Using Best–Worst Scaling Survey to Investigate the Relative Importance of Attributes Associated with Public Hospital Outpatient Appointments

Introduction Obtaining patient input before healthcare redesign improves patient experience. The Townsville Hospital and Health Service, a regional Australian public health service, seeks to reduce the long wait list for medical specialist appointments by introducing allied health substitution models of care for low-acuity patients. This paper describes a best worst scaling survey conducted to refine attributes associated with outpatient appointments which will be used in a future discrete choice experiment (DCE). Methods A literature review was conducted to identify attributes associated with medical specialist outpatient appointments and allied health substitution models. An object (or case 1) best worst scaling (BWS) survey was designed using blocks of a balanced incomplete block design and analysed using multinomial logit and mixed logit models. Patients waiting at local specialist outpatient clinics were invited to complete the survey via an iPad. The interviewer collected field notes, which were analysed using content analysis. Results A total of 12 attributes were identified in the literature review and one from local discussion. The 167 completed responses demonstrated the ranking of attributes were diagnostic accuracy, symptom relief, continuity of care, satisfaction with care, healthcare professional, manner and communication, time on waitlist and onward referral. The least important attributes were reassurance offered, appointment wait time, cost and appointment duration. Conclusions This BWS survey allows us to reduce the attributes for inclusion in the DCE from 13 to 8. Diagnostic accuracy and symptom relief were of most importance, and appointment wait time and duration were of least importance. This suggests that patients would be willing to be attend different models of care such as allied health primary contact model if clinical outcomes were equivalent to the current medical-led models.