Abstract
Increasing public commitment to organ donation is critical to improving donor kidney availability for end-stage renal disease patients desiring transplant. This study surveyed (N = 1339) African Americans, measuring perceived pros relative to cons of organ donation, to evaluate an existing Transtheoretical Model decisional balance scale and associations between decisional balance and expressing donation intentions. Findings supported the existing scale structure. More positive decisional balance ratios were associated with 1.76 times the odds of expressing intentions (95% confidence interval = 1.52–2.04). Pros were more strongly linked to donation intentions than cons. Greater understanding of organ donation decision-making is valuable for informing interventions that encourage donation.
Introduction
Kidney transplantation is the therapy of choice and the recommended treatment over dialysis for many end-stage renal disease (ESRD) patients in the United States. Support for transplant is guided by evidence that individuals who undergo transplant experience improved quality of life, greater participation in life activities, longer survival, and lower incurred medical expenses (Evans et al., 1985; Patzer et al., 2014; Patzer and Pastan, 2014; Purnell et al., 2013; Southeastern Kidney Council, 2014; Tonelli et al., 2011). Despite wide support for kidney transplantation, the unadjusted kidney transplant rate for dialysis patients in the United States staggers below 4 percent (United States Renal Data System, 2015).
Insufficient donor organ supply is the primary reason why kidney transplantation rates are low in the United States. Significant growth in the prevalence of ESRD, longer survival of those with ESRD, and related upsurges in the number of patients on the kidney transplant waiting list have been well documented in recent decades to describe the growing need for donor kidneys (United States Renal Data System, 2015). This demand far exceeds that for other organs, with more than 82 percent of the 119,454 Americans on the organ transplant waiting list requiring kidneys (US Organ Procurement and Transplantation Network and Scientific Registry of Transplant Recipients, 2016). Unfortunately, the rising need for donor kidneys has not been matched by comparable increases in donation. In 2013, the active kidney waiting list was 2.7 times greater than the supply of deceased donor organs (United States Renal Data System, 2015).
Substantial efforts have been made to increase the availability of donor organs by improving public commitment to deceased organ donation. These initiatives have included programs that educate and empower community members about becoming organ donors (Arriola et al., 2010; Callender and Miles, 2001), interventions that increase opportunities for expressing donation intentions (Arriola et al., 2010; Callender and Miles, 2001), and policies establishing state donor registries that better document residents’ intentions to donate (Callender et al., 2001; The National Conference of Commissioners on Uniform State Laws, 2017). Modest increases in the number of deceased kidney donors have been observed since 2003 and may be attributed to the success of the various multi-level strategies implemented to improve public commitment to organ donation (United States Renal Data System, 2015). Still, reported rates of organ donation and intentions to donate remain far below what is required to accommodate the burgeoning demand for kidney transplant, indicating the need for further research addressing the factors influencing organ donation decision-making (United States Renal Data System, 2015).
Numerous initiatives to improve public commitment to organ donation have focused on African Americans, given that African American organ donation rates are low relative to their disproportionately higher rates of ESRD and overrepresentation on the kidney waiting list (Kurz et al., 2007). African Americans make up 13.3 percent of the US population and 16 percent of deceased organ donors but represent 31 percent of ESRD patients and 34 percent of the kidney waitlist (United States Renal Data System, 2015; US Department of Health and Human Services, 2014). Improving African Americans’ donation rates may benefit the organ donation system as a whole by increasing organ availability and by providing a more heterogeneous donor pool (Committee on Organ Procurement and Transplantation Policy, 1999). Improving donation rates may benefit African American ESRD populations in particular, by increasing the likelihood that minorities can successfully match with a donor organ once on the kidney waiting list.
Factors associated with African Americans’ decision-making regarding becoming an organ donor have been well studied. Findings demonstrate that negative attitudes and beliefs about organ donation, including fear of pre-mature declaration of death, distrust of the medical system, and concerns of racial inequalities in the organ allocation system, are linked to reluctance to becoming a donor (Boulware et al., 2002a; Callender et al., 1982; DuBay et al., 2014; Kurz et al., 2007; Morgan, 2004, 2006; Park, 1998; Russell et al., 2012). Research also shows that more positive donation perceptions, such as stronger belief in the effectiveness of transplant, greater awareness of the need for transplant, and knowing someone who has received a transplant are associated with increased willingness to donate (Andrews et al., 2016; Arriola et al., 2008; Britt et al., 2017; Callender et al., 1991; Creecy and Wright, 1990; Davis et al., 2005; Morgan, 2004; Resnicow et al., 2012; Siminoff and Sturm, 2000). The literature examining factors associated with African Americans’ donation intentions is compelling; however, research overwhelmingly analyzes separately the negative factors that impede donation and the positive perceptions that facilitate donation (Davis et al., 2005; DuBay et al., 2014; Morgan, 2006; Morgan and Miller, 2002). A more integrated approach that examines the ways in which the perceived positive aspects of organ donation are weighed against the negative aspects may yield more insightful explanations of African Americans’ organ donation decision-making processes.
The Transtheoretical Model’s (TTM) decisional balance construct provides a useful framework for concurrently examining positive and negative perceptions guiding organ donation decision-making by rating individuals’ perceived pros relative to their cons of performing pro-donation behaviors (DiClemente and Prochaska, 1982; Hall et al., 2007; Waterman et al., 2006). In application, this construct suggests that individuals who perceive greater pros than cons of organ donation will be more likely to become donors.
Existing literature supports the use of the TTM decisional balance construct for assessing health decision-making among African American adults. Studies that have successfully measured African Americans’ decisional balance for donation-related behaviors include the work of Burditt et al. (2009) exploring motivations for blood donation and the research by Waterman et al. (2010) examining intentions to receive a deceased kidney transplant. Studies that have developed and tested measures for assessing decisional balance regarding organ donation support the use of the TTM construct but have primarily included predominantly White study samples (Robbins et al., 1999, 2001).
Few studies have examined decisional balance among African Americans when determining whether to become organ donors. One notable exception is the research conducted by Hall et al. (2007) with students from four historically Black colleges in North Carolina. This study adapted and refined scales originally developed for a mostly White study sample to create valid and reliable measures that are culturally appropriate for assessing the decisional balance constructs with African American college students (Hall et al., 2007). This research has significantly contributed to the field of organ donation through the development of a 14-item scale validated for measuring the pros and cons of becoming an organ donor among African American college students. However, further research is needed to test the measure among African American adults.
The purpose of this article is to build upon existing literature examining African American organ donation decision-making by (1) testing the construct validity of an existing psychosocial scale developed to measure TTM decisional balance for becoming an organ donor and (2) examining the extent to which the relative weighing of pros and cons is associated with self-reported intentions to donate, among African American adults. It is expected that further research of the interrelationships among these constructs will inform intervention messages that are effective in increasing African Americans’ commitment to organ donation.
Methods
Study design and sample
Giving About Choices in Transplantation and Sharing (ACTS) is a population-based study designed to assess whether culturally sensitive educational materials are associated with increased donation-related knowledge and improved attitudes and beliefs toward organ and tissue donation and increased registration on the state donor registry among residents of 16 ZIP codes in the metropolitan area of a large southeastern city. The state donor registry and 2010 Census data were used to identify ZIP codes with the highest percentages of African American residents and the lowest percentages of registered donors. Data for this article were taken from baseline assessments completed by 1339 participants enrolled in the study. Eligible study participants self-identified as Black or African American, were 18 years of age or older and resided in one of the 16 study ZIP codes. Written informed consent was obtained from all individuals prior to study participation. This study was approved by the Emory University Institutional Review Board (IRB0059938).
Questionnaire content
Study participants completed a questionnaire that measured decisional balance related to organ donation, self-reported donation intentions, and demographic characteristics (e.g. age, gender, educational attainment, employment, and marital status). Questionnaire scales are described below.
Organ donation decisional balance
This survey item was assessed using the survey prompt, “Please share your feelings about becoming an organ donor and select an answer based on its importance to your decision of whether to become a donor.” A total of 14 items were used to assess participants’ decisional balance related to organ donation (shown in Table 1). These items were previously developed and validated by Hall et al. (2007) to assess decisional balance regarding deceased organ and tissue donation among African American college students. Survey items ask about seven possible pros and seven possible cons of organ donation. Scaled response options for each item were (1) Not at All Important, (2) Somewhat Important, (3) Neutral, (4) Important, and (5) Very Important. Pros score and cons score summary variables were created by summing all responses for the seven corresponding decisional balance survey items for each of the two constructs, with possible summary scores ranging from 7 to 35. Next, decisional balance scores were created by dividing participants’ summed pros score by their summed cons score, with possible ratio scores ranging from 0.2 to 5.0. Ratio scores higher than 1 indicate more positive decisional balance.
Donation decisional balance survey items and CFA loadings.
CFA: confirmatory factor analysis.
Self-reported expressing donation intentions
Participants responded to one survey item, “I have expressed my donation intentions by the following.” Response options were “Signing a donor card or registering online,” “Designating myself on my driver’s license,” “Sharing my wishes with a loved one/family member,” “None of the above.” These methods of expressing donation intentions have been strongly linked to success in becoming an organ donor in previous studies (Robbins et al., 2002; Rodrigue et al., 2006; Siminoff et al., 2001). A dichotomous variable was created to assess expressing donation intentions for data analysis. Responses indicating any method of expressing donation intentions using any of the listed methods were coded as “1,” and responses that did not indicate expressing donation intentions were coded as “0.”
Data analysis
All data analyses were conducted using SPSS 24.0 and SPSS Amos 25.0, with p-values less than 0.05 indicating statistical significance. Researchers began by analyzing the frequency distributions of all categorical variables. Normality was assessed for all continuous variables using skewness and kurtosis tests.
Confounding was evaluated for all factors by conducting chi-square tests and t-tests, as appropriate, to assess associations of demographic variables (age, gender, education, employment, and marital status) with decisional balance scores and with the study outcome (self-reported expressing donation intentions).
Confirmatory factor analysis (CFA) was conducted using SPSS Amos 25.0 to address the first study aim testing the underlying factor structure of the decisional balance measure developed for examining organ donation decision-making with our study sample. Multiple models were tested and compared to identify the model that best represented the factor structure. For each of the models, the following three fit indices were examined: (1) the confirmatory fit index (CFI), (2) the root mean square error of approximation (RMSEA), and (3) the standardized root mean square residual (SMSR). Traditionally, CFI values above 0.80 indicate good fit, while values above 0.90 suggest excellent fit (Tabachnick and Fidell, 2001). For RMSEA and SMSR, values below 0.08 generally indicate good model fit (Hu and Bentler, 1998; MacCallum et al., 1996).
Logistic regression was conducted, addressing the second study aim, to examine the extent to which the decisional balance of pros relative to cons of organ donation was associated with expressing donation intentions. Generalized estimating equations (GEEs) were used to adjust for clustering at the community worker level. In an initial exploratory step, a model was developed that included patients’ summary pros scores and summary cons scores separately, controlling for age, gender, highest level of education, employee status, and marital status. Subsequently, a second logistic model was created that included the relative decisional balance score and controlled for age, gender, highest level of education, employment status, and marital status.
Results
Demographics for the 1339 participants included in data analysis are presented in Table 2. Study participants were mostly female (59.0%) and not married (78.2%). The mean age of participants was 48.59, with age normally distributed across the sample. The study sample was largely educated for 12 or more years, with only 7.8 percent completing less than high school. Around 40 percent of participants were currently employed. Participants had a mean pros score of 26.74 and a mean cons score of 16.75 contributing to a positive mean relative decisional balance score of 10.07. Pros scores, cons scores, and decisional balance scores were normally distributed. Unadjusted mean decisional balance scores were highest among participants who were women, more educated, employed part- or full-time, and married. Of the participants, n = 658 (49.1%) reported having indicated their donation intentions verbally or in written form, while n = 681 (50.1%) did not report having previously indicated their donation intentions.
Demographic characteristics of study participants (N = 1339).
Percentages do not sum to zero due to missing responses.
The study’s first aim of examining the underlying factor structure of an existing psychosocial scale developed to measure TTM decisional balance for becoming an organ donor was assessed using CFA methods. Around 14 decisional balance items were included in CFA analysis. A total of 1178 of the 1339 study participants were included in CFA analysis, with excluded participants resulting from one or more missing variables for decisional balance scale items. Sensitivity analysis was conducted and confirmed that excluded participants were otherwise similar to included participants in demographics and study outcomes.
Four models were tested and compared in CFA analysis to identify the model that best represented the factor structure: (1) a null (comparative) model with no latent factors, (2) a one-factor model, (3) a two-factor model with uncorrelated pros and cons, (4) a two-factor model with correlated pros and cons (Hall et al., 2007). The fourth model, containing correlated pros and cons demonstrated the best fit, with CFI = 0.905, RMSEA = 0.070, and SRMR = 0.054. The CFA demonstrated factor loadings ranging from 0.59 to 0.76 for pros and 0.39 to 0.75 for cons. The correlation coefficient between pros and cons was −0.11. Cronbach’s alphas of 0.83 for pros and 0.79 for cons were determined and support the internal reliability of both scales. The confirmatory factor loadings and Cronbach’s alphas for the selected model are presented in Table 1.
The second research aim examined the extent to which the balance of perceived pros relative to the perceived cons of organ donation was associated with expressing donation intentions. The exploratory analyses separately examined the associations of the perceived pros and cons of organ donation with expressing donation intentions. Logistic regression using GEE demonstrated that each one-unit increase in the pros of donation increased the odds of expressing donation intentions by 1.13 times (confidence interval (CI) = 1.09–1.15). Conversely, each one-unit increase in the cons of donation reduced the odds of expressing donation intentions by 0.96 times (CI = 0.94–0.98; see Table 3).
Multivariable logistic regression analysis for decisional balance associated with expressing donation intentions.
CI: confidence interval.
All models control for age, gender, marital status, employment, and educational status.
p < 0.001, **p < 0.0001.
Subsequent analysis examined the relationship between the decisional balance ratio (pros to cons) and expressing donation intentions. Findings from logistic regression with GEE demonstrated an association between decisional balance and expressing donation intentions. For every one-unit increase in decisional balance, the odds of participants expressing donation intentions increased by 75 percent, after controlling for age, gender, education, employment, and marital status (95% CI = 1.52–2.04; see Table 3).
Discussion
Understanding donation decision-making behaviors among African Americans is imperative to increasing commitment to donation. This study sought to test the decisional balance constructs (pros and cons) as they pertain to becoming an organ donor and explore associations between decisional balance and expressing donation intentions among a sample of African American adults. The first aim evaluated the underlying factor structure measuring the pros and cons of becoming an organ donor, as theorized and validated among African American college students (Hall et al., 2007) and its applicability among African American adults.
CFA indicated that when tested with this study’s population, the previously developed 14-item decisional balance scale maintains a similar two-factor structure, is internally consistent, and demonstrates excellent model fit. These findings support the examined decisional balance scale as a valuable tool for better understanding African American adults’ perceived pros and cons of organ donation.
One notable difference in the factor structure determined by this project compared to those presented in previous studies is that the correlation coefficient between pros and cons was negative (–0.11). This observed weak, inverse relationship differs from the positive correlation relationship reported by the Hall et al (2007) project, from which this study’s items were drawn, and by other similar studies (Waterman et al., 2010, 2015). Although this study’s finding diverges from that of previous studies, its indication that perceiving more positive aspects of donation is linked to perceiving fewer negative aspects is well aligned with TTM concepts. Furthermore, the inverse relationship between pros and cons identified in the study is fitting for this project’s context, given that, on average, participants had positive decisional balance regarding organ donation.
The second aim explored the relationship of the decisional balance constructs (pros, cons, and in combination) and expressing donation intentions. As anticipated, a higher perception of pros of organ donation was associated with greater odds of expressing donation intentions, and a higher perception of cons was associated with a reduced likelihood of expressing donation intentions. This finding is demonstrative of the relationship between decisional balance and behavior change predicted by the TTM and provides further evidence for the validity of the pros and cons constructs identified in the best-fitting structural model (Prochaska et al., 1992).
When considering the relative weighing of pros and cons of organ donation, we found the pros to be more strongly linked to expressing donation intentions than cons. We posit from our analyses that interventions aiming to promote the benefits of organ donation may be more effective in increasing pro-donation behaviors than interventions aiming to discount the barriers to organ donation. This hypothesis aligns with the TTM, which supports that behavioral pros outweigh the cons as a person moves closer to action (Glanz et al., 2008). There are an abundance of initiatives aiming to improve commitment to organ donation, with many focusing on diffusing negative attitudes toward donation (Callender and Miles, 2010). Such initiatives are backed by study findings demonstrating prominent barriers to organ donation as concerns about inequalities in the organ allocation system (Arriola et al., 2008; Boulware et al., 2007), whether being recognized as an organ donor might hasten one’s death (McNamara et al., 1999; Siminoff and Sturm, 2000) and religious beliefs about needing the body to remain whole to enter heaven (Boulware et al., 2002b; Callender et al., 2002; Callender and Miles, 2001).
While the barriers to organ donation commitment are undeniable, our findings suggest that interventions with an emphasis on promoting the gains of organ donation rather than solely focusing on dispelling the barriers to donation may be most effective in increasing organ donation commitment. One way of potentially enhancing the benefits (or pros) of organ donation commitment through behavioral interventions involves promoting the altruistic aspect of this health behavior. Organ donation is a health behavior characterized by low-vested interest, high attitudinal ambivalence, and oftentimes a lack of personal benefit (Siegel and Alvaro, 2009). Altruism is not readily promoted in most organ donation interventions, partly because its relationship with donation expression is somewhat undefined, given the overall difficulty in measuring altruism (Morgan and Miller, 2002). Thus, more research is needed to understand how to promote altruism in organ donation to potentially create more effective interventions in this area.
Creating interventions that focus on promoting the gains of organ donation may be particularly important in improving commitment to organ donation among African Americans. The “cons” measures in this study explored impediments to donation around social norms, religion, and fairness in organ allocation. Prior studies have shown that many of these barriers are related to the persistence of myths and misconceptions that are deeply rooted in cultural, familial, religious, and societal beliefs among African Americans (Boulware et al., 2002a; Callender et al., 1997; Kurz et al., 2007) and are strong predictors of reluctance to donate (Kurz et al., 2007; Siminoff and Sturm, 2000). Furthermore, value-based barriers (such as our item, “organ donation is against my religious beliefs”) do not always lend themselves to objective arguments (Morgan and Miller, 2002) and are less changeable through behavioral interventions, furthering the argument for promoting the benefits of organ donation in future interventions.
There are limitations of this study that should be noted. One significant limitation is that the TTM measure’s stages of change and self-efficacy were not measured. The absence of these data precluded opportunities to explore the relationship between participants’ donation-related decisional balance and their readiness for and confidence in becoming organ donors. Such analyses could lend this project added value in helping public health practitioners to tailor intervention strategies to their target populations (Glanz et al., 2008; Waterman et al., 2010).
Additional TTM measures were not collected in this study in an effort to minimize survey participant fatigue, as it was beyond the scope of this project. This study is part of a larger evaluation of an intervention that sought to improve organ donation intentions, thus shaping the outcome used in this study. The authors aim to incorporate stages of change and self-efficacy measures in future, more comprehensive studies on this topic.
An additional study limitation is related to the response options developed for the survey item used to capture the primary outcome variable—expressed donation intentions. Participants were asked to select the methods they used to communicate their donation intentions. Two of the three response options (excluding “none of the above”), “signing a donor card or registering online” and “designating myself on my driver’s license,” indicate participants are communicating their commitment to becoming a donor. However, it is likely that some individuals who selected the third option, sharing their wishes with a loved one/family member, were actually informing their loved ones of their disinterest in becoming an organ donor. Despite this concern, results from this survey item were maintained in the data analysis, given evidence that talking with family members about organ donation is linked to increased willingness to donate and greater likelihood of becoming a donor (Rodrigue et al., 2006).
We neither assume causality in this study, as it was cross-sectional in design, nor can we assume that our sample of African Americans is representative of all African Americans. Also, we used convenience sampling and only considered participants from a small geographical area in Georgia. Given the cross-sectional design, we can also assume that participants were at different stages of change along the TTM, which could account for some of the variability in decisional balance scores. Finally, we only considered self-reports of past expression of donation intentions, which may differ from actual behaviors.
Despite the described limitations of this study, there are also substantial strengths. This research is one of few studies to analyze the pros and cons of expressing donation intentions in combination. Prior research across health behaviors has primarily focused on analyzing pros and cons of organ donation separately. As a construct of the TTM, decisional balance provides a foundation for elucidating individuals’ readiness for behavior change. Consequently, assessing the pros and cons of decisional balance together provides a more comprehensive, theory-based method for understanding organ donation decision-making that can, in turn, better inform future health promotion intervention strategies (Glanz et al., 2008).
Identifying new ways to promote organ donation among African American adults is imperative to address the growing need for available organs. Our findings support a theoretically based and relatively underused strategy for developing interventions that effectively encourage organ and tissue donation among African Americans. Greater understanding of decisional balance among African American adults in relationship to expressing donation intentions will be valuable for informing future interventions aimed to improve organ donation registration rates and increase access to life-saving transplants.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) (Grant # 5R01DK079713-10).
