Secondary Logo

Journal Logo

Patient Preferences for Outcomes After Kidney Transplantation: A Best-Worst Scaling Survey

Howell, Martin PhD1,2; Wong, Germaine PhD1,2,3; Rose, John PhD4; Tong, Allison PhD1,2; Craig, Jonathan C. PhD1,2; Howard, Kirsten PhD2

doi: 10.1097/TP.0000000000001793
Original Clinical Science—General
Free
SDC
Social Media Collection

Background The care of kidney transplant recipients involves a balance between maximizing graft survival and serious adverse outcomes. This study aimed to quantify patients’ preferences and trade-offs for important outcomes after transplantation.

Methods A best-worst scaling survey, analyzed by multinomial-logit models, was used to calculate normalized preference scores (0, best; 1, worst), for varying years of graft duration and risk of dying before graft failure, cancer, cardiovascular disease, diabetes, infection, anxiety/depression, diarrhoea/nausea, and weight gain. Willingness to trade years of graft survival to minimize the risk of adverse outcomes was calculated.

Results Ninety-three transplant recipients from 2 Australian transplant units and an on-line panel (aged 18-69 years (mean time since transplantation, 7 years) completed the survey. Graft loss at 1 year was the least desirable outcome (mean preference value, 0.0:95% confidence intervals, −0.05 to 0.05) and worse than a 100% risk of dying before graft loss (0.17: 0.12-0.23). Graft duration of 5 years had the same preference scores (ie, as bad) as the maximum risk of all adverse outcomes including a 100% risk of dying before graft failure. To achieve zero risk of cancer, dying, and cardiovascular disease participants were only willing to trade 3.1(2.1 to 4.7), 1.7(1.1 to 2.5), and 1.2(0.8 to 1.8) years of graft survival, respectively, and less than 1 year for all other outcomes.

Conclusions Transplant recipients regarded graft loss as worse than death and showed minimal willingness to trade a reduction in this outcome with an improvement in any other outcome.

This small online survey completed by 93 kidney transplant recipients suggests that graft loss is perceived as worse outcome than death, with minimal willingness to trade any year of graft survival for reduction in the risk of cancer, cardiovascular disease, or death. Supplemental digital content is available in the text.

1 Centre for Kidney Research, The Children’s Hospital at Westmead, Westmead, NSW, Australia.

2 Sydney School of Public Health, The University of Sydney, Sydney, NSW, Australia.

3 Centre for Transplant and Renal Research, Westmead Hospital, NSW, Australia.

4 Institute for Choice, University of South Australia Business School, NSW, Australia.

Received 13 July 2016. Revision received 27 March 2017.

Accepted 7 April 2017.

M.H. is supported by a National Health and Medical Research Council Capacity Building Grant ID 571372. A.T. is supported by the National Health and Medical Research Council Fellowship (ID 1106716). The funding organization had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review, or approval of the article.

The author declares no conflicts of interest.

M.H. contributed to the study design, conducted the survey, conducted analyses and wrote the article. G.W. contributed to the study design, data analyses, and article preparation and review. J.R. contributed to the study design, data analyses, and article preparation and review. A.T. contributed to the study design and article preparation and review. J.C. contributed to study design and article preparation and review. K.H. contributed to the study design, data analyses, and article preparation and review. All authors had full access to all data and analysis.

Correspondence: Martin Howell, PhD, Centre for Kidney Research, The Children’s Hospital at Westmead, Westmead NSW 2145, Sydney, Australia. (martin.howell@health.nsw.gov.au).

Supplemental digital content (SDC) is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (www.transplantjournal.com).

The care of kidney transplant recipients involves a balance between maximizing graft survival and serious adverse outcomes including cancer, cardiovascular disease (CVD) and infection associated with immunosuppression.1-3 Although there has been a marked improvement in short-term graft and patient survival with advancements in drugs and surgery, long-term graft survival has remained relatively static over the past 2 decades, and there remains an uncertainty in the influence of treatment regimens on long-term outcomes.4-6

Understanding patient preferences is key to ensuring that, as far as possible, treatment regimens reflect patient values and variation in tolerance of risk.7-9 Because preferences are underpinned by beliefs of consequences, patients should be provided with an unbiased assessment of uncertainties, benefits, and harms.10,11 Before transplantation, patient expectations rightly focus on the positive aspects12; however, preferences change with age health and personal circumstance.8,13,14 It is therefore important to understand preferences before and after transplantation. Prior research has quantified the frequency and severity of adverse effects associated with long-term immunosuppression.15-18 Kidney transplant recipients have indicated a strong focus on graft survival, aversion to returning to dialysis, and willingness to accept side effects and adverse outcomes as a necessary part of the treatment.19-24

Opportunities for shared decisions before and after transplantation have been identified.25-28 Knowledge of preferences after transplantation provides insights into patient views of the benefits and harms of immunosuppression, level of tolerance of side effects and sharing decisions about the management of adverse outcomes.27,28 Despite the role for patient preferences, there is scarce data about kidney transplant recipients preferences and trade-offs they may be willing to make to reduce the impact of adverse outcomes through change, or withdrawal of, immunosuppression, and the risk of graft loss. This study specifically focuses on preferences after transplantation.

The aim of this study was to evaluate preferences and trade-offs transplant recipients may be willing to accept to avoid adverse outcomes of long-term immunosuppression using a quantitative technique, a best-worst case scaling survey (BWS). A BWS is a type of discrete choice experiment increasingly used to evaluate preferences for health programs and interventions,29 including complex treatments, such as immunosuppression after transplantation.

Back to Top | Article Outline

MATERIALS AND METHODS

Participant Selection

Adult kidney transplant patients (18 years and older) were recruited using a convenience sampling approach from 2 large transplant units in Sydney, Australia, and an Australia wide on-line panel administered by an external organization (Survey Sampling International). English-speaking patients able to provide informed consent were eligible. Recruitment took place by M.H. who attended day clinics and approached patients with follow-up by email or letter. Patients not attending the clinics on these days were contacted by email, letter or phone. A dummy question was used to ensure that only transplant recipients participated from the community panel. Ethics approval was obtained from the Human Research Ethics Committee, Western Sydney Local Health Network, NSW Health (HREC2009/6/4.15 (2956)/AU/RED09/WMEAD/56).

Back to Top | Article Outline

Best-Worst Scaling

In best-worst case scaling participants are shown a list of health outcomes and asked to choose the “best” and “worst.” The outcomes are varied across multiple profiles with preferences estimated from choices by multinomial logit (MNL) regression.29,30 Survey design and analysis are underpinned by random utility theory.31

Based on findings of a qualitative study21 and pilot BWS,32 9 outcomes representing those most likely to be important to patients were included. Eight outcomes (dying before graft loss, cancer [other than skin], CVD, diabetes, serious infection, excessive weight gain, severe diarrhea/nausea, and severe anxiety/depression) were presented as the risk of occurrence after transplantation, whereas graft survival was expressed as years. Outcome levels covered clinically plausible ranges (Figure 1).

FIGURE 1

FIGURE 1

Back to Top | Article Outline

Survey Design

A statistically efficient experimental design method33 informed by the pilot32 was used to combine outcome levels into profiles. Following recommended practice, parameter estimates from the pilot are used to maximize the designs statistical efficiency.31 The more efficient the design, the smaller the sample size required. The final design indicated statistical significance of main effects (P < 0.05) would be achieved with a sample size less than 50. Furthermore, a recognized simplified sample size estimation method based on the number of outcomes, outcome levels, and choice sets of the final design34,35 suggests a minimum sample size of 70. To minimize the number of choice sets, respondents were asked to make 4 choices “best,” “worst,” “next-best,” and “next-worst.” The design included 4 blocks of 10 profiles with participants randomly assigned by the survey software to 1 block, whereas paper versions were distributed in approximately equal numbers. Online surveys prevented an attribute being marked best and worst. Respondents were also asked to consider the profile as a hypothetical treatment: “If you were offered a treatment that resulted in all of the above outcomes, would you take it?” (Figure 1). Self-reported demographic and medical details including medication, comorbidities, dialysis duration before transplantation, donor type, number of transplants, and time since last transplant were collected. Paper versions were handed out in the clinic or mailed. All were self-completed without assistance.

An estimate of major errors was made by counting the number of times 25 years graft duration was chosen as a “worst” outcome and 100% risk of dying as a “best” outcome. All survey responses were included except where choice selection was unclear or missing (in paper versions).

Back to Top | Article Outline

Analysis

MNL regression models were estimated using Python BIOGEME software.36 Utility functions were constructed for each outcome and each of the choices “best.” “second-best,” “worst,” “second-worst” (Appendix S1, SDC,http://links.lww.com/TP/B441). Outcome levels were dummy coded and choices assumed to be sequential, “best,” then “worst,” “next-best,” and “next-worst” with the chosen outcome unavailable for subsequent selection.37 The middle outcome levels were set as reference values. Age, sex, comorbidities, time on dialysis, time since transplantation, and number of transplants were included as covariates. Mean regression coefficients for outcome levels and 95% confidence limits were estimated using a MNL model.37,38

Outcome level preference scores were obtained by normalizing the mean coefficients to the range 0 to 1, where 0 is least and 1 most preferred and can be compared across all attributes. Within an outcome the magnitude of change between preference scores indicates sensitivity to change and indifference where scores are equal.

To examine effects of covariates on preferences, coefficients were estimated separately for “best” and “worst” choices for each outcome relative to weight gain. To assist in interpretation, coefficients are expressed as odds ratios. An odds ratio less than 1 indicates a lower probability of the outcome being selected and greater than 1 a higher probability, all else being equal. We would expect a priori, that a “best” choice would be for lower rather than higher risks of adverse outcomes and longer rather than shorter graft survival with the converse for “worst” choices.

Responses to the question “If you were offered a treatment that resulted in all of the above outcomes, would you take it?” were modeled using a mixed-MNL model with panel specification with “no” as the reference. All outcome coefficients were assumed to be normally distributed; however, to maintain parsimony, the final model only included random parameters with significant standard deviations (P < 0.05) (Appendix S2, SDC,http://links.lww.com/TP/B441) with the remainder entered as fixed parameters. Odds ratios greater than 1 indicate a higher probability that “yes” will be chosen as the outcome level becomes more favorable, that is, as years of duration increase or as risk reduces.

Benefit/harm trade-offs for graft survival and risk of adverse outcomes were estimated from the ratio of the coefficients for each adverse outcome and graft survival39 (Appendix S3, SDC, http://links.lww.com/TP/B441) with confidence limits estimated using the Krinsky-Robb procedure.39,40 The benefit/harm trade-off is the sensitivity to the risk of adverse events expressed on the common scale of years of graft survival. The benefit/harm trade off was calculated for 0%, 10%, 30%, and 50% risk of occurrence of the adverse outcomes and can be interpreted as the years of graft survival that hypothetically would be traded to ensure the specified risk.

Back to Top | Article Outline

RESULTS

Participant Characteristics

Ninety-three patients completed the BWS, with 30 (32%) and 63 (68%) from the community panel and transplant clinics, respectively. It is not possible to estimate a meaningful response rate because not all patients enrolled at the clinics could be contacted and the number of transplant recipients on the panel is not recorded. Of the 103 transplant recipients contacted in the clinics, 63 (61%) completed the survey. Respondents were aged between 18 and 69 years (mean, 50.5; SD, 9.7), 59 (63%) were male, 88 (95%) spoke English as the first language, and 60 (65%) completed high school. Time since the last transplant ranged from 0.9 to 31 years (mean, 7.7; SD, 5.7) and 49 (53%) received deceased donor grafts. Sixty-nine (74%) were on dialysis before transplant, with 38 (41%) for more than 2 years. Maintenance immunosuppression was dominated by prednisone (84%), tacrolimus (62%), and mycophenolate mofetil (66%). The most common comorbidities were hypertension (75%), high cholesterol (50%), diabetes (33%), and skin cancer (22%). Multiple comorbidities were reported by the majority with 61 (66%) reporting 2 or more comorbidities (Table 1). Patient characteristics were comparable to the prevalent kidney transplant population reported for 2013 in the Australian and New Zealand Dialysis and Transplant Registry (ANZDATA) registry for age, sex, number of transplants and immunosuppression (Table 1). The proportion of living donor transplants was higher in those completing the survey (47%) compared with the 2013 prevalent population (36%).

TABLE 1

TABLE 1

Back to Top | Article Outline

Survey Completion

Of 93 surveys completed, 71 (76%) were online, and all 10 scenarios were completed. Missing choices and double entries occurred for 139 (4%) of the 3720 selections. The unintuitive “best” choice (100% chance of dying) and “worst” choice (25-year graft survival) occurred for 2 (1.1%) of 178 choices and 8 (4.5%) of 176 choices, respectively.

Back to Top | Article Outline

Preference Scores

Preferences scores, with a range of 0 (worst) to 1 (best), are plotted in Figure 2 (see Table S1, SDC,http://links.lww.com/TP/B441 for regression coefficients) and show relative importance within and between outcomes. The worst (least desirable) outcome was graft failure after 1 year with a preference score of 0.0 (95% confidence interval, −0.05 to 0.05), whereas the best (most desirable) was 0% risk of dying before graft failure with a preference score of 1.0 (0.92-1.08). Consistent with expectations, as risks of adverse outcomes increased, preferences decreased, and as duration of graft survival increased, preferences increased.

FIGURE 2

FIGURE 2

Comparison of preferences between outcomes implies that graft loss after 1 year was worse than a 100% risk of dying before graft failure. Five years of graft survival was as important as 100% to 75% risk of dying and anxiety/depression, 100% risk of diarrhea/nausea, 50% to 30% risk of cancer, 50% risk of CVD, diabetes, and serious infection. Fifteen years of graft survival was as important as 0% risk of weight gain; 0 to 10% risk of diabetes, 10% risk of cancer, CVD, and infection; 25% risk of dying, and diarrhea/nausea, and 30% risk of infection. Finally, 25 years of graft survival was as important as 0% risk of cancer, CVD, depression/anxiety, infection, and diarrhea/nausea. Only 0% risk of dying before graft failure was more important.

Back to Top | Article Outline

Sensitivity to Change in Levels

On average, respondents were most sensitive to changes at low-risk values and less sensitive or indifferent to changes at high risk (Figure 2). This was particularly evident for dying with a functioning graft where the preference at 0% was more than double that at 25% (values of 1.0 [0.92-1.08] and 0.41 [0.36-0.46], respectively). In contrast, respondents were indifferent to change between 75% and 100% risk (scores of 0.20 [0.14-0.25] and 0.17 [0.12-0.23], respectively). Variation in sensitivity to change was also evident for cancer, CVD, and anxiety/depression. Preferences for graft survival showed similar sensitivity over 1 to 25 years with scores of 0.0 (−0.05 to 0.05), 0.19 (0.14-0.23), 0.46 (0.42 to 0.51), and 0.72 (0.66 to 0.78), respectively (Figure 2).

Back to Top | Article Outline

Benefit/Harm Trade-off

Benefit/harm trade-off between the risks of an adverse outcome and years of graft survival are plotted in Figure 3. The trade-off for 0% risk of cancer, dying, CVD, diabetes and depression/anxiety, was 3.10 (2.10-4.69), 1.66 (1.12-2.51), 1.21 (0.82-1.83), 0.87 (0.59-1.32), and 0.81 (0.55-1.22) years graft survival, respectively. Willingness to trade graft survival declined as the risk increased. The trade-off for 30% risk for the same outcomes were 0.09 (0.06-0.13), 0.38 (0.26-0.57), 0.28 (0.19-0.42), 0.21 (0.14-0.31), and 0.27 (0.18-0.41) years, respectively.

FIGURE 3

FIGURE 3

Back to Top | Article Outline

Influence of Respondent Characteristics

The influence of respondent characteristics on choices is indicated by the odds ratios for best and worst choices (Table 2). On average, with increasing age respondents were less likely to select a low risk of infection, diarrhea/nausea, depression/anxiety, and diabetes as best (odds ratios of 0.94 (0.92-0.97), 0.94 (0.92-0.97), 0.95 (0.91-0.98), 0.95 (0.93-0.98), respectively) and more likely to select short graft survival as worst with an odds ratio 1.07 (1.03-1.12). This suggests lower concern for these outcomes and greater concern for avoiding short graft survival with increasing age. Compared with men, women were more likely to select a low risk of depression/anxiety, CVD and cancer, and long graft survival as best (odds ratios of 2.77 [1.39-5.54], 2.89 [1.45-5.74], 2.38 [1.19-4.73], and 1.68 [1.06 to 2.66], respectively) and more likely to select a high risk of diarrhea/nausea and depression/anxiety as worst (odds ratios of 2.47 [1.08-5.64] and 2.13 [1.02-4.48], respectively). On average, women had a greater concern for these outcomes and long graft survival compared with men. Increasing comorbidity was associated with more concern for long graft survival and less for short graft survival, diabetes, and dying. Increasing years on dialysis before transplantation was associated with greater concern for long graft survival and CVD. Increasing years since the last transplant was associated with greater concern for cancer and less for long graft survival and CVD. Finally, having had more than 1 transplant was associated with greater concern for long graft survival, dying before graft failure, cancer, CVD, diabetes, and anxiety/depression.

TABLE 2

TABLE 2

Back to Top | Article Outline

Acceptance of Treatment Outcomes

Graft survival and dying were the only outcomes associated with selection of treatment outcome profiles (P < 0.05). Odds ratios indicated an increasing likelihood (1.54, 1.22-1.94) of accepting the treatment for every year of graft survival and a decreasing likelihood (0.96, 0.93-0.91) of acceptance for a 25% increase in the risk of dying. Odds ratios are provided in Table S2, SDC,http://links.lww.com/TP/B441.

Back to Top | Article Outline

DISCUSSION

This BWS demonstrates the overwhelming importance of graft survival to transplant recipients, which dominates all other outcomes. Quantifying choice is inherently a difficult concept but this consistent finding is exemplified by the following results. Very short duration of graft survival (1 year) is worse than any outcome, including a 100% risk of dying before graft failure. Five to 15 years of graft survival is regarded by transplant recipients as equivalent to a high risk of serious adverse outcomes including dying before graft failure. It is only beyond 15 years of graft survival that low risks of serious adverse outcomes are of equal or greater importance. There is a willingness to trade only a small number of years of graft survival to avoid adverse outcomes, and only to achieve an improbably low risk. Graft survival was the only attribute positively associated with accepting the hypothetical treatment.

The importance of graft survival is consistent with expectations, but not the magnitude of dominance with 15 years duration for a functioning graft being preferred to a low risk of all outcomes including dying with a functioning graft. Follow-up in kidney transplantation trials is mostly less than 5 years after transplantation.41-45 For example, in a recent review of the 233 randomized controlled trials published between January 2003 and December 2015, 64% were less than 2 years duration and only 13% were more than 4 years.46 There are logistical issues in ascertaining long-term outcomes; nonetheless, if research is to inform decision-making, it should align with patient expectations.

After dying, cancer was the most important adverse outcome for preference and trade-offs while diabetes was least important. This may reflect aversion to cancer21,47 and/or limited patient experience of cancer (4% of respondents). The low importance of diabetes may reflect high prevalence (33%), with familiarity and normalization.48 Zero risk of depression/anxiety and diarrhea/nausea was as important as 25 years graft survival and more important for women compared with men. Patients are willing to accept side effects as necessary,19-21,23 and tend not to discuss them with clinicians, leading to underestimation of frequency and severity.49 Our study suggests greater focus on prevalence and severity of side effects in research, and clinical care is warranted.

Willingness to trade years of graft survival to avoid adverse outcomes is strongly dependent on the level of risk. The benefit/harm trade-off to achieve a 0% risk of cancer and dying before graft failure was 3.1 and 1.7 years, respectively, and less than 1 year for a 20% risk. This has implications for risk communication and understanding preferences, values, and risk tolerance. Interventions cannot eliminate all risk and contextualizing absolute event rates across alternate regimens, or comparison with patients on dialysis or the general population may be preferred.50

The importance of graft survival and adverse events varies with age, sex, dialysis, comorbidities, number of kidney transplants, and duration since transplantation. Variability of preferences should be anticipated given that they are underpinned by experience, knowledge, values, and beliefs.51 A recent qualitative study found that some transplant recipients were cautious about treatment options involving changes to immunosuppression, which was shaped by prior experience of dialysis, age, sex and accepting side effects as a “necessary evil.”24 Patient-centered care requires acknowledgment of patients' informed preferences and avoidance of preferences framed by “ill-informed” beliefs.24,52 However, although the BWS provides an insight into preferences and priorities, it cannot be concluded that transplant recipients would seek to maximize duration of graft survival at all costs. Nor can it be assumed that recipients would choose not to return to dialysis.

Transplant patient preference studies have mostly addressed acceptance of donor organs.53-56 A maximum difference scale (similar to a BWS) used to assess waiting list patient and clinicians' priorities for deceased donor organs identified organ quality and function as most important.56 The longer time on dialysis, the less important was graft function. However, adverse outcomes and graft/patient survival after transplantation were not included. In a discrete choice experiment with transplant waitlist patients,55 older age and dialysis time was associated with a greater likelihood of accepting high-risk organs. Attributes included donor age, longer waitlist time, and the risk of HIV infection, but patient/graft survival was omitted. A survey of Slovenian dialysis patients, who chose to remain on dialysis instead of transplantation, identified the prime reason for their decision as concern for adverse and uncertain outcomes after transplantation. Most important were cancer, diabetes, and infection, but patient and graft survivals were not included in the survey.54 Acceptance of reduced life expectancy after acute rejection treatment compared with no treatment and return to dialysis was assessed using a standard gamble technique53 among transplant recipients and waitlist patients. Consistent with our study, preferences were found to vary with patient age, sex, and time on dialysis and the waitlist. The utility of returning to dialysis was zero for 14% of the transplant recipients, meaning they would rather die than return to dialysis. All these studies presented unrealistic scenarios or restricted attributes from which preferences have been elicited. In contrast, a multiattribute BWS combines adverse outcomes, including mortality with duration of graft survival. Preferences reported in our study reflect individual experiences of posttransplant immunosuppression including adverse outcomes and are relevant to the long-term care of transplant recipients. However, understanding preferences of waitlist patients is critical to pretransplant care and should be addressed in future studies.26-28

There are limitations with this study. Although sex and age were representative of adult kidney transplant patients, there is a bias to educated, English-speaking white patients. The proportion of recipients with living donor transplants was higher in the survey (47%) compared with the 2013 prevalent population (36%) reported by ANZDATA (Table 1). The maintenance immunosuppression, although reflecting current practice in Australia and New Zealand,57 may limit generalizability. Our sample was a convenience sample of patients attending 2 large transplant units in Australia which limits generalizability. However, the overwhelming focus on graft survival has been a consistent finding,21,24,58,59 and our study provides additional insights not previously possible. We have assumed symmetry of choices to allow for smaller sample sizes and less complex surveys; however, choice may be influenced by positive and negative framing.37,38 The analysis assumes choices are made sequentially, whereas for all or some scenarios patients may chose the worst, then best or simultaneously.38 Different assumptions about symmetry and order of selection could give slightly different estimates particularly where an attribute is consistently chosen as for dying and graft survival. To reduce respondent burden, we considered it more important to allow flexibility in the way individuals made selections rather than forcing a prespecified order. The survey did not include outcomes after graft loss. Mortality may be higher than that for transplant naive patients,60 graft intolerance syndrome may occur,61 and there may be a lower likelihood of being waitlisted with longer waiting times62 all of which could influence preferences. Finally, respondent medical details were self-reported.

In conclusion, elicitation of preferences and trade-offs, expressed quantitatively for outcomes after transplantation, has demonstrated graft survival to be the outcome of prime importance for kidney transplant patients. Adverse outcomes of transplantation and drug-related side effects were also regarded as important, but graft survival of 5 to 15 years duration is, on the basis of preferences elicited, worse than any other outcome including dying. If this is the case, recipients may rather not receive this “gift of life” if it is going to be “taken away from them” during a period they regard as unacceptably short (in this case, 5-15 years duration). This focus on graft survival is reflected in the limited willingness to trade graft duration, which is contingent on achieving a very low risk of any serious outcome (ie, less than 10%). Given uncertainty in predicting long-term outcomes, understanding patient preferences is essential to the development of treatment regimens for long-term immunosuppression. Research questions and outcomes should be aligned with patient priorities and preferences; and patient-centered care should acknowledge patients' preferences.

Back to Top | Article Outline

ACKNOWLEDGMENTS

The authors would like to thank the kidney transplant patients from Westmead Hospital, and Royal Prince Alfred Hospital, Sydney, Australia, who generously gave their time and shared their opinions.

Back to Top | Article Outline

REFERENCES

1. Webster AC, Craig JC, Simpson JM, et al. Identifying high risk groups and quantifying absolute risk of cancer after kidney transplantation: a cohort study of 15,183 recipients. Am J Transplant. 2007;7:2140–2151.
2. Kasiske BL, Snyder JJ, Gilbertson D, et al. Diabetes mellitus after kidney transplantation in the United States. Am J Transplant. 2003;3:178–185.
3. Matas A, Gillingham KJ, Humar A, et al. 2202 kidney transplant recipients with 10 years of graft function: what happens next? Am J Transplant. 2008;8:2410–2419.
4. Lamb KE, Lodhi S, Meier-Kriesche HU. Long-term renal allograft survival in the United States: a critical reappraisal. Am J Transplant. 2011;11:450–462.
5. Chadban S, Morris R, Hirsch HH, et al. Immunosuppression in renal transplantation: some aspects for the modern era. Transplant Rev (Orlando). 2008;22:241–251.
6. Opelz G, Dohler B. Influence of immunosuppressive regimens on graft survival and secondary outcomes after kidney transplantation. Transplantation. 2009;87:795–802.
7. Barry MJ, Edgman-Levitan S. Shared decision making–pinnacle of patient-centered care. N Engl J Med. 2012;366:780–781.
8. Epstein RM, Peters E. Beyond information: exploring patients' preferences. JAMA. 2009;302:195–197.
9. Kramer M, Bauer W, Dicker D, et al. The changing face of internal medicine: patient centred care. Eur J Intern Med. 2014;25:125–127.
10. Mühlbacher A, Juhnke C. Patient preferences versus physicians’ judgement: does it make a difference in healthcare decision making? Appl Health Econ Health Policy. 2013;11:163–180.
11. Siegel CA. Shared decision making in inflammatory bowel disease: helping patients understand the tradeoffs between treatment options. Gut. 2012;61:459–465.
12. Cleemput I, Kesteloot K, De Geest S, et al. Health professionals' perceptions of health status after renal transplantation: a comparison with transplantation candidates' expectations. Transplantation. 2003;76:176–182.
13. Blinman P, King M, Norman R, et al. Preferences for cancer treatments: an overview of methods and applications in oncology. Ann Oncol. 2012;23:1104–1110.
14. Ubel PA, Loewenstein G, Schwarz N, et al. Misimagining the unimaginable: the disability paradox and health care decision making. Health Psychol. 2005;24:S57–S62.
15. Ekberg H, Kyllonen L, Madsen S, et al. Increased prevalence of gastrointestinal symptoms associated with impaired quality of life in renal transplant recipients. Transplantation. 2007;83:282–289.
16. Kugler C, Geyer S, Gottlieb J, et al. Symptom experience after solid organ transplantation. J Psychosom Res. 2009;66:101–110.
17. Moons P, Vanrenterghem Y, Hooff JP, et al. Health-related quality of life and symptom experience in tacrolimus-based regimens after renal transplantation: a multicentre study. Transpl Int. 2003;16:653–664.
18. Ponticelli C, Colombo D, Novara M, et al. Gastrointestinal symptoms impair quality of life in Italian renal transplant recipients but are under-recognized by physicians. Transpl Int. 2010;23:1126–1134.
19. Buldukoglu K, Kulakac O, Kececioglu N, et al. Recipients' perceptions of their transplanted kidneys. Transplantation. 2005;80:471–476.
20. Goldade K, Sidhwani S, Patel S, et al. Kidney transplant patients' perceptions, beliefs, and barriers related to regular nephrology outpatient visits. Am J Kidney Dis. 2011;57:11–20.
21. Howell M, Tong A, Wong G, et al. Important outcomes for kidney transplant recipients: a nominal group and qualitative study. Am J Kidney Dis. 2012;60:186–196.
22. Nilsson M, Forsberg A, Backman L, et al. The perceived threat of the risk for graft rejection and health-related quality of life among organ transplant recipients. J Clin Nurs. 2011;20:274–282.
23. Prasad G, Nash MM, McFarlane PA, et al. Renal transplant recipient attitudes toward steroid use and steroid withdrawal. Clin Transplant. 2003;17:135–139.
24. Boaz A, Morgan M. Working to establish “normality” post-transplant: a qualitative study of kidney transplant patients. Chronic Illn. 2014;10:247–258.
25. Fortnum D, Smolonogov T, Walker R, et al. “My kidneys, my choice, decision aid”: supporting shared decision making. J Ren Care. 2014;41:81–87.
26. Gordon EJ, Butt Z, Jensen SE, et al. Opportunities for shared decision making in kidney transplantation. Am J Transplant. 2013;13:1149–1158.
27. Ross LF, Zenios S, Thistlethwaite JR. Shared decision making in deceased-donor transplantation. Lancet. 2006;368:333–337.
28. Su X, Zenios SA, Chertow GM. Incorporating recipient choice in kidney transplantation. J Am Soc Nephrol. 2004;15:1656–1663.
29. Lancsar E, Louviere J, Donaldson C, et al. Best worst discrete choice experiments in health: methods and an application. Soc Sci Med. 2013;76:74–82.
30. Flynn TN. Valuing citizen and patient preferences in health: recent developments in three types of best-worst scaling. Expert Rev Pharmacoecon Outcomes Res. 2010;10:259–267.
31. Janssen EM, Segal JB, Bridges JF. A framework for instrument development of a choice experiment: an application to type 2 diabetes. Patient. 2016;9:465–479.
32. Howell M, Wong G, Rose J, et al. Eliciting patient preferences, priorities and trade-offs for outcomes following kidney transplantation: a pilot best-worst scaling survey. BMJ. 2016;6:e008163.
33. Bliemer MC, Rose JM. Construction of experimental designs for mixed logit models allowing for correlation across choice observations. Transp Res B: Methodol. 2010;44:720–734.
34. de Bekker-Grob EW, Donkers B, Jonker MF, et al. Sample size requirements for discrete-choice experiments in healthcare: a practical guide. Patient. 2015;8:373–384.
35. Rose JM, Bliemer MC. Sample size requirements for stated choice experiments. Transportation. 2013;40:1021–1041.
36. Bierlaire M. A free package for the estimation of discrete choice models. In: Paper Presented at: 3rd Swiss Transportation Research Conference. Ascona, Switzerland: 2003. http://transp-or.epfl.ch/publications-proceedings.php?List=1.
37. Rose J. Interpreting discrete choice models based on best-worst data: a matter of framing. In: Transportation Research Board 93rd Annual General Meeting. Washington, DC: 2014. https://trid.trb.org/view.aspx?id=1289746.
38. Giergiczny M, Hess S, Dekker F, et al. Testing the consistency (or lack thereof) between choices in best-worst surveys. In: Third International Choice Modelling Conference. Sydney, Australia: 2013. https://www.researchgate.net/publication/269464985_Testing_the_consistency_or_lack_thereof_between_choices_in_best-worst_surveys.
39. de Bekker-Grob EW, Rose JM, Bliemer MC. A closer look at decision and analyst error by including nonlinearities in discrete choice models: implications on willingness-to-pay estimates derived from discrete choice data in healthcare. Pharmacoeconomics. 2013;31:1169–1183.
40. Krinsky I, Robb AL. Three methods for calculating the statistical properties of elasticities: a comparison. Empiric Econ. 1991;16:199–209.
41. Merion RM, Ashby VB, Wolfe RA, et al. Deceased-donor characteristics and the survival benefit of kidney transplantation. JAMA. 2005;294:2726–2733.
42. Pascual J, Zamora J, Galeano C, et al. Steroid avoidance or withdrawal for kidney transplant recipients. Cochrane Database Syst Rev. 2009:CD005632.
43. Webster AC, Lee VW, Chapman JR, et al. Target of rapamycin inhibitors (TOR-I; sirolimus and everolimus) for primary immunosuppression in kidney transplant recipients. Cochrane Database Syst Rev. 2006;19:CD004290.
44. Webster AC, Ruster LP, McGee R, et al. Interleukin 2 receptor antagonists for kidney transplant recipients. Cochrane Database Syst Rev. 2010:CD003897.
45. Webster A, Woodroffe RC, Taylor RS, et al. Tacrolimus versus cyclosporin as primary immunosuppression for kidney transplant recipients. Cochrane Database Syst Rev. 2005:CD003961.
46. Howell M, Yeo R, Tong A, et al. Completeness of reporting of adverse events in trials of maintenance immunosuppression in kidney transplantation: a systematic review. Nephrol Dial Transplant. 2017;32:1586–1594.
47. Rosenbaum L. “Misfearing”—culture, identity, and our perceptions of health risks. N Engl J Med. 2014;370:595–597.
48. Schwartz CE, Sprangers MA. Methodological approaches for assessing response shift in longitudinal health-related quality-of-life research. Soc Sci Med. 1999;48:1531–1548.
49. Peters T, Spinola K, West J, et al. Differences in patient and transplant professional perceptions of immunosuppression-induced cosmetic side effects. Transplantation. 2004;78:537–543.
50. Fagerlin A, Zikmund-Fisher BJ, Ubel PA. Helping patients decide: ten steps to better risk communication. J Natl Cancer Inst. 2011;103:1436–1443.
51. Hausman D. Preferences, value, choice, and welfare. Cambridge, UK: Cambridge University Press; 2012.
52. Murray MA, Bissonnette J, Kryworuchko J, et al. Whose choice is it? Shared decision making in nephrology care. Semin Dial. 2013;26:169–174.
53. Girardi V, Schaedeli F, Marti HP, et al. The willingness of patients to accept an additional mortality risk in order to improve renal graft survival. Kidney Int. 2004;66:375–382.
54. Nizič-Kos T, Ponikvar A, Buturović-Ponikvar J. Reasons for refusing kidney transplantation among chronic dialysis patients. Ther Apher Dial. 2013;17:419–424.
55. Reese PP, Tehrani T, Lim MA, et al. Determinants of the decision to accept a kidney from a donor at increased risk for blood-borne viral infection. Clin J Am Soc Nephrol. 2010;5:917–923.
56. Solomon DA, Rabidou N, Kulkarni S, et al. Accepting a donor kidney: an evaluation of patients' and transplant surgeons' priorities. Clin Transplant. 2011;25:786–793.
57. McDonald S, Clayton P, Hurst K. ANZDATA Registry thirty fifth annual report. http://www.anzdata.org.au/anzdata/AnzdataReport/35thReport/2012_annual_report.pdf. Published 2012.
58. Sautenet B, Tong A, Manera K, et al. Developing consensus-based priority outcome domains for trials in kidney transplantation: a multinational Delphi survey with patients, caregivers and health professionals. Transplantation. 2017;101:1875–1888.
59. Orr A, Willis S, Holmes M, et al. Living with a kidney transplant: a qualitative investigation of quality of life. J Health Psychol. 2007;12:653–662.
60. Rao PS, Schaubel DE, Jia X, et al. Survival on dialysis post–kidney transplant failure: results from the scientific registry of transplant recipients. Am J Kidney Dis. 2007;49:294–300.
61. Delgado P, Diaz F, Gonzalez A, et al. Intolerance syndrome in failed renal allografts: incidence and efficacy of percutaneous embolization. Am J Kidney Dis. 2005;46:339–344.
62. Marcén R, Teruel JL. Patient outcomes after kidney allograft loss. Transplant Rev (Orlando). 2008;22:62–72.

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.