ABSTRACT
This study examined how lifestyle factors and gender affect kidney allocation to transplant patients by 99 British and Singaporean participants. Thirty hypothetical patients were generated from a combination of six factors (alcohol intake, smoking frequency, weight, exercise frequency, diet, and gender) and randomly paired four times. Participants saw 60 patient pairings and, in each pair, chose which patient would receive treatment priority. A Bradley-Terry model was used to derive coefficients for each factor per participant. A mean factor score (MFS) was then calculated across all participants for each factor. Participants gave lower priority to patients who drank more, were overweight, smoked more and exercised less. A patient’s diet and gender had no significant effect on allocation. There were no significant cross-cultural differences. There were moderate correlations between participants’ self-reported pre- and post-experiment ordering of decision criteria, and these measures and factor coefficients, suggesting a modest level of decision-making consistency. Between participants, moderate levels of concordance with respect to factor importance were observed for self-reported orderings of factors, and weaker agreement for model-derived coefficients. Very similar results were obtained for both British and Singaporean participants, and the implications of the findings are discussed.
DATA AVAILABILITY
This is obtainable from the last author upon request
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author(s).
ETHICS
This was sought and obtained (UCL EP/2018/007)
REGISTRATION
This paper was not pre-registered with the journal
Notes
1 In 2017, the WHO age-standardized estimated prevalence of tobacco and cigarette smoking among those aged 15 years or more was 16.5% and 14.8% respectively in Singapore, compared to 19.8% and 17.5% respectively in the UK (WHO, 2019).
2 In 2016, alcohol per capita consumption (of pure alcohol) was 2.0 liters in Singapore, compared to 11.4 liters in the UK. The prevalence of alcohol use disorders and alcohol dependence was 1.1% and 0.5% respectively in Singapore, compared to 8.7% and 1.4% respectively in the UK (WHO, 2018).
3 Further participant demographic information collected is available from the first author on request.
4 There was a high correlation between the Bradley-Terry model derived factor coefficient estimates and their z-values for each of the six factors across all participants. The correlations are shown in Appendix B.
5 There are only 95 participants as four participants gave all six factors the same ratings and/or ranking and thus a Kendall’s correlation coefficient could not be calculated.