Falls and delirium in older inpatients: Work-as-imagined, work-as-done and preferences for clinical decision support systems

Abstract Background Falls and delirium are common complications in older inpatients. Implementation of fall and delirium guidelines is complex and may be facilitated by Clinical Decision Support Systems (CDSSs). Our study aimed to understand the differences between guidelines (work-as-imagined) and actual care (work-as-done) and how these can impact the design of a CDSS. Methods We used Functional Resonance Analysis Method (FRAM) as structured method to visualize work-as-imagined and work-as-done and to develop an initial CDSS design. As input for work-as-imagined, we used national guidelines. To identify work-as-done and CDSS preferences, we conducted semi-structured interviews in two rounds. We identified differences between guidelines and actual care by creating a FRAM model for work-as-imagined and one for work-as-done. CDSS preferences were coupled to activities in the FRAM models and to the identified barriers and facilitators for following guidelines. Results Eighteen clinicians participated in 24 individual and small-group interviews. For both falls and delirium, we found substantial differences between work-as-imagined and work-as-done. Several CDSS opportunities to improve fall and delirium care were identified including reminders for screening, an order set to facilitate risk assessment, automatically-generated advice with personalized preventive interventions and support to facilitate medication reviews. The required CDSSs addressed both barriers and facilitators. Conclusion In our study, work-as-done for falls and delirium differed substantially from work-as-imagined and the preferred CDSSs would address both barriers and facilitators. Furthermore, our study showed that FRAM is a suitable tool to identify differences between actual care and guidelines and to design CDSSs from a user-centred approach.