AN OBSERVATION AUDIT TOOL FOR NURSE-TO-NURSE BEDSIDE CLINICAL HANDOVER

Objectives: 1. Examine the reliability, validity and usability of a multi-purpose tool for training, coaching and auditing nurse-to-nurse bedside handover across an organisation and. 2. Evaluate the quality of nurse-to-nurse bedside handover practices. Methods: A naturalistic, descriptive, mixed method study conducted in three stages: Stage 1- face and content validity of an existing tool were examined using literature review and focus group with five local nursing and midwifery experts. Stage 2- the tool was revised and pilot tested; the final 24-criterion tool included 52 items providing illustrative exemplars of expected handover behaviours related to content, process and environment aspects of handover practice, evaluated using a 3-point scale. Stage 3- an observation audit of 199 ‘real’ handover events across 5 wards; in 72 handover events, independent second observers were used to examine tool reliability; and unstructured field notes captured data on tool usability. Descriptive statistics were used to analyse audit data; the Kappa statistic was used to measure inter-observer agreement to examine reliability. Qualitative description was used for qualitative analyses. Results: Content validity was supported by both literature and expert review. Whilst also supported, face validity and usability were impacted by the volume of items and the relevance of the exemplar behaviours to the local context of practice. Use of Kappa statistic (K) to measure inter-observer agreement revealed 30 illustrative items were acceptable with ‘moderate’ or ‘good’ agreement (K 0.41 or higher); 14 items had ‘fair’ agreement (K 0.40 to 0.00); and of the remaining items, 4 had poor inter-observer agreement (K < 0.00), and K could not be calculated for 4 items. Interobserver agreement was acceptable for 70.5% (n=11) of content items, 58.8% (n=10) of process items and 44.4% (n=8) of environment items. The organisation set a benchmark for illustrative behaviours to be observed in 80% of handovers. Analyses revealed only 11.7% (n=2) behaviours related to handover content, 17.6% (n=3) behaviours related to handover process, and no behaviours related to environment were observed in more than 80% of handovers. Conclusion: This research provides a model and tool for ongoing quality improvement that addresses the three key components of nurse handover practice: content, process and environment. The content and face validity of the nurse-to-nurse bedside handover tool was acceptable, however use of context specific exemplars could improve face validity and usability, particularly when evaluating behaviours related to handover processes and environmental safety. Reducing the volume of exemplar behaviours may improve inter-observer agreement. Findings reveal a need to improve understanding about the scope of handover practice, particularly the recognition of process and environment influences on patient safety. Future research is needed to validate the tool across a wide range of handover settings. Measuring nurse handover practices in the ‘real world’ is complex. This research revealed limited understanding of the scope of nurse handover practice, and provides a tool to facilitate education, coaching and quality audit.