Acute hospital admissions from nursing homes. Rates and characteristics; unwarranted variation and effects of interventions to reduce them

Abstract

Background: The geriatric nursing home population is vulnerable to acute incidences in their health condition, due to longevity, multiple chronic illnesses and a low level of physical and mental function. Aims: The overall aims of this study were to describe acute hospital admissions from nursing homes in Norway; investigate and explain the variation in admission rates between nursing homes, and systematically review the effects of interventions to reduce hospitalisations from nursing homes. Methods: Population-based observational data were used for papers I and II, with electronic patient journals as the main source of data. For two consecutive years (01.01.2007-31.12.2008), acute admissions from 38 nursing homes were identified through ambulance records (n=2 451). Paper I reports annual incidence rates of hospitalisation, describe characteristics of hospital stays and reports mortality. Paper II describes variation in hospitalisation between all nursing homes in one municipality and explores explanatory associations. Paper III is a systematic review of the effects of interventions to reduce hospitalisations from nursing homes. Results: The mean annual hospital admission incidence rate was 0.62 admissions per person-year. Of all admissions 55% were related to falls and respiratory and circulatory diseases. Length-of-stay was median 3 days. In-hospital and 30-days mortality rates were high; 16% and 29%, respectively. A nine-fold variation in admission rates were found among the nursing homes. Annual rates correlated significantly with size and percentage short-term beds of the institutions, but no associations were found between rate and ownership, suburb or age of the residents. There was a strong correlation between nursing homes’ admission rates two years in a row. The literature search generated 6 250 individual records. Four systematic reviews of high quality and five primary studies were included, evaluating 11 different interventions. The overall quality of the evidence was graded low or very low. Conclusions: Nursing home residents are hospitalised at a very high rate in this setting, and a more than nine-fold variation suggests the presence of unwarranted variation. Some interventions to reduce hospitalisations show promising results, but the evidence is of too low quality to provide clear recommendations.