Evaluating Barriers and Transtheoretical Model Constructs in Kidney Transplant Decision-Making

Living donor kidney transplant (LDKT) is an effective treatment for kidney disease but is underutilized compared to other treatment options. Understanding factors that influence LDKT decision-making has potential to enhance intervention effectiveness and increase pursuit of living transplant. The Transtheoretical Model (TTM) is one model of behavior change that has been applied to transplant decisionmaking in kidney disease populations. TTM constructs in this area have some empirical support, but evidence suggests that aspects of socioeconomic status also impact the decision to pursue LDKT. The purpose of this study was to test a cross-sectional model of readiness for pursuing LDKT that was theoretically based in the TTM. Key socioeconomic status (SES) variables were incorporated into the model of TTM change constructs in an effort to study a wider range of variables that may improve understanding of LDKT decision-making. Data were utilized from the completed baseline sample of Your Path to Transplant, a longitudinal randomized control trial that aimed to enhance decisionmaking to pursue kidney transplant by delivering TTM Stage-matched expert system coach-delivered feedback (N = 799 ESRD patients). Prior to model testing, multivariate analysis of variance (MANOVA), analysis of variance (ANOVA), and chi-square tests were performed to examine relationships within TTM constructs (Stage of Change, Decisional Balance, and Self-Efficacy for pursuing LDKT) and between TTM constructs and five socioeconomic variables. Results revealed significant relationships between TTM variables, but no significant relationships were observed between TTM and SES variables. Analyses between SES variables revealed significant relationships with small effect sizes. Stepwise binary logistic regression was performed to test two models of readiness for pursuing LDKT (Stage of Change: Pre-Action Stages or Action). The replicated TTM model demonstrated expected relationships between independent and dependent TTM constructs. Decisional Balance and Self-Efficacy were related to Stage of Change, and later Stages exhibited significantly greater Self-Efficacy and Pros for pursuing LDKT and significantly lower Cons (χ (3) = 20.83, p < .001, R = .047, 95% CI [.01, .08]). In the full model, no statistically significant relationships were observed between TTM constructs and SES variables. Findings from this study support the successful replication of TTM constructs in a large and diverse sample of ESRD patients. The replicated model demonstrated key differences in perceptions and motivations between patients who were in Action compared to Pre-Action Stages. However, this study was unable to detect significant improvement in model fit with the addition of SES variables. Future research should examine the LDKT readiness model longitudinally, and test for relationships with SES variables over time.


LIST OF TABLES
Treatment for ESRD involves dialysis or a kidney transplant from a deceased or living donor. Dialysis is the standard of care for ESRD with wide accessibility and health insurance coverage (Farney, Doares, Kaczmorski, Rogers, & Stratta, 2010;USRDS, 2016). Despite universal use, dialysis is the treatment option associated with the poorest health outcomes. Compared to kidney transplant, dialysis is associated with decreased quality of life and reduced lifespan (Mange, Joffe, & Feldman, 2001).
Chronic dialysis treatment also increases the likelihood of future transplant failure, which affects patients who receive long-term dialysis while waiting for an available kidney transplant (Wolfe et al., 1999).
Kidney transplant is a more effective treatment for ESRD, with a projected life span increase of three to 17 years compared to dialysis (USRDS, 2016;Wolfe et al., 1999). Transplant recipients report greater quality of life, including reduced pain and decreased fatigue, and have fewer lifestyle restrictions with discontinuation of dialysis (Neipp et al., 2006).

Deceased donor kidney transplant (DDKT) is the most common transplant
option, accounting for two-thirds of all kidney transplants (USRDS, 2016). However, the demand for DDKT greatly outweighs supply. The average time spent on a wait list is 3.5 years, which may exceed the life expectancy of an ESRD patient (USRDS, 2016;Schold et al., 2009). In addition to a prolonged waiting period, DDKTs take longer to achieve kidney function once transplanted and have higher failure rates than living transplants (Cecka, 1998).
The most effective treatment for ESRD is living donor kidney transplant (LDKT) where a patient receives a kidney from a matched living donor. Studies have found that LDKT recipients achieve better health outcomes than DDKT or dialysis. In the immediate postoperative period, LDKT is associated with better graft survival and earlier kidney functioning than DDKT (Cecka, 1998 The process of pursuing LDKT is more complex than DDKT because it involves finding a healthy living donor after completing the transplant referral and candidacy evaluation common to both transplants. Patients interested in LDKT must seek a potential donor from their network of family, friends and community. Potential donors undergo testing to determine match suitability, a process that may be repeated until a match is found. Because a patient's transplant is contingent upon finding a suitable donor, the level of engagement after the transplant evaluation has potential to affect whether the patient obtains an LDKT.

Transtheoretical Model of Behavior Change.
The transtheoretical model of behavior change (TTM) is an empirically supported theoretical framework that has been used to understand and guide interventions to support high quality decision-making to pursue kidney transplantation (Waterman et al., 2014). The TTM is a decision-making model of intentional health behavior change oriented towards wide-reaching public health interventions, in which malleable behavior change processes are targeted to increase readiness to enact behavior change (Prochaska & Velicer, 1997).
The key organizing principle of the TTM is the conceptualization of behavior change as a process that occurs over time, which is the temporal dimension of the model (Prochaska & Velicer, 1997 Decisional Balance is defined as the relative weighing of Pros and Cons towards making a health behavior change (Prochaska & Velicer, 1997 (Velicer et al., 2012). This crossover of valuing benefits more greatly than the costs of making a behavior change typically occurs in the Contemplation Stage (Hall and Rossi, 2008).
Situation-specific Self-Efficacy, derived from Bandura's self-efficacy theory (1977), is the second intermediate construct in the TTM. In the context of health behavior change, Self-Efficacy is defined as an individual's confidence to make or sustain a behavior change across high-risk or difficult situations (Prochaska & Velicer, 1997). Self-Efficacy is expected to increase as individuals progress through later Stages of Change (Velicer et al., 2012).

Transtheoretical Framework of Decision-Making for Pursuing LDKT
Over decades of research, the TTM has served as a framework to evaluate change mechanisms in a wide range of health behaviors across varied populations, including cadaveric organ donation and blood donation (Robbins et al., 2001;Burditt et al., 2009) (Waterman et al., 2013). In a survival analysis that investigated receipt of a living donor kidney as the main outcome, patients in later Stages of Change (Action or Maintenance) at the onset of the transplant process were significantly more likely to have received an LDKT six years later compared to patients in earlier Stages of Change (Hazard Ratio = 4.3, 95% CI = 2.7, 6.8) (Waterman et al., 2013). This finding demonstrated the importance of readiness as a framework for understanding patients' decision-making to pursue transplant, but also as a point of intervention that could increase LDKT utilization.
Refined measures of Stage of Change, Decisional Balance, and Self-Efficacy for pursuing LDKT have since been developed and validated across two samples of patients with ESRD (Waterman et al., 2015). Results from validation testing have demonstrated the expected trends across Stages of Change, as outlined above (Waterman et al., 2015).
While preliminary findings support this application of the TTM in modeling LDKT decision-making, it is unclear whether the model includes all relevant barriers previously found to impact the process of receiving a LDKT. Findings from a number of retrospective studies indicate that key aspects of socioeconomic status (SES) serve as barriers to transplant, which are particularly relevant when addressing ongoing racial disparities and inequities in LDKT receipt (Lockwood, Bidwell, Werner, & Lee, 2016). However, TTM studies do not typically test the utility of socio-demographic variables, possibly due to the expectation that concrete measures of SES influence broader constructs, such as Self-Efficacy, to have their effects on behavior change.
Despite this hypothesis, it may be that LDKT decision-making is a considerably more complex process than is the case for many health behaviors because of the combined influence of individual-level motivation, external resources, and health implications for the living donor. Moreover, it is important to study the role of specific socioeconomic barriers to not only recognize and respond to consistent findings across key retrospective studies, which suggest that SES strongly relates to low LDKT rates in minority populations, but to also evaluate whether SES improves our understanding of TTM change processes in a model of readiness for pursuing LDKT.

Inequities in LDKT Receipt
A key issue in kidney transplantation is unequal access to living and deceased donor transplants in underserved and minority populations. Non-White ESRD patients receive considerably fewer kidney transplants than White patients despite higher rates of ESRD, which is due in part to higher rates of diabetes and hypertension (OPTN 2017c (Hall, 2011;Lockwood et al., 2016;Navaneethan & Singh, 2006).

Socioeconomic Barriers to Transplant.
Socioeconomic status is defined as the relative social standing of an individual or group, often measured as a combination of education, income, and occupation (APA, 2017). In minority samples, studies have associated lower SES with a variety of circumstances that serve as barriers to kidney transplant (Gore, Danovitch, Litwin, Pham, & Singer, 2009;Navaneethan & Singh, 2006). However, it is less clear whether conditions that coincide with lower SES serve as universal barriers to transplant across race and ethnicity, or have a magnified effect in certain groups. One study found that patients of lower SES faced similar barriers towards transplant evaluation and receipt regardless of race (Sieverdes et al., 2015). Another retrospective study of 41,000 ESRD patients identified a number of universal and race-specific characteristics associated with decreased likelihood of LDKT receipt, which included older adults, African Americans, those with less education, those who lived in lower income areas, or those insured by Medicare insurance rather than private health insurance (Gore et al., 2009).
Components of SES have been found to impact every step of the transplant process. Patients with lower SES have been found to experience inadequate access to healthcare at early stages of kidney disease, difficulty completing the transplant evaluation, reduced likelihood of undergoing transplant surgery, and difficulty affording immunosuppressive medication after insurance coverage ends (Purnell et al., 2013;Waterman et al., 2013).

Poverty. Poverty influences both disease progression and course of treatment
for ESRD. Poverty is associated with higher risk for hypertension and diabetes, two conditions that damage kidney function (Crews, Pfaff, & Powe, 2013). Moreover, poverty is a predictor for the development of chronic kidney disease, with greater influence in Black patients than White (Crews, Charles, Evans, Zonderman & Powe, 2010).
Poverty has been found to affect many aspects of the transplant process, including reduced likelihood of referral for evaluation and completing the evaluation process . Other indications of SES, such as education level, employment, and insurance type, have also been found to affect likelihood of transplant receipt. Full-time employment is one factor associated with increased likelihood of LDKT receipt and improved graft survival, even when controlling for health insurance (Petersen et al., 2008;Sandhu et al., 2013). In this area of research, full-time employment has been studied as a function of mental and physical health status, education level, or financial resources (Sandhu et al., 2013).
Education. Education level has broad implications for future health outcomes due to moderating effects on other facets of SES, such as employment, income, and living conditions. Retrospective studies have identified poorer ESRD outcomes in those with low levels of education, including increased risk for conditions that could influence ESRD onset or complicate treatment, such as diabetes and coronary heart disease (Green & Cavanaugh, 2015).
In the ESRD population, lower educational levels have been associated with Health Insurance Type. Research suggests that health insurance type may partially explain minority disparities in transplantation (Johansen, Zhang, Huang, Patzer, & Kutner, 2012;Schold et al., 2011). Lack of private health insurance is associated with reduced likelihood of referral for evaluation and completing the evaluation process if referred (Schold et al., 2011). Findings by Patzer et al. (2012) suggest an increased likelihood of transplant receipt with private insurance, as 43.8% of their sample had private insurance at transplant referral and 74% had private insurance at transplant.
Insurance type influences the treatment options available to a person with ESRD. Dageforde et al. (2015) and Kazley et al. (2014) found that lack of coverage for immunosuppressive medication deterred patients from pursuing transplant. For example, Medicare provides coverage for dialysis, but only covers the first three years of immunosuppressive medications post-transplant (Farney et al., 2010). Without insurance coverage for expensive medications, a patient must decide whether they can afford anti-rejection drugs for the rest of their life.

Purpose of Study.
This study evaluated whether the integration of key socioeconomic barriers into an established TTM model enhanced our understanding of decision-making for pursuing LDKT in a diverse sample of kidney patients who were at different Stages of Change. Typically, behavior change studies using the TTM have not incorporated socio-demographic variables into main analyses, but likely considered SES to understand complexities of a behavior, identify important sample characteristics, or develop construct measures that included relevant barriers (Kazley, Simpson & Chavin, 2012;Prochaska et al., 2004). The addition of socioeconomic variables into an existing TTM model was an opportunity to evaluate whether empirically relevant socioeconomic variables were related to measures of behavior change.
This study aimed to supplement current research by (1) replicating previously established relationships between the TTM constructs Stage of Change, Decisional Balance, and Self-Efficacy, (2) examining the degree to which socioeconomic variables were related to three central TTM constructs, and (3) examining whether a model of readiness for pursuing LDKT was significantly improved with the inclusion of evidence-driven socioeconomic variables.
While this study evaluated the usefulness of socioeconomic barriers in understanding LDKT decision-making, it also provided an additional evaluation of TTM constructs in a relatively nascent content area. If measures of SES, in conjunction with TTM constructs, improve our understanding of a complex decisionmaking process, it may be important to incorporate a socioeconomic perspective into future models of health behavior change.

Sample.
This study involved secondary data analysis of baseline data from the Your Path to Transplant study (Waterman et al., 2014). Your Path to Transplant (YPT) is a longitudinal randomized control trial with one aim of reducing racial disparities in LDKT by measuring and providing feedback on transplant decision-making and knowledge compared to a usual care education control group (Waterman et al., 2014).
YPT is a computer-tailored intervention primarily delivered via telephonic coaching in which validated measures of TTM decision-making constructs were used to create

Measures.
Demographics. The baseline demographics available for analysis included age, sex, race and ethnicity, dialysis status, and presence of hypertension and diabetes (Table 3).

Transtheoretical Model. TTM measures included Stage of Change, Decisional
Balance, and Self-Efficacy, and were created and validated for LDKT decisionmaking (Waterman et al., 2015). Between two separate samples of ESRD patients, the scales demonstrated strong internal reliability and validity, and relationships between constructs were found to be externally valid and comparable to similar models of health decision-making (Waterman et al., 2015;Hall et al., 2007;Plummer et al., 2001).
LDKT Readiness. Stage of Change was measured as self-reported readiness for pursuing LDKT. Seen in Table 1, the staging algorithm classified participants into four Stages of Change: Precontemplation (I am not considering taking actions in the next six months to pursue living donation), Contemplation (I am considering taking actions in the next six months to pursue living donation), Preparation (I am preparing to take actions in the next 30 days to pursue living donation), and Action (I am taking actions to pursue living donation) (Waterman et al., 2015). Construct validity was tested by examining whether patients in Action reported completion of certain LDKT behaviors compared to Pre-Action Stages, which included seven common behaviors such as "Generally talk to people about my interest in transplant" and "Accept someone's offer to donate" (Waterman et al., 2015).
Results of the analyses showed that patients in Action had completed more LDKT behaviors than those in earlier Stages, and Action could be predicted by certain behaviors, such as sharing a need for a living donor to a larger community (Waterman et al., 2015).
In invariance testing, Stage was invariant for gender and education level (Brick et al., 2016). Stage distribution varied significantly by race/ethnicity, with Black participants more likely to have been in Pre-Action Stages of Change, but this variance was consistent with LDKT trends in minority groups (Brick et al., 2016).
Lastly, movement through the Stages of Change revealed that Pros increased 0.92 SD, Cons decreased 0.29 SD, and Self-Efficacy increased 0.80 SD from Precontemplation to Action (Waterman et al., 2015). These relationships are congruent with changes by Stage in previous TTM models across a range of health behavior changes (Hall & Rossi, 2004 (Waterman et al., 2015).

Self-Efficacy for Living Donation.
The Self-Efficacy scale is a six-item measure of participants' confidence in their ability to pursue LDKT even when faced with a variety of difficult situations (Waterman et al., 2015). This measure includes statements such as "You don't know how to discuss living donation with potential donors," and "A potential living donor who was evaluated did not match you." All Self-Efficacy scale items are listed in Table 2. Self-Efficacy to continue pursuing LDKT, despite the given barrier, is rated on a five-point scale ranging from 'Not at all Confident' (1) to 'Completely Confident' (5). For the analyses in this study, the six Self-Efficacy items were summed to create a single variable. Self-Efficacy scores ranged from 6 to 30, with a mean score of 21.07 (SD = 6.64).
Previous validation testing found this measure to be internally consistent (Sample 1: a = 0.90, Sample 2: a = 0.88) and Self-Efficacy increased 0.80 SD from Precontemplation to Contemplation, consistent with previous TTM measures (Prochaska et al., 1994).

Socioeconomic Variables.
Evidence suggests that socioeconomic barriers may derail the transplant process, and may contribute to lower transplant utilization among patients with fewer socioeconomic resources (Gore et al., 2009  Employment status was unable to be used for analysis due to a greatly reduced sample size when the income sources were organized into independent categories. In this measure, participants were asked to select as many types of employment as applicable, which resulted in substantial overlap between income sources. Independent income groups reduced the sample by 34.5% (n = 523), and results may not have been representative of the total sample. Health Insurance was also excluded from analysis despite evidence that health insurance significantly impacts LDKT receipt (Gore et al., 2009;Schold et al., 2011).
In preparation for analysis, this variable was grouped into four separate categories, without reducing sample size ('Single private health insurance,' 'Single government insurance,' 'Multiple private,' or 'Multiple government'). However, the quality of health insurance could not be determined because of substantial changes in insurance markets during the course of data collection that may have reduced coverage discrepancies between private and government insurance plans. Results would not be interpretable without additional information on potential coverage differences between government versus private health insurance plans.

Hypotheses and Planned Analyses.
All statistical analyses were conducted using SPSS Version 24.

Preliminary Analyses. All participants from treatment and control baseline
surveys were included in the sample. Socio-demographic analyses were used to describe the sample by examining means and frequencies of TTM constructs (Pros,

Cons, and Self-Efficacy), SES variables, and health measures across four Stages of
Change and by race/ethnicity. In addition, a series of chi-square tests were conducted to evaluate independence between SES variables; these tests aided in interpretation of results from Hypotheses 2 and 3.

Hypothesis 1: Replicating established transtheoretical model relationships.
Relationships between TTM constructs of LDKT decision-making were compared to the established TTM relationships found in measure development as well as TTM models developed in other health settings (Waterman et al., 2015;Hall & Rossi, 2008). It was predicted that relationships between baseline measures of Stage of Change, Decisional Balance, and Self-Efficacy would be consistent with previous studies of health behavior change.
Specifically, we expected Decisional Balance and Self-Efficacy to be significantly related to Stage of Change. For Decisional Balance, we predicted that higher Cons would be associated with earlier Stages of Change, Precontemplation and Contemplation, and higher Pros would be associated with later Stages of Change, Preparation and Action. We predicted that Self-Efficacy would increase across Stages of Change, with greater Self-Efficacy in later Stages of Change.
Analyses. Multivariate analysis of variance (MANOVA) was used to evaluate group differences between four Stages of Change by Decisional Balance and Self-Efficacy. Separate analysis of variance (ANOVA) tests were used to examine differences between Stage groups. Tukey's post hoc tests were used to test for significant differences between earlier and later Stages of Change. Standardized Tscores were used to clarify relationships, and differences in T-score standard deviations were used to aid in comparison with previous TTM models.
Hypothesis 2: Relationships between socioeconomic and TTM variables.

2.a. Decisional Balance and Self-Efficacy.
This analysis explored the degree to which socioeconomic variables were related to two TTM constructs, Self-Efficacy and Decisional Balance. We predicted that indications of greater SES (higher level of education, lower income vulnerability, feeling safe in one's neighborhood, owning a washer and dryer, and owning a vehicle) would be associated with greater Self-Efficacy, greater Pros for pursuing LDKT, and lower Cons.

Analysis 2a. With each of the SES variables, a series of one-way ANOVAs
were conducted to evaluate strength and direction of relationships with three continuous TTM variables: Pros, Cons, and Self-Efficacy. .001, Φ = .16; χ 2 (1) = 4.99, p < .05, Φ = .10). The odds of diabetes were almost twice as great for Hispanic patients than Whites (Table 5).

Self-Efficacy.
A follow-up ANOVA test for Self-Efficacy revealed significant differences between Stage groups (F(3, 795) = 18.50, p < .001, η 2 = .085), with a small to medium effect size. Follow up Tukey's post hoc tests revealed significantly greater Self-Efficacy among those in Action versus Precontemplation.
Standardized T-scores of Pros, Cons, and Self-Efficacy scales (M = 50, SD = 10) were calculated to assist with the interpretation of TTM relationships. T-score differences between Precontemplation and Action indicated that Pros increased 0.81 SD, Cons decreased 0.45 SD, and Self-Efficacy increased 0.91 SD (Table 6). In Figure 1, mean T-scores for Self-Efficacy, Pros, and Cons are graphed across the Stages of Change.

Hypothesis 2: Relationships between socioeconomic and TTM variables.
A series of ANOVA and chi-square tests were conducted to evaluate whether significant relationships exist between SES and TTM variables, as well as the direction of such relationships.

2.a. Decisional Balance and Self-Efficacy. Analyses involved three one-way
ANOVA tests per SES variable. We expected indications of greater SES to be related to higher Self-Efficacy, higher Pros, and lower Cons.

Education Level. Education Level was significantly related to Decisional
Balance variables. No significant relationships were detected between Education Level and Self-Efficacy (F(4, 794) = 2.29, p > .05).
Education Level was significantly related to Pros (F(4, 267.8) = 4.07, p < .01, η 2 = .02). The effect size was small. Contradictory to our hypothesis, Education and subjects with a professional or graduate degree (n = 74, M = 15.4), but was not significantly different than subjects with a college degree (n = 204, M = 18).

Financial Security. No significant relationships were detected between
Financial Security and Pros, Cons, or Self-Efficacy (Table 7).
Neighborhood Safety. Neighborhood Safety was not significantly related to Pros, Cons, or Self-Efficacy (Table 7).
Washer/Dryer. Owning a Washer or Dryer was not significantly related to Pros, Cons, or Self-Efficacy (Table 7).
Vehicle. Owning or having access to a vehicle was not significantly related to Pros or Self-Efficacy for pursuing LDKT (Table 7). A significant relationship was detected between access to a Vehicle and Cons for pursuing LDKT, with a small effect size, F(1, 797) = 5.35, p < .05, η 2 = .001.

2.b. Stage of Change.
A series of chi-square analyses were conducted to evaluate the relationship between Stage of Change and five SES variables. Seen in Table 8, no statistically significant relationships were detected.

Hypothesis 3: Modeling TTM Constructs with Socioeconomic Variables.
Two binary logistic regression models were tested to evaluate whether the addition of five SES measures into an existing TTM model significantly improved Stage grouping in this sample. We predicted that model fit would be significantly improved with the inclusion of SES variables in the TTM model, and that indications of greater SES would be related to being in Action compared to Pre-Action Stages.

Preliminary Analyses.
Relationships between SES variables were tested over a series of nine chi-square tests, with effect size reported with Cramer's V or Phi coefficients ( Table 9). Results of the chi-square tests revealed that all five socioeconomic variables were significantly related to each other at the p < .001 significance level. Effect sizes for this set of analyses were small and ranged from .14 to .21. Small effect sizes supported the inclusion of all five SES variables into the tested model, as the variables appeared to have measured related dimensions of SES without substantial overlap.

Analysis 3.a.
Stepwise binary logistic regression analysis was used for model testing. The first step sought to replicate previous TTM relationships and the second step evaluated whether the addition of SES variables improved model fit.
Step one included Decisional Balance and Self-Efficacy as IVs, and baseline Results from step two of logistic regression analysis detected no statistically significant improvement in model fit with the addition of SES variables (Table 10).
Non-significant Wald tests indicated that effects on Stage grouping were undetectable at the .05 significance level. Without the addition of significant SES variables, this second model was essentially identical to the replicated TTM model in step 1, with a minor increase in correct classification to 58.9% (Pre-Action: 75.6%, Action: 38%).
Odds ratios for Pros, Cons, and Self-Efficacy remained unchanged from step one.

DISCUSSION
The pursuit of living donor kidney transplant requires resources and opportunities that may not be available to many patients with lower socioeconomic status. This study examined relationships between readiness for pursuing living donor transplant and socioeconomic barriers that could negatively influence efforts to pursue transplant from a living donor. Socioeconomic barriers did not add to a model of readiness for pursuing LDKT when analyzed with a set of dynamic TTM constructs.
Results from MANOVA and ANOVA analyses supported the use of the transtheoretical model of readiness for pursuing LDKT, and the relationships found between the TTM constructs were successfully replicated in a new sample of ESRD patients. Consistent with prior research, Self-Efficacy and Pros were the main drivers of Stage progression, while Cons had a weaker effect on readiness (Prochaska, 1994).
From participants in Precontemplation to those in Action, Self-Efficacy increased 0.91 SD, Pros increased 0.81 SD, and Cons decreased 0.45 SD. This relationship is presented in Figure 1, and is consistent with the results reported in initial TTM measure development for pursuing LDKT (Waterman et al., 2015). The Pros and Cons intersected between Contemplation and Preparation Stages, consistent with the relationship of these variables seen in a meta-analytic review of cross-sectional TTM models in other behavior change areas (Hall & Rossi, 2008).
Moreover, the replicated model demonstrated dynamic relationships between constructs when modeled with logistic regression. For every one-unit increase, the odds of being in Action increased 1.05 for Pros and 1.04 for Self-Efficacy, and decreased 0.96 for Cons. Therefore, the odds of being in Action were 6.3 times greater for patients who valued the Pros of LDKT as 'Very Important' to their decision to pursue transplant compared to 'Moderately Important.' These findings suggest that patients in Action were clearly distinguished from earlier Stages by their perception of transplant benefits and degree of confidence for pursuing LDKT.
The successful model replication in this study provides additional evidence that the TTM can model decision-making processes in complex behavior change areas. Pursuing LDKT is a particularly complex set of behaviors because a patient's success relies on the behavior of a second person, the living donor. Moreover, the staging algorithm did not identify a defining behavior that clearly separated Action from Pre-Action, which typically requires the adoption or extinction of a specific behavior. Lastly, being in Action does not guarantee that a patient will receive an LDKT despite engaging in LDKT behaviors. While patient-level change plays an important role in LDKT receipt, factors outside the patient's control also weigh heavily on outcomes, such as matching with a living donor or meeting certain health criteria.
Transplant readiness has been identified as an important predictor of LDKT receipt, and a body of literature has identified racial and socioeconomic disparities in LDKT utilization (Gore et al., 2009;Waterman et al., 2013). However, this is the first study to directly test whether the inclusion of socioeconomic variables would improve our understanding of LDKT readiness. After testing variables independently and as covariates, this study did not find significant improvement in model fit when socioeconomic variables were added to the readiness model. Socioeconomic status did not appear to account for differences between patients in Pre-Action Stages compared to those in Action. However, it is possible that the cross-sectional constraints of the present study design limited our ability to detect relationships with socioeconomic variables.

Race & Differences in Stage
When socio-demographic variables were analyzed across racial-ethnic groups,  Petersen et al., 2008).

Future Directions.
While readiness for pursuing LDKT did not appear to be related to socioeconomic barriers or advantage in this study, more research is needed to substantiate these findings. An important next step is to examine the LDKT readiness model longitudinally, which would be important for determining predictive relationships within the model. Researchers should also test for relationships between socioeconomic barriers and readiness longitudinally.
Additionally, future research should investigate the relationship between readiness for pursuing LDKT and transplant receipt. While readiness was found to predict transplantation in a previous study, this relationship has not been established with the validated TTM scales used in the present study (Waterman et al., 2013;Waterman et al., 2015).
In conclusion, living donor kidney transplant has proven to be the treatment of choice for enhancing wellbeing and survival of patients with chronic renal failure.
Transplant research continues to investigate inequities that impact the decision to pursue kidney transplants from living donors. Reducing barriers to kidney transplant, with focus on the dynamic factors that are emphasized by the TTM, has potential to lead to greater use of LDKT and ultimately improve treatment outcomes for end-stage renal disease.