Reducing hospital admissions from nursing homes: a systematic review

Longevity, chronic illness, frailty and deficits in activities of daily living are common characteristics of the geriatric nursing home population. These features are predispositions to a trajectory of health with acute incidences which raises the question about acute care hospitalisation. Acute flares in nursing home residents’ health may call for services not necessarily available in the nursing homes, such as diagnostic procedures, particular interventions or a shift towards end-of-life care. Indeed, studies from a range of different countries with well-developed nursing home sectors have demonstrated that acute hospital admissions occur commonly, with annual rates from 9% up to 60% [1‐7]. Noteworthy, large variations in hospital admission rates from nursing homes are not only observed between countries, but also within countries and in small geographic areas [1, 8, 9].

Adding to this picture, a number of studies have pointed to the detrimental impacts that hospitalisations may have on elderly people, including iatrogenic illnesses, like infections to functional and cognitive decline [10‐16]. Additionally, the nursing home population is appointed to account for many potentially unnecessary hospitalisations, with estimates between 19–67% [17‐20]. As such, a reduction of hospital admissions among nursing home residents may potentially serve a dual benefit of improving care for residents, as well as reducing use and monetary cost of specialist health care.

Although it is strongly communicated that nursing home residents represent an overuse of specialist services [17‐20] it is not clear what strategies can best substitute hospitalisations. Thus, enforced by healthcare reforms that warrant for a shift in the provision of health care from specialist to primary care settings, there is an increasing interest for care models that can replace frequent and perhaps unnecessary use of hospital admissions from nursing homes [21, 22]. Still, it is not clear what strategies can best substitute hospitalisations, to achieve the twofold aim of providing high quality services and reducing cost in specialist health care.

The objective of this systematic review is therefore to summarise the effects of interventions to reduce acute hospitalisations from nursing homes.

This is an update of a systematic review published in Norwegian by the Norwegian Knowledge Centre for the Health Services [23]. A protocol for the first version, including eligibility criteria, search strategy and methods of analysis, was developed in advance and made available in PROSPERO [24].

We set out to systematically review the effects of interventions to reduce acute hospital admissions from nursing homes. Four systematic reviews and five primary studies were included, evaluating a total of eleven different interventions. Overall, using GRADE, the quality of the evidence for all outcomes was low or very low. In systematic reviews, the quality of evidence reflects the extent of confidence that an estimate of effect is correct [37]. As such, our confidence in the findings is weak. Still, we believe that this review is an important contribution as the first truly systematic and transparent approach to the topic. Further, several of the included studies showed promising effects on hospital admission, but were downgraded, in many cases because of the relatively few included patients. Among the seven interventions to structure or standardise treatment, a reduction in hospital admissions was found for four of them. This was the case for two out of three advance care planning interventions, one intervention to enhance the use of palliative care services and one where a care pathway for lower respiratory tract infections was tested. For the three remaining interventions in this category; an ACP-intervention involving social workers, one multifaceted implementation of a national guideline for the treatment of pneumonia and a care protocol for residents with COPD, a statistical significant difference in hospitalisation between the intervention and control group was not found. Two single studies tested geriatric specialist services, both involving flexible and add-on special competence and human resources to the care in nursing homes. Both of these reported fewer hospitalisations in favour of the intervention. Two Cochrane reviews respectively tested influenza vaccination among residents and health personnel. The case of vaccinating residents, although many studies were identified, only observational design studies were found, making it infeasible to draw conclusive inferences from the findings. Also, noteworthy, all of the studies failed to show an effect on laboratory-confirmed influenza, raising serious doubt in the inherent conceptual mechanism of the intervention. Further, it is not clear whether promoting influenza vaccination among health personnel makes a difference on hospitalisations of nursing home residents.

Few evaluations are conducted on the effects of interventions to reduce hospital admissions from nursing homes. Eleven evaluated interventions were identified, but none were tested more than once with a rigorous study design. Although the quality of evidence was low for all comparisons in this review, some of the interventions had effects on reducing hospital admissions. These interventions, such as advance care planning, palliative care, care pathways and geriatric specialist services, may represent important aspects of nursing home care to reduce hospital admissions and should be studied further. Our findings suggest an evidence-policy gap, where current policies and practices are lacking evidence-based management strategies to underpin them.