Shared decision making: developing the OPTION scale for measuring patient involvement.

BACKGROUND A systematic review has shown that no measures of the extent to which healthcare professionals involve patients in decisions within clinical consultations exist, despite the increasing interest in the benefits or otherwise of patient participation in these decisions. AIMS To describe the development of a new instrument designed to assess the extent to which practitioners involve patients in decision making processes. DESIGN The OPTION (observing patient involvement) scale was developed and used by two independent raters to assess primary care consultations in order to evaluate its psychometric qualities, validity, and reliability. STUDY SAMPLE 186 audiotaped consultations collected from the routine clinics of 21 general practitioners in the UK. METHOD Item response rates, Cronbach's alpha, and summed and scaled OPTION scores were calculated. Inter-item and item-total correlations were calculated and inter-rater agreements were calculated using Cohen's kappa. Classical inter-rater intraclass correlation coefficients and generalisability theory statistics were used to calculate inter-rater reliability coefficients. Basing the tool development on literature reviews, qualitative studies and consultations with practitioner and patients ensured content validity. Construct validity hypothesis testing was conducted by assessing score variation with respect to patient age, clinical topic "equipoise", sex of practitioner, and success of practitioners at a professional examination. RESULTS The OPTION scale provided reliable scores for detecting differences between groups of consultations in the extent to which patients are involved in decision making processes in consultations. The results justify the use of the scale in further empirical studies. The inter-rater intraclass correlation coefficient (0.62), kappa scores for inter-rater agreement (0.71), and Cronbach's alpha (0.79) were all above acceptable thresholds. Based on a balanced design of five consultations per clinician, the inter-rater reliability generalisability coefficient was 0.68 (two raters) and the intra-rater reliability generalisability coefficient was 0.66. On average, mean practitioner scores were very similar (and low on the overall scale of possible involvement); some practitioner scores had more variation around the mean, indicating that they varied their communication styles to a greater extent than others. CONCLUSIONS Involvement in decision making is a key facet of patient participation in health care and the OPTION scale provides a validated outcome measure for future empirical studies.