Defining appropriateness of emergency department attendance: a New Zealand perspective

Abstract

Aim: To explore NZ emergency medical and nursing staff understandings of Emergency Department (ED) 'appropriateness‘ in relation to patient presentations. Background: This study was developed in order to contribute to the knowledge and practice responses around ED overcrowding. Overcrowding of hospitals EDs is an international problem, and one that has given rise to a number of policy and practice interventions. A commonly cited contributing factor is the impact of 'inappropriate‘ or non urgent patients on the ED capacity and management of patient flow. Suggestions have been made that this group could be identified and potentially re-directed or managed in other ways, thus reducing the burden on the ED workforce and resources. What is often overlooked, however, is the need to first identify what is meant by the term 'inappropriate‘. Much of the research and recommendations currently available is based on the assumption that this definition is already known and agreed on. This thesis disputes that, and seeks to explore the degree of agreement possible within a group of expert NZ emergency care clinicians. Research Design: This study was positioned within a Mixed Methods approach utilising a consensus research technique, the Delphi process, to gather data. Consideration in choosing the research design was to identify a process that would maximise the opportunity to formulate a consensus definition or set of criteria which could potentially be applied to the practice setting. A three round Delphi survey was used, including rated statements, and free text responses. Thematic and simple descriptive statistical analysis was used in evaluating the findings from this study. Participants: An expert panel was established, consisting of 59 peer nominated clinical experts from ten tertiary level NZ EDs. This included 23 physicians and 36 nurses working in EDs with an average patient throughput of 30,000 or greater per annum. Results: Fifty nine individuals commenced the study, with 49 completing all three rounds of the Delphi, providing a retention rate of 84%. A NZ specific consensus (set at 70%) generated 18 conditions, which if the only cause for presentation in an otherwise well individual were found to be indicative of 'inappropriateness‘ for ED presentation. Agreement was not reached in regard to a sample international tool, or to a number of other conditions and situations presented in international studies. The NZ criteria were more specific than those suggested in the wider literature, and even where consensus was reached, there was evidence of 'outlier‘ opinions with strong viewpoints. The NZ experts asked for additional contextual information in relation to clinical assessment and eight additional non clinical factors were identified as impacting on decision making and assessment of appropriateness. Conclusion: Following application of a rigorous consensus process, limited agreement regarding appropriateness of patient presentations was reached by a NZ panel of expert clinicians. The NZ experts did not endorse the internationally derived definitions or criteria that were presented. There were several concerns expressed; including the ability to apply clinical criteria in the absence of social context, and how much relative weighting each of these aspects should receive. The essential subjectivity associated with some definitions was acknowledged and whether these could be applied prospectively was questioned. Recommendations were identified for clinical practice, policy development and further research. In particular, the importance of acknowledging existing assumptions, the need to question the evidential base for these, and the risks associated with failing to clarify the NZ perspective.