The use of fall prevention guidelines in German hospitals – a multilevel analysis Kathrin Raeder MA RN, 1 Ute Siegmund MA RN, 2 Ulrike Grittner PhD, 3 Theo Dassen PhD RN 4 and Cornelia Heinze PhD RN 5 1 Student, 2 Master Student, 4 Professor, 5 Assistant Professor, Centre for the Humanities and Health Sciences, Department of Nursing Science, Charité – Universitätsmedizin Berlin, Berlin, Germany 3 Statistician, Charité Centre 4 – Therapy and Research, Department of Biometrics and Clinical Epidemiology, Charité – Universitätsmedizin Berlin, Berlin, Germany Keywords fall, fall-related injuries, guideline, hospital, multivariate analysis, prevention Correspondence Cornelia Heinze Centre for the Humanities and Health Sciences Department of Nursing Science Charité – Universitätsmedizin Berlin Charitéplatz 1, 10117 Berlin Germany E-mail: cornelia.heinze@charite.de Accepted for publication: 20 November 2008 doi:10.1111/j.1365-2753.2009.01143.x Abstract Aims Falls and fall-related injuries are major problems in hospitals. The aim of this study was to examine the impact of fall prevention guidelines on falls and fall-related injuries in hospitals. Method A cross-sectional study was conducted in German hospitals. Some 28 hospitals participated with a total of 5046 patients. Eleven of these hospitals had already imple- mented a fall prevention guideline, 10 were in the process of developing such a guideline and seven hospitals were not using any fall prevention guideline at all. A standardized questionnaire was used on the individual patient level to obtain details regarding the socio-demographic background, falls, fall-related injuries and other problems relevant to nursing. A further questionnaire referred to the use of fall prevention guidelines in the individual hospitals. Data specific to falls were analysed both on hospital level and on ward and patient level by means of a multilevel logistic model. Results The univariate analyses suggest that patients in hospitals that are using guidelines are more likely to fall [odds ratios (OR) = 1.19, confidential interval (CI) = 0.65–2.18] than in hospitals that do not use any guideline (reference category) or are still in the developing stage (OR = 0.82, KI = 0.77–0.87). If, in a multivariate analysis, the ward level and indi- vidual patient variables (age, disorientation, confusion, incontinence) are included, the following results are obtained: the probability of falls in hospitals not using guidelines (reference category) is higher than in hospitals developing a guideline (OR = 0.86, KI = 0.58–1.28) or using a guideline (OR = 0.71, KI = 0.44–1.14). The differences are even more distinct regarding the injuries resulting from a fall that require medical treatment. The probability of these injuries is significantly lower in hospitals using guidelines (OR = 0.27, KI = 0.09–0.85) than in hospitals developing a guideline (OR = 0.61, KI = 0.24–1.54) or not using any guidelines at all (reference category). Conclusion The present results of the multilevel analysis show that falls and fall-related injuries can be reduced by the implementation of fall prevention guidelines. Introduction A fall is a significant patient problem, especially in elder patients. Apart from pain, injuries and restrictions in mobility falls also have psychological consequences, for instance social isolation, depression or fear of falling [1]. There is no evidence that using restraints and side rails, which were historically important in fall prevention, can prevent falls, may result in injuries or even lead to death [2–4]. Another problem that should not be underestimated is the fear of a recurring fall [5]. The frequency of falls differs significantly between hospitals. International figures show that they range 1.4–17.9 falls (particular in geriatric hospitals) per 1000 patient days (K. Raeder & U. Siegmund, unpublished) [6,7]. For quite some time, international discussions on quality in nursing care are focusing on the development of guidelines for the implementation as quality assessment instruments [8–10]. Since the 1990s, falls have received increased attention in nursing research. The World Health Organisation, for example, published a report on the problem of falls in 2004, and the Organisation for Economic Co-operation and Development included falls in its list of quality indicators [11,12]. Other internationally recognized associations are dealing with this subject such as the American Journal of Evaluation in Clinical Practice ISSN 1356-1294 © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd, Journal of Evaluation in Clinical Practice 16 (2010) 464–469 464