The majority of hospital in-patients are older people, and many of these are at increased risk of readmission, which can be an adverse outcome for the patient. Currently there is poor understanding as to how best to reduce the risk of readmission. We searched MEDLINE, EMBASE and the Cochrane library for high quality review articles about readmissions. Each review was quality assessed by two reviewers. Grouped data and evidence from original papers is cited with 95% confidence intervals when possible. Nine review studies of sufficient quality were included. Two addressed risk factors for readmission, which included: age, poor functional status prior to admission, length of stay during the index admission, depression, cognitive impairment, malnutrition, social support and social networks/support. The seven other reviews addressed interventions to reduce readmission, which included: discharge planning, post-discharge support, post-discharge case management, and nutritional supplementation. It is possible to identify older people at risk of readmission using well-established risk factors; discharge planning, post-discharge support and nutritional interventions appear to be effective in reducing readmission. Combined interventions appear to be more effective than isolated interventions.

The degree of social support available to older persons who are institutionalised is under-researched. This study investigated the structural and functional support exchanges with their social network members of 72 nursing home residents in Hong Kong (58 women, 14 men). They were asked to identify their network members, to evaluate the degree to which each one was important in their lives, and to rate the support received from and provided to each individual. The participants reported few network members (average 2.6) and in many cases neither a spouse nor children were included. Only one-fifth of the participants reported a social network member in the nursing home, and most of those who did nominated a member of the staff. There were also few friends in their networks. On the whole, the participants were comparatively socially isolated. The findings were explained in terms of the shame associated by the Chinese with placement in an institutional home, cultural patterns of social support, changes in children's filial attitudes, home placement policies, and the management practices that accentuate the distance between the older person and family members around the time of institutionalisation. These inculcate a feeling of abandonment, and discourage family visits as well as social interactions within the home.

Population ageing and constraints on public sector spending for older people with long-term health problems have led policy makers to turn to the social networks of older people, or the ‘informal sector’, as a source of long-term care. An important question arising from this policy shift is whether these social networks have the resources to sustain the high levels of care that can be required by older people with chronic health problems. In the face of both dire warnings about the imminent demise of the informal sector, and concurrent expectations that it will be the pillar of community long-term care, it is timely to undertake a critical analysis of the caring capacity of older people's social networks. In this paper we argue that the best way to understand the caring capacity of informal networks of frail older people is to establish their membership and caring capacity. It is useful to make conceptual distinctions between ‘social’, ‘support’, and ‘care-giving’ networks. We argue that transitions of networks from social through support to care roles are likely to show systematic patterns, and that at each transition the networks tend to contract as the more narrowly defined functions prevail. A focus on ‘care networks’, rather than the more usual ‘care dyads’, will move forward our understanding of the caring capacity of the informal sector, and also our ability to forge sound social and health policies to support those who provide care.