Decisional equipoise is not decisional conflict: avoiding the false clarity bias in the evaluation of decision aids and Shared Decision Making processes.

摘要

Purposes: To question the use of criteria related to decisional ‘sureness’ in evaluating decision aids and Shared Decision Making processes,- as occurs in the Decisional Conflict Scale (items 10-12) and its reduced form SURE (item 1) - on the ground that decisional equipoise is a legitimate outcome that should not affect an evaluation of decision quality or decision support. To investigate the empirical extent of equipoise in a decision aid trial and explore the relationship between degree of equipoise and decision quality as self-assessed by MyDecisionQuality, a dually-personalised instrument that does not include such ‘sureness’ criteria.Methods: In the fixed criteria arm of an Australian trial of decision aids relating to PSA testing for prostate cancer 727 respondents supplied their importance weights for 5 criteria (loss of lifetime, needless biopsy, and bowel, urinary and sexual problems). Combining these criteria weights with the best available evidence on the performance of the two options (have or not have a PSA test) on each criterion, produced a personalised expected value score for each option on a 0 to 1 scale. The Absolute Difference (AD) between the two option scores for each individual was calculated and those whose AD was .01 or less were defined as being in effective equipoise. These ADs were then correlated with respondent’s scores on the self-weighted and self-rated instrument, MyDecisionQuality, which contains 8 items (Options, Effects, Importance, Chances, Trust, Support, Control, Commitment). Results: For 64 % of the 727 respondents the Absolute Difference was less than .01, leaving them in effective equipoise. For 52% it was under .005. The correlation between AD and MDQ score was .06, showing that there was no influence of the degree of equipoise on the self-reported assessment of decision quality.Conclusion: Evaluations of decision aids and Shared Decision Making processes, indeed any decision making process, should accept decisional equipoise is a possible and legitimate outcome, even after full and unbiased processing of evidence and preferences. The instruments used should not reward unwarranted ‘sureness’ or ‘decisional conflict reduction’, which potentially lead to a ‘false clarity’ bias.