• respiratory infection

Bu et al 1 report their findings from using data collected by an established longitudinal study of older people (the English Longitudinal Study of Ageing, ELSA), linked to administrative information, to analyse loneliness, social disengagement and isolation as risk factors for hospital admissions related to respiratory disease. Both survival and sensitivity analyses were used on a sample of c. 4500 respondents to conclude that living alone and social disengagement are risk factors for such admissions, but there was no significant statistical evidence to indicate that low social contact or loneliness are.

Previous research, including the ELSA wave 5 report,2 has acknowledged ‘… strong cross-sectional associations between psychological well-being and health, particularly in relation to chronic illness and disability, although with variations across different aspects of well-being.’ These findings mirror those from Bu et al, acknowledging that both physical and psychological health have multiple domains and determinants. The wave 5 report included findings on chronic lung disease among other long-standing limiting illnesses, whereas the study’s wave 4 report3 looked at the relationship between cardiovascular disease and loneliness, along with other measures of well-being.

Bu et al recognise that although ELSA is a rich and diverse source of information, it does not provide the necessary data alone to facilitate detailed analysis on specific respiratory conditions and the associated use of hospital services. Linking to administrative mortality and hospital records provides much more depth to the data, which could (in theory) be developed further by linking to primary care data (potentially from the Clinical Practice Research Datalink), medications and prescription data (ditto) and Accident and Emergency admissions (from Hospital Episode Statistics, HES). However, using linked data inevitably means that the respondents who did not consent to data linkage (1040 respondents from the core ELSA sample of 8310) cannot be used in the subsequent analysis and, for HES data, a further loss of those for whose hospital records are incomplete or cannot be linked. All of these ‘losses’ of respondents affect the extent to which the findings can be generalised across all older people in England. Also, this is an observational study and so although a range of potential confounding factors have been taken into account, causality cannot be assumed. There may be further residual confounding as social engagement (being more active than social contact) is arguably determined by both the availability of relevant activities and the physical ability of the respondent to participate.

Indeed, one of the authors of this paper has recently written4 on the outcomes of the British Lung Foundation Patient Passport. They write that ‘pulmonary rehabilitation (PR) is one of the highest value chronic obstructive pulmonary disease (COPD) interventions. Referral rates to and delivery of PR remain inadequate, both according to national audit data and in the present patient survey’; only 34% of respondents said they had discussed PR with their care provider. There is evidently scope for improvement here both from a clinical point of view but also in terms of risk factors for hospital admissions, as PR is in itself a form of social engagement.