Adaption and validation of the Working Alliance Inventory for General Practice: qualitative review and cross-sectional surveys.

Background: Relational aspects of primary care are important, but we have no standard measure for assessment. The 'working alliance' incorporates elements of the therapeutic relationship, shared decision-making, goal setting and communication skills. The Working Alliance Inventory (short form) (WAI-SF) has been used in adult psychology, and a high score on the survey is associated with improved outcomes for clients. Objective: To adapt the WAI-SF for use between GPs and patients and to test its concurrent validity with measures of shared decision-making and the doctor-patient relationship and discriminant validity with measures of social desirability. Methods: Two rounds of online survey feedback from 55 GPs and 47 patients were used to adapt the WAI-SF - the WAI-GP. The tool was then completed by 142 patients in waiting rooms after seeing their GP and by 16 GPs at the end of their session. Concurrent validity with measures of shared decision-making and patient-doctor depth of relationship was determined using Spearman Rho correlations. Patients also completed two social desirability surveys, and discriminant validity with WAI-GP was assessed. Results: Following feedback, the survey was re-worded to remove phrases that were perceived as judgmental or irrelevant. The patient measure of the WAI-GP was strongly correlated with Dyadic OPTION (rho = 0.705, P = 0.0001) and Patient-Doctor Depth of Relationship scale (rho = 0.591, P = 0.0001) and not with measures of social desirability. Conclusion: The psychometric properties of the WAI-GP support its use for measuring GP-patient alliance. Possibilities for use include assessing the influence of therapeutic alliance on the effectiveness of interventions.