested that this correlation is observed because an OPO representative has extensive training in handling challenging family dynamics, the occurrence of death, and providing education regarding organ donation. However, one of the most interesting findings of this study is that we observed that the first approach by the OPO was significantly associated with lower authorization rates compared with an approach by the local health care provider. We suspect multifactorial causes for this observation. Our population is largely rural with higher rates of poverty and fewer years of formal education compared with the US averages. These factors influence cultural attitudes and biases in our population toward organ donation that could explain how these families may trust and rely on their own health care provider and less on the OPO representative, who may be perceived as an “outsider.”17 Excluding potential donors who made a predeath donor designation eliminated potential selection bias of those who had actively made a decision to donate. However, the same variables remained associated with familial authorization in this sensitivity analysis. Including this subgroup analysis helped to confirm the validity of our analysis of the entire data set. There are limitations in this study. This was a retrospective case-control study using historical data from an approximate 6-y period. Studies have documented that African-Americans and other minorities have significantly lower rates of authorization for donation compared with Caucasians.12,14 Although there was a trend toward similar findings in our study, our population is homogenous in its ethnic breakdown. We suspect that this finding did not reach statistical significance because of lower numbers of African-Americans in our study population. An additional limitation is that actual donor household income was not collected by the OPO. Thus, median county income was collected by county of residence and is limited by data collection practices used by the US Census Bureau. A number of studies have used this technique to assess socioeconomic status.12,18,19 We acknowledge that county populations are not necessarily homogenous. Over a large sample size, however, results may be somewhat more generalizable for meaningful conclusions. No effort was made to identify potential donor religious affiliations, which may contribute to specific attitudes and beliefs toward organ donation. There were not any data collected in follow-up with potential donor families regarding their thoughts about why they chose to donate or not.
94
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
In summary, our study evaluated an underserved population with unique characteristics to identify factors associated with nonauthorization. These data may be used to identify populations at risk for nondonation, a population where the OPO may choose to allocate increased resources during the family approach. Importantly, donor registration correlated most strongly with higher authorization rates. These results suggest that focused public educational efforts regarding organ donation in populations with unfavorable socioeconomic considerations, such as populations with high poverty levels or fewer years of formal education, may be beneficial in improving donor registration. However, it is noted that great care must be taken to prevent increasing mistrust with the health care system, donation, and transplantation and that very focused and population-specific educational materials be provided.14,20 Additional studies are warranted on methods on how best to accomplish this. Focused educational efforts in areas and groups where older individuals may reside and engage in daily activities may also be beneficial in improving donor registration. It has indeed been shown that targeted efforts through various media campaigns and culturally sensitive education programs can make a positive impact on donor registration.21 In addition, in-hospital assessments to include techniques of approaching families for authorization (i.e., decoupling) can also play a significant role in increasing authorization for donation. Education of and collaboration with local health care providers may also be a very effective method of improving authorization rates.
Acknowledgment The authors acknowledge the Kentucky Organ Donor Affiliates for providing the database of potential organ donor referrals. They also acknowledge the Kentucky Office of Vital Statistics and West Virginia Health Statistics Center for providing the database of death certificates. Authors’ contributions: M.S., A.G., J.B., R.G., and D.D. contributed to conception and design of the study. M.S., V.V., and A.G. contributed to data acquisition. M.S., V.V., A.G., M.D., J.B., R.G., and D.D. contributed to analysis and interpretations. M.S., V.V., A.G., and D.D. drafted the manuscript. M.S., V.V., A.G., M.D., J.B., R.G., and D.D. critically revised the manuscript. M.S., A.G., M.D., J.B., R.G., and D.D. contributed to statistical analysis and interpretation.
Disclosure The authors reported no proprietary or commercial interest in any product mentioned or concept discussed in the article.
references
1. Number of patients Waiting for organ transplant. Richmond, Virginia: United Network for Organ Sharing; 2016.
2. Donors Recovered in the U.S. Richmond, Virginia: OPTN: US Department of Health and Human Services; 2014. 3. Brown CV, Foulkrod KH, Dworaczyk S, et al. Barriers to obtaining family consent for potential organ donors. J Trauma. 2010;68:447e451. 4. Salim A, Velmahos GC, Brown C, Belzberg H, Demetriades D. Aggressive organ donor management significantly increases the number of organs available for transplantation. J Trauma. 2005;58:991e994. 5. Salim A, Martin M, Brown C, et al. Using thyroid hormone in brain-dead donors to maximize the number of organs available for transplantation. Clin Transplant. 2007;21:405e409. 6. Chon WJ, Josephson MA, Gordon EJ, et al. When the living and the deceased cannot agree on organ donation: a survey of US organ procurement organizations (OPOs). Am J Transplant. 2014;14:172e177. 7. UNOS Star Files. OPTN database. Richmond, Virginia: OPTN: US Department of Health and Human Services; 2014. 8. Ebadat A, Brown CV, Ali S, et al. Improving organ donation rates by modifying the family approach process. J Trauma Acute Care Surg. 2014;76:1473e1475. 9. Dubay DA, Redden DT, Haque A, et al. Do trained specialists solicit familial authorization at equal frequency, regardless of deceased donor characteristics? Prog Transplant. 2013;23:290e296. 10. Siminoff LA, Burant CJ, Ibrahim SA. Racial disparities in preferences and perceptions regarding organ donation. J Gen Intern Med. 2006;21:995e1000. 11. Salim A, Berry C, Ley EJ, et al. The impact of race on organ donation rates in Southern California. J Am Coll Surg. 2010;211:596e600. 12. DuBay D, Redden D, Haque A, et al. Is decedent race an independent predictor of organ donor consent or merely a surrogate marker of socioeconomic status? Transplantation. 2012;94:873e878. 13. Gortmaker SL, Beasley CL, Sheehy E, et al. Improving the request process to increase family consent for organ donation. J Transpl Coord. 1998;8:210e217. 14. Regalia K, Zheng P, Sillau S, et al. Demographic factors affect willingness to register as an organ donor more than a personal relationship with a transplant candidate. Dig Dis Sci. 2014;59:1386e1391. 15. National donor designation Report card. Donate Life America; 2014. Available at: https://www.donatelife.net/wp-content/ uploads/2016/06/Report-Card-2014-44222-Final.pdf. 16. Rodrigue JR, Cornell DL, Howard RJ. Pediatric organ donation: what factors most influence parents’ donation decisions? Pediatr Crit Care Med. 2008;9:180e185. 17. Traino HM, Alolod GP, Shafer T, Siminoff LA. Interim results of a national test of the rapid assessment of hospital procurement barriers in donation (RAPiD). Am J Transplant. 2012;12:3094e3103. 18. Adler JT, Hyder JA, Elias N, et al. Socioeconomic status and ethnicity of deceased donor kidney recipients compared to their donors. Am J Transplant. 2015;15:1061e1067. 19. Mohan S, Mutell R, Patzer RE, Holt J, Cohen D, McClellan W. Kidney transplantation and the intensity of poverty in the contiguous United States. Transplantation. 2014;98: 640e645. 20. Reibel M, Olmo C, Andrada S, Koertzen J. Deep demographics: understanding local Variation in donor registration. Prog Transplant. 2016;26:191e198. 21. Salim A, Ley EJ, Berry C, et al. Effect of community educational interventions on rate of organ donation among Hispanic Americans. JAMA Surg. 2014;149:899e902.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Socioeconomic factors as predictors of organ donation Malay B. Shah, MD,a,* Valery Vilchez, MD,a Adam Goble, BS,a Michael F. Daily, MD,a Jonathan C. Berger, MD, MHS,a Roberto Gedaly, MD,a and Derek A. DuBay, MDb a
Division of Abdominal Transplantation, College of Medicine, University of Kentucky, Lexington, Kentucky Division of Abdominal Transplantation, College of Medicine, Medical University of South Carolina, Charleston, South Carolina
b
article info
abstract
Article history:
Background: Despite numerous initiatives to increase solid organs for transplant, the gap
Received 9 March 2017
between donors and recipients widens. There is little in the literature identifying socio-
Received in revised form
economic predictors for donation. We evaluate the correlation between socioeconomic
28 July 2017
factors and familial authorization for donation.
Accepted 10 August 2017
Methods: A retrospective analysis of adult potential donor referrals between 2007 and 2012
Available online xxx
to our organ procurement organization (OPO) was performed. Potential donor information was obtained from the OPO database, death certificates, and the US Census Report. Data on
Keywords:
demographics, education, residence, income, registry status, cause and manner of death,
Organ donation
as well as OPO assessments and approach for donation were collected. End point was fa-
Socioeconomic
milial authorization for donation.
Familial authorization
Results: A total of 1059 potential donors were included, with an overall authorization rate of
Transplantation
47%. The majority was not on the donor registry (73%). Younger donors (18-39 y: odds
Disparities
ratio [OR] ¼ 4.9, P < 0.001; 40-60 y: OR ¼ 2.1, P < 0.001), higher levels of education (college: OR ¼ 2.5, P ¼ 0.005; graduate studies: OR ¼ 3.9, P ¼ 0.002), prior listing on the donor registry (OR ¼ 10.3, P < 0.001), and residence in counties with lower poverty rates than the US rates (OR ¼ 1.7, P ¼ 0.02) were independently associated with higher authorization rates. Decoupling (OR ¼ 3.1, P < 0.001) and donation first mentioned by the local health care provider (OR ¼ 1.8, P ¼ 0.01) were also independently associated with higher authorization rates. Conclusions: Donor registration correlated most strongly with the highest authorization rates. These results indicate that public educational efforts in populations with unfavorable socioeconomic considerations may be beneficial in improving donor registration. Collaborations with local providers as well as OPO in-hospital assessments and approach techniques can help with improving authorization rates. ª 2017 Elsevier Inc. All rights reserved.
Image is from the Donor Memorial Wall at the University of Kentucky AB Chandler Hospital. * Corresponding author. Liver Transplantation University of Kentucky 740 South Limestone, K301 Lexington, Kentucky 40536-0284. Tel.: þ1 859 323-4661; fax: þ1 859 257-3644. E-mail address: [email protected] (M.B. Shah). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.08.020
shah et al socioeconomic predictors of donation
Introduction The need for transplantable organs continues to increase faster than the supply of organs. As of April 27, 2016, there are 120,963 patients in need of solid organ transplant in the United States.1 There has unfortunately been a relative stagnation of potential organ donors nationally since 2007.2 The organ shortage continues to be a worsening public health crisis. Numerous programs such as the Organ Donation Breakthrough Collaborative and Transplant Growth and Management Collaborative have been implemented to identify best practices associated with increases in organ donation.3 In addition, aggressive management of potential organ donors, as well as the use of resuscitative techniques, has helped maintain the donor pool and salvage of transplantable organs.4,5 Despite these initiatives and changes in the medical management of donors, the gap between donors and recipients continues to widen. This leads to several questions regarding the organ donation. Have organ procurement organizations (OPOs) maximized the number of potential donors? Have OPOs and transplant providers maximized collaborations to educate the public about organ donation? While these are questions that are examined elsewhere, we attempted to answer the question of whether transplant providers have identified at-risk populations for poor familial authorization rates for organ donation. Most decedents eligible for organ donation are not registered organ donors, and their families are approached for authorization for organ donation.6,7 Studies have attempted to identify specific barriers that may negatively impact rates of familial authorization for donation. Some studies have indicated that the technique, timing, sequence, and who approaches the family of the decedent for consent significantly impact the ability to obtain authorization.8,9 Other studies have examined decedent demographics and social characteristics that are associated with familial authorization for donation. Non-Caucasian race has repeatedly been associated with nondonation, whereas medical causes of death have also been shown to be associated with nondonation.3,10-12 Unfortunately, there is a gap in the literature regarding other socioeconomic predictors of donation. It is therefore important to identify the characteristics of eligible decedents who are unlikely to donate. Identification of these characteristics can allow sufficient resource allocation to prehumous phase and an optimal familial approach in the posthumous phase. The purpose of this investigation is to examine a large OPO decedent referral database. The goal is to measure the association between socioeconomic factors and familial authorization, while controlling for variables known to be associated with familial authorization (e.g., race and approach factors). We hypothesize that lower socioeconomic status and reduced education are associated with decreased familial authorization.
Methods A retrospective analysis of referrals to our local OPO, the Kentucky Organ Donor Affiliates (KODA), between May 2007
89
and December 2012 was performed. All potential organ donor referrals, aged 18 y or older, were included. Candidate donor characteristics were extracted from the KODA database. Variables collected included age, gender, ethnicity, registry status, hospital where the donor died, cause of death, manner of death (medical or trauma), which family member(s) was approached for donation, who first mentioned donation (family, local provider, or OPO), whether or not the family had an understanding of the hopelessness of their loved one’s situation (nonsurvivable injury), and whether or not decoupling of the pronouncement of death from the approach for donation occurred.13 Family understanding of hopelessness was assessed by the OPO by asking the decedent’s family what they understood about the condition of their loved one. If it was deemed that the family did not comprehend the nonsurvivable condition, then the process for authorization was halted, and the family was approached later. Additional information about the potential donor was obtained from the death certificates from the Kentucky Office of Vital Statistics and West Virginia Health Statistics Center. These documents provided candidate donors’ home address, marital status, education level, leading cause of death, and manner of death. Home address was used to determine county of residence, geographic area of residence (Appalachia versus non-Appalachia), and mileage from donor residence to the hospital of death. Additional data were collected from the US Census Report, which included median county household income and percentage of individuals below the poverty line. The median county household income was used as a surrogate for decedent family income in the given year of the donor death. The percentage of individuals below the poverty line was obtained in the given year of the donor death. The primary outcome of the analysis was familial authorization for organ donation. A case-control comparison was performed between potential donors who had authorization for donation and those whose authorization was declined. A univariable analysis for all variables was performed using the Student’s t-test for continuous variables and Pearson’s c2 test with Yates correction for categorical variables. Variables with a P value < 0.05 were entered into the multivariable logistic regression analysis. Statistical significance was set at P < 0.05. The institutional review board was consulted, and it was determined that the study qualified for a waiver of informed consent. This was due to the fact that a retrospective analysis was performed on decedents, and no family members were contacted to obtain any additional information beyond data already collected by KODA and available in death certificates.
Results During the 6-y study period, there were 1059 potential organ donor referrals. Table 1 summarizes the demographics of our study population. There were a higher percentage of males in our study. Our study population was predominantly Caucasian. The average age of the potential donor was 48 15 y. Most of the patients in our population had at least a high school level of education, followed by college level. The median household income was $39,915 8667, with the majority
90
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
Table 1 e Demographics (n [ 1059). Demographics
n (%)
Gender Male
613 (58)
Female
452 (42)
Race Caucasian African-American
946 (89) 94 (9)
Age category 18-39
305 (29)
40-60
502 (47)
>60
252 (24)
Education 8th grade 9-12th grade
90 (9) 90 (9)
High school grad/General Educational Development
430 (41)
College
219 (21)
Grad school
45 (4)
Marital status Married
508 (48)
Divorced
219 (21)
Single
250 (24)
Widowed
81 (7)
Appalachia residence Yes
445 (42)
No
613 (58)
Income category <$35,000
323 (31)
$35,000-$45,000
434 (41)
>$45,000
301 (28)
County/state poverty Below
512 (48)
Above
547 (52)
County/US poverty Below
151 (14)
Above
908 (86)
United Network for Organ Sharing cause of death Cerebrovascular accident
513 (48)
Head trauma
284 (27)
Anoxia
220 (21)
Other
40 (4)
Manner of death Medical
661 (62)
Trauma
389 (37)
Registry status Yes
163 (15)
No
767 (73)
Unknown status
129 (12)
Overall authorization Yes
501 (47)
No
558 (53)
of potential donors falling below $45,000. The average poverty rate of potential donors was 19.5% 6.3, compared with the overall state poverty rate of 17.8% 1.5 and the US poverty rate of 14.6% 1.2. The majority of potential donors resided in counties with poverty rates higher than the US rates, whereas a considerable number lived in Appalachia. Only 15% of potential donors were on the donor registry. Familial authorization was obtained in 47% of the study population. A univariable analysis was performed between decedents for which familial authorization was obtained (“authorization” group) and compared with those where familial authorization was declined (“decline” group). Table 2 represents decedent characteristics. There were no significant differences between gender and race. However, significantly higher authorization rates were observed in decedents aged between 18-39 y (P < 0.001), those who were on the donor registry (P < 0.001), and had a head injury as a cause of death (P < 0.001) or traumatic manner of death (P < 0.001). In addition, significantly higher authorization rates were observed in decedents with the following socioeconomic factors: single (P ¼ 0.001), college or higher levels of education (P < 0.001), residents of non-Appalachian geographic areas (P ¼ 0.001), hospital located in non-Appalachian regions (P ¼ 0.02), family median household incomes >$45,000 (P < 0.001), and residents of counties with poverty rates lower than the state (P < 0.001) and national (P < 0.001) poverty rates. Table 3 represents characteristics of family members approached for donation and shows no differences in familial authorization based on gender of the family member approached for donation nor in the distance traveled from the potential donor residence to the hospital of death. Approaching the parent of a potential donor was associated with significantly higher rates of authorization compared with approaching a spouse, sibling, or child (P < 0.001). Significantly higher rates of authorization were seen when the approach for donation was decoupled from the declaration of death (P < 0.001) and when the family understood the hopelessness of the situation of their loved one (P ¼ 0.002). Higher rates of authorization were also seen in potential donors when the family initially inquired about the possibility of donation (P < 0.001). A multivariable logistical regression was performed to identify factors independently associated with increased conversion rates (Table 4). Again, younger donors (18-39 y: odds ratio [OR] ¼ 4.9, confidence interval [CI] ¼ 3.1-7.7, P < 0.001; 40-60 y: OR ¼ 2.1, CI ¼ 1.4-3.1, P < 0.001), as well as candidates with higher levels of education (college: OR ¼ 2.5, CI ¼ 1.3-4.6, P ¼ 0.005; graduate studies: OR ¼ 3.9, CI ¼ 1.6-9.4, P ¼ 0.002), had significantly higher rates of authorization. Prior listing on the donor registry was also associated with higher rates of authorization (OR ¼ 10.3, CI ¼ 6.0-17.7, P < 0.001), as did residing in counties with overall poverty rates lower than the US rates (OR ¼ 1.7, CI ¼ 1.1-2.6, P ¼ 0.02). Decoupling the approach for donation from the declaration of death was associated with higher rates of authorization (OR ¼ 3.1, CI ¼ 1.8-5.4, P < 0.001). Donation first mentioned by a local health care provider was also associated with higher authorization rates compared with the first approach by the OPO
91
shah et al socioeconomic predictors of donation
Table 2 e Univariable analysis of potential donor characteristics between decline and authorization groups. Donor factors impacting authorization
Decline Authorization (n ¼ 558), (n ¼ 501), n (%) n (%)
Table 2 e (continued ) Donor factors impacting authorization P Other
Male
325 (53)
288 (47)
Female
233 (52)
213 (48)
493 (52)
453 (48)
56 (60)
38 (40)
18-39
102 (33)
203 (67)
40-60
272 (54)
230 (46)
>60
184 (73)
68 (27)
8th grade
64 (71)
26 (29)
9-12th grade
49 (54)
41 (46)
235 (55)
195 (45)
College
79 (36)
140 (64)
Grad school
18 (40)
27 (60)
Married
281 (55)
227 (45)
Divorced
120 (55)
99 (45)
Single
105 (42)
145 (58)
51 (63)
30 (37)
Yes
262 (59)
183 (41)
No
295 (48)
318 (52)
0.8
Race
African-American
Medical
399 (60)
262 (40)
Trauma
156 (40)
233 (60)
<0.001
Discussion
<0.001
Marital status 0.001
Appalachia residence 0.001
Registry status Yes
19 (12)
144 (88)
No
456 (60)
311 (40)
<$35,000
194 (60)
129 (40)
$35,000-$45,000
235 (54)
199 (46)
>$45,000
128 (42)
173 (58)
Below
239 (47)
273 (53)
Above
319 (58)
228 (42)
<0.001
<0.001
<0.001
Education
Widowed
13 (32)
(OR ¼ 1.8, CI ¼ 1.1-2.9, P ¼ 0.01). And, although it did not reach statistical significance, there was a trend toward head trauma (P ¼ 0.06) and potential donor residence outside of Appalachia (P ¼ 0.07) being associated with higher rates of authorization. To eliminate a potential selection bias of registered donors confounding our results, a subgroup analysis of 767 potential donors was performed analyzing only potential donors who were not on the donor registry. The univariable analysis is not included in this summary, but a multivariable analysis was performed using only factors from the univariable analysis that achieved statistical significance. Table 5 illustrates the multivariable analysis for factors associated with increased rates of authorization in potential donors not on the registry. Again, younger donors (18-39 y: OR ¼ 5.4, CI 3.3-8.6, P < 0.001; 40-60 y: OR ¼ 2.2, CI ¼ 1.4-3.3, P < 0.001), as well as candidates with higher levels of education (college: OR ¼ 2.9, CI ¼ 1.4-5.9, P ¼ 0.003; graduate studies: OR ¼ 3.4, CI ¼ 1.2-9.4, P ¼ 0.02), had significantly higher rates of authorization. Decoupling was again associated with higher rates of authorization (OR ¼ 3.2, CI ¼ 1.7-5.9, P < 0.001). Donation first mentioned by a local health care provider was again associated with higher authorization rates compared with that mentioned by the OPO (OR ¼ 1.8, CI ¼ 1.1-3.1, P ¼ 0.02). Although not quite reaching statistical significance, there was a trend observed toward higher authorization rates in potential donors residing in counties with poverty rates lower than the US rates.
0.3
Age category
High school grad/General Educational Development
27 (68)
P
Manner of death
Gender
Caucasian
Decline Authorization (n ¼ 558), (n ¼ 501), n (%) n (%)
Income category
County/state poverty <0.001
County/US poverty Below
59 (39)
92 (61)
Above
499 (55)
409 (45)
Yes
151 (59)
105 (41)
No
407 (51)
396 (49)
Cerebrovascular accident
305 (60)
208 (40)
Head trauma
106 (37)
178 (63)
Anoxia
120 (55)
100 (45)
<0.001
After controlling for known factors associated with familial authorization, we were able to demonstrate several new socioeconomic factors not previously recognized as being associated with familial authorization. Multiple factors clearly contribute to disparities in donation and authorization rates. To begin to identify at-risk groups for lower conversion rates, a two-phase approach examining prehospital and in-hospital assessments as separate entities is prudent.
Hospital region in Appalachia 0.02
United Network for Organ Sharing cause of death <0.001
(continued)
Prehospital assessment It is well established that deceased donation rates are improved when the potential donor has made a previous decision to donate prehumously.14,15 Prior designation of donor registration likely leads to easing the burden of the decisionmaking process for donation of the potential donor’s family.3 We observed that prior donor designation was associated with an 11-fold increased odds of authorization (P < 0.001).
92
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
Table 3 e Univariable analysis of family member characteristics between decline and authorization groups. Factors involved with family decision making
Decline Authorization (n ¼ 558), (n ¼ 501), n (%) n (%)
Table 4 e Multivariable analysis of predictors for increased rates of authorization. Predictors P
373 (57)
283 (43)
Local health care provider
85 (54)
73 (46)
Family
61 (31)
137 (69)
185 (54)
161 (46)
Female
304 (51)
288 (49)
Spouse
236 (56)
188 (44)
Parent
72 (33)
147 (67)
Sibling
42 (58)
30 (42)
<0.001
139 (62)
84 (38)
0.5
<0.001
467 (51)
455 (49)
No
24 (63)
14 (37)
Yes
409 (49)
427 (51)
No
72 (67)
36 (33)
<25 miles
254 (52)
237 (48)
>25 miles
303 (54)
263 (46)
1.08
0.52-2.24
0.8
High school grad/General Educational Development
1.42
0.80-2.52
0.2
College
2.46
1.31-4.60
0.005
Graduate studies
3.91
1.62-9.44
0.002
Donation first mentioned by
Local health care provider
1.80
1.13-2.86
0.01
Family
2.55
1.71-3.80
<0.001
1.33
0.97-1.82
0.07
Anoxia
0.91
0.62-1.34
0.6
Head trauma
1.43
0.98-2.08
0.06
10.28
5.97-17.69
<0.001
3.11
1.80-5.36
<0.001
1.70
1.11-2.62
0.02
40-60
2.11
1.44-3.10
<0.001
18-39
4.88
3.08-7.74
<0.001
Residence in Appalachia Yes (referent) No
Family understanding of hopelessness Yes
9-12th grade
OPO (referent)
Family member approached
Children
P value
<8th grade (referent)
Gender of family member approached Male
95% CI
Education
Donation first mentioned by OPO
Odds ratio
0.002
United Network for Organ Sharing cause of death CVA/stroke (referent)
Decoupled <0.001
Registry status
Miles from home
No (referent) 0.8
Yes Decoupling No (referent) Yes County/US poverty
However, there are a multitude of other potential factors that, once identified and targeted, can potentially increase donor authorization rates. We also demonstrate that increasing education levels are significantly associated with increased rates of authorization for organ donation. This was significant in potential donors with college or higher levels of education. It has been suggested that individuals with higher education levels have more positive perceptions of organ donation.14 Our data demonstrate a stepwise decrease in familial authorization with increasing decedent age. The highest frequency of familial authorization was observed in the youngest decedent group, 18-39 y of age. Our data also demonstrate that parental approach was significantly associated with increased authorization rates, which may account for the increased familial authorization in the youngest decedents. This particular observation is likely multifactorial. Although parental approach has been examined in the pediatric donor population to some degree, factors that influence parental consent warrant additional study.16 In potential donors older than 60 y, there is likely to be more spousal, child, or sibling involvement who potentially have their own preconceived notions regarding donation and the fact that their loved one is simply “too old” to be a suitable donor.3 To further identify at-risk populations for poor authorization rates, we examined poverty levels in potential donors. We looked at poverty rates in counties where the potential donor
Above (referent) Below Age (y) >60 (referent)
CVA ¼ cerebrovascular accident.
resided in and compared that with the US and state poverty rates. Potential donors who resided in counties with higher poverty rates than the US rates were significantly less likely to donate. It is not surprising that counties with higher poverty levels potentially correlate with fewer years of formal education and fewer jobs with higher financial benefit. However, even when controlling for education and other socioeconomic factors, county poverty rate higher than the US rate remained independently associated with poor authorization rates. Donation rates were also examined in potential donors who resided in Appalachia. Appalachia is known as an area with high poverty rates, high unemployment, and health care disparities. Although residence in Appalachia was strongly correlated with poor authorization rates in a univariable analysis, this did not reach statistical significance in the multivariable analysis. This suggests that residence in
shah et al socioeconomic predictors of donation
Table 5 e Multivariable analysis of predictors for increased rates of authorization in donors not registered for donation. Predictors
Odds ratio
95% CI
P value
Education <8th grade (referent) 9-12th grade
1.18
0.53-2.65
0.6
High school grad/General Educational Development
1.62
0.84-3.11
0.1
College
2.90
1.44-5.85
0.003
Graduate studies
3.35
1.20-9.36
0.02
Local health care provider
1.83
1.10-3.05
0.02
Family
2.96
1.89-4.63
<0.001
3.16
1.69-5.91
<0.001
1.56
0.98-2.48
0.06
40-60
2.15
1.40-3.30
<0.001
18-39
5.35
3.32-8.60
<0.001
Donation first mentioned by OPO (referent)
Decoupling No (referent) Yes County poverty rates Higher than US rates (referent) Lower than US rates Age (y) >60 (referent)
Appalachia is a surrogate marker for poverty, which is associated with poor authorization rates. Our data suggest that targeting economically disparate areas may be beneficial in increasing donor registry.
In-hospital assessment Numerous studies have examined factors impacting authorization rates when potential donors are in the hospital. The OPO techniques for approach, involvement of an OPO family support representative, cause and manner of death, and donor hospital education programs have been examined by others.3 Similar to other OPOs, KODA uses a structured approach when contacted by the local health care team about a potential donor. This approach includes (1) assessment of the family; (2) assessment with the local health care team regarding family visitation with their loved one, timing of the approach, privacy for discussions, and introduction of KODA staff to the donor family; (3) assessing family understanding of the condition of their loved one; (4) providing education and allowing for questions regarding organ donation; (5) allowing family time to make a decision and providing informed consent for organ donation. Similar to other studies, our study also demonstrated that cerebrovascular accidents and anoxia as the causes of death were significantly associated with poor authorization rates.3 This may be explained in that these causes of death are typically seen in older individuals, which is an independent negative predictor for donation. This suggests that medical reasons
93
for death are surrogate markers for increased age, which is an important independent negative risk factor for donation. Decoupling is a strategy employed by OPOs as a way to approach grieving family members. Conceptually, this involves separating the declaration of death to the family from the approach for donation. Our study clearly demonstrates that when the technique of decoupling is employed, there is a significantly higher rate of authorization for donation. It has been suggested that the success of using this technique may come from comforting and letting the family members cope the process of grieving their loved one. Also interesting are the rates of familial authorization when examining who made the first approach for donation. It is widely reported that the first approach for donation by an OPO representative is strongly correlated with increased rates of authorization compared with a local health care provider’s approach.8,9 It has been suggested that this correlation is observed because an OPO representative has extensive training in handling challenging family dynamics, the occurrence of death, and providing education regarding organ donation. However, one of the most interesting findings of this study is that we observed that the first approach by the OPO was significantly associated with lower authorization rates compared with an approach by the local health care provider. We suspect multifactorial causes for this observation. Our population is largely rural with higher rates of poverty and fewer years of formal education compared with the US averages. These factors influence cultural attitudes and biases in our population toward organ donation that could explain how these families may trust and rely on their own health care provider and less on the OPO representative, who may be perceived as an “outsider.”17 Excluding potential donors who made a predeath donor designation eliminated potential selection bias of those who had actively made a decision to donate. However, the same variables remained associated with familial authorization in this sensitivity analysis. Including this subgroup analysis helped to confirm the validity of our analysis of the entire data set. There are limitations in this study. This was a retrospective case-control study using historical data from an approximate 6-y period. Studies have documented that African-Americans and other minorities have significantly lower rates of authorization for donation compared with Caucasians.12,14 Although there was a trend toward similar findings in our study, our population is homogenous in its ethnic breakdown. We suspect that this finding did not reach statistical significance because of lower numbers of African-Americans in our study population. An additional limitation is that actual donor household income was not collected by the OPO. Thus, median county income was collected by county of residence and is limited by data collection practices used by the US Census Bureau. A number of studies have used this technique to assess socioeconomic status.12,18,19 We acknowledge that county populations are not necessarily homogenous. Over a large sample size, however, results may be somewhat more generalizable for meaningful conclusions. No effort was made to identify potential donor religious affiliations, which may contribute to specific attitudes and beliefs toward organ donation. There were not any data collected in follow-up with potential donor families regarding their thoughts about why they chose to donate or not.
94
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 8 8 e9 4
In summary, our study evaluated an underserved population with unique characteristics to identify factors associated with nonauthorization. These data may be used to identify populations at risk for nondonation, a population where the OPO may choose to allocate increased resources during the family approach. Importantly, donor registration correlated most strongly with higher authorization rates. These results suggest that focused public educational efforts regarding organ donation in populations with unfavorable socioeconomic considerations, such as populations with high poverty levels or fewer years of formal education, may be beneficial in improving donor registration. However, it is noted that great care must be taken to prevent increasing mistrust with the health care system, donation, and transplantation and that very focused and population-specific educational materials be provided.14,20 Additional studies are warranted on methods on how best to accomplish this. Focused educational efforts in areas and groups where older individuals may reside and engage in daily activities may also be beneficial in improving donor registration. It has indeed been shown that targeted efforts through various media campaigns and culturally sensitive education programs can make a positive impact on donor registration.21 In addition, in-hospital assessments to include techniques of approaching families for authorization (i.e., decoupling) can also play a significant role in increasing authorization for donation. Education of and collaboration with local health care providers may also be a very effective method of improving authorization rates.
Acknowledgment The authors acknowledge the Kentucky Organ Donor Affiliates for providing the database of potential organ donor referrals. They also acknowledge the Kentucky Office of Vital Statistics and West Virginia Health Statistics Center for providing the database of death certificates. Authors’ contributions: M.S., A.G., J.B., R.G., and D.D. contributed to conception and design of the study. M.S., V.V., and A.G. contributed to data acquisition. M.S., V.V., A.G., M.D., J.B., R.G., and D.D. contributed to analysis and interpretations. M.S., V.V., A.G., and D.D. drafted the manuscript. M.S., V.V., A.G., M.D., J.B., R.G., and D.D. critically revised the manuscript. M.S., A.G., M.D., J.B., R.G., and D.D. contributed to statistical analysis and interpretation.
Disclosure The authors reported no proprietary or commercial interest in any product mentioned or concept discussed in the article.
references
1. Number of patients Waiting for organ transplant. Richmond, Virginia: United Network for Organ Sharing; 2016.
2. Donors Recovered in the U.S. Richmond, Virginia: OPTN: US Department of Health and Human Services; 2014. 3. Brown CV, Foulkrod KH, Dworaczyk S, et al. Barriers to obtaining family consent for potential organ donors. J Trauma. 2010;68:447e451. 4. Salim A, Velmahos GC, Brown C, Belzberg H, Demetriades D. Aggressive organ donor management significantly increases the number of organs available for transplantation. J Trauma. 2005;58:991e994. 5. Salim A, Martin M, Brown C, et al. Using thyroid hormone in brain-dead donors to maximize the number of organs available for transplantation. Clin Transplant. 2007;21:405e409. 6. Chon WJ, Josephson MA, Gordon EJ, et al. When the living and the deceased cannot agree on organ donation: a survey of US organ procurement organizations (OPOs). Am J Transplant. 2014;14:172e177. 7. UNOS Star Files. OPTN database. Richmond, Virginia: OPTN: US Department of Health and Human Services; 2014. 8. Ebadat A, Brown CV, Ali S, et al. Improving organ donation rates by modifying the family approach process. J Trauma Acute Care Surg. 2014;76:1473e1475. 9. Dubay DA, Redden DT, Haque A, et al. Do trained specialists solicit familial authorization at equal frequency, regardless of deceased donor characteristics? Prog Transplant. 2013;23:290e296. 10. Siminoff LA, Burant CJ, Ibrahim SA. Racial disparities in preferences and perceptions regarding organ donation. J Gen Intern Med. 2006;21:995e1000. 11. Salim A, Berry C, Ley EJ, et al. The impact of race on organ donation rates in Southern California. J Am Coll Surg. 2010;211:596e600. 12. DuBay D, Redden D, Haque A, et al. Is decedent race an independent predictor of organ donor consent or merely a surrogate marker of socioeconomic status? Transplantation. 2012;94:873e878. 13. Gortmaker SL, Beasley CL, Sheehy E, et al. Improving the request process to increase family consent for organ donation. J Transpl Coord. 1998;8:210e217. 14. Regalia K, Zheng P, Sillau S, et al. Demographic factors affect willingness to register as an organ donor more than a personal relationship with a transplant candidate. Dig Dis Sci. 2014;59:1386e1391. 15. National donor designation Report card. Donate Life America; 2014. Available at: https://www.donatelife.net/wp-content/ uploads/2016/06/Report-Card-2014-44222-Final.pdf. 16. Rodrigue JR, Cornell DL, Howard RJ. Pediatric organ donation: what factors most influence parents’ donation decisions? Pediatr Crit Care Med. 2008;9:180e185. 17. Traino HM, Alolod GP, Shafer T, Siminoff LA. Interim results of a national test of the rapid assessment of hospital procurement barriers in donation (RAPiD). Am J Transplant. 2012;12:3094e3103. 18. Adler JT, Hyder JA, Elias N, et al. Socioeconomic status and ethnicity of deceased donor kidney recipients compared to their donors. Am J Transplant. 2015;15:1061e1067. 19. Mohan S, Mutell R, Patzer RE, Holt J, Cohen D, McClellan W. Kidney transplantation and the intensity of poverty in the contiguous United States. Transplantation. 2014;98: 640e645. 20. Reibel M, Olmo C, Andrada S, Koertzen J. Deep demographics: understanding local Variation in donor registration. Prog Transplant. 2016;26:191e198. 21. Salim A, Ley EJ, Berry C, et al. Effect of community educational interventions on rate of organ donation among Hispanic Americans. JAMA Surg. 2014;149:899e902.