Background: Approximately 152,000 falls are reported in acute hospitals every year causing significant morbidity and mortality with an estimated cost to the NHS of £2.3 billion per annum. (NPSA, 2007). Existing methods for identifying those at risk are poor, preventing appropriate targeting of resources.
Innovation: We identified seven risk factors for falls in residential home residents1. These were MMSE < 17, impulsive behaviour, poor standing balance, use of a walking frame, fall in the previous year and use of antidepressants and hypnotics/anxiolytics. This pilot study evaluated whether these risk factors remained predictive in the acute hospital setting. Delirium was also assessed.
All new admissions to the Health and Ageing Unit in a one month period were assessed for these risk factors. Routinely collected STRATIFY2 scores were also recorded for comparison. Falls during admission were followed up using hospital reporting systems and electronic patient records.
Evaluation: After excluding re-attenders, data was collected on 69 consecutive admissions with 16 fallers identified. Only impulsivity was a statistically significant determinant of falling in multivariate analysis (OR 7.29 95% CI 1.62-32.79). This remained significant when adjusted for length of stay. Compared to STRATIFY, the care home tool had a greater area under the curve (0.61 (95%CI 0.44-0.78) vs. 0.81 (95%CI 0.67-0.96)).
Conclusion: This pilot study found that the care home tool was more effective at identifying fallers than STRATIFY. This provides proof of principle for a larger study, powered to demonstrate the superiority of the new tool over existing tools and to detect whether any differences in the risk factors collected other than impulsivity exist.