Introduction: Falls are the most commonly reported patient-safety incident in mental health settings for older people with approximately 36,000 falls reported from these settings per year. Risk of falling is exacerbated by mental health problems, such as impaired mental status due to dementia and depression, and their treatments. However, we have very little understanding of either the context or best way to prevent falls in these settings. This paper will present phase 1 findings from a mixed-methods exploratory case study.
Methods: Methods included retrospective analysis of reported falls for a 12 month period, non-participatory observation (300+ hours), focus-groups with staff (n=5) and interviews with senior managers (n=6).
Findings: We will use the data to explore the temporal and spatial nature of falls within 5 acute mental health settings. We found falls were not evenly distributed across the course of the day, with peaks occurring in the morning, and early afternoon. Staff reflections on this data during focus group discussions, and our observations of the temporally-determined flow of activity and use of space by staff, patients and visitors to the ward have enabled us to understand some of the patterns and possible factors that contribute to falls in these settings.
Conclusions: Use of routinely collected data, enhanced by staff reflection and observation can aid understanding of factors contributing to falls in acute mental health settings. Data of this nature could be used by organisations seeking to manage risk, improve patient safety, and develop appropriate interventions.
Original languageEnglish
Publication statusPublished - 12 Jul 2012
EventBritish Society of Gerontology - Keele, United Kingdom
Duration: 11 Jul 201213 Jul 2012


ConferenceBritish Society of Gerontology
Country/TerritoryUnited Kingdom


  • Older people
  • falls prevention


Dive into the research topics of 'The temporal and spatial nature of falls in acute mental health settings'. Together they form a unique fingerprint.

Cite this