The psychometric properties of Observer OPTION5, an observer measure of shared decision making

Abstract

ObjectivesObserver OPTION5 was designed as a more efficient version of OPTION12, the most commonly used measure of shared decision making (SDM). The current paper assesses the psychometric properties of OPTION5. MethodsTwo raters used OPTION5 to rate recordings of clinical encounters from two previous patient decision aid (PDA) trials (n=201; n=110). A subsample was re-rated two weeks later. We assessed discriminative validity, inter-rater reliability, intra-rater reliability, and concurrent validity. ResultsOPTION5 demonstrated discriminative validity, with increases in SDM between usual care and PDA arms. OPTION5 also demonstrated concurrent validity with OPTION12, r=0.61 (95%CI 0.54, 0.68) and intra-rater reliability, r=0.93 (0.83, 0.97). The mean difference in rater score was 8.89 (95% Credibility Interval, 7.5, 10.3), with intraclass correlation (ICC) of 0.67 (95% Credibility Interval, 0.51, 0.91) for the accuracy of rater scores and 0.70 (95% Credibility Interval, 0.56, 0.94) for the consistency of rater scores across encounters, indicating good inter-rater reliability. Raters reported lower cognitive burden when using OPTION5 compared to OPTION12. ConclusionsOPTION5 is a brief, theoretically grounded observer measure of SDM with promising psychometric properties in this sample and low burden on raters. Practice implicationsOPTION5 has potential to provide reliable, valid assessment of SDM in clinical encounters.