• accidental falls;
  • cognitive impairment;
  • dementia;
  • epidemiology;
  • hospital;
  • prevalence

Aims and objectives.  The objective of this study was to compare fall rates in older hospital inpatients with and without cognitive impairment. Relationships between age, gender, mobility, cognitive impairment, care dependency, urinary incontinence and medical disciplines were investigated.

Background.  Falls are common in older people with cognitive impairment, but studies in the hospital setting are rare.

Design.  A secondary analysis of three nationwide prevalence studies in German hospitals from the years 2005, 2006 and 2007 was conducted.

Method.  Trained staff nurses used a standardised instrument to collect data about accidental falls within the last two weeks in their institutions and about other patient characteristics. Data from 9246 patients aged 65 years or older from 37 hospitals were analysed.

Results.  The fall rate for cognitively impaired patients was 12·9%, while only 4·2% of older persons without cognitive impairment experienced a fall. Comparison between medical disciplines showed great differences concerning fall risk for confused and non-confused inpatients. In multivariate logistic regression analysis, the odds-ratio association of cognitive impairment and falls was 2·1 (CI 1·7–2·7). Higher age (OR 1·5, CI 1·2–1·9), greater care dependency (OR 1·6, CI 1·1–2·1), reduced mobility (OR 2·6, CI 1·9–3·7) and being a patient on a geriatric ward (OR 1·8, CI 1·1–2·9) were also statistically significant predictors in this model.

Conclusions.  Cognitively impaired older people constitute a high-risk group for accidental falls in hospitals.

Relevance to clinical practice.  Fall prevention strategies in the hospital setting should address cognitively impaired inpatients as an important high-risk group.