Community-based interventions for older people with complex needs: time to think again?

It is conventional wisdom that primary care for community-dwelling older people needs to shift from vertical, disease-oriented system of care to horizontal, problem-based, person-centred, goal-oriented care, integrating all healthcare providers. Although plausible, there is little evidence that such horizontal care, when embedded into routine practice, is effective in reducing disability and improving function, quality of life or well-being among older people. The negative findings of the ISCOPE trial, published in this issue [1], add to the argument that conventional wisdom may be wrong.

We can learn much about complex, community-based interventions for older people from pioneering work done in general practice in the 1980s [2]. These early studies explored ways to organise anticipatory care for older people, while acknowledging the iatrogenic risks of treating unimportant abnormalities and of medicalising old age. Brief, non-intrusive strategies for predicting functional problems during routine consultations were sought and tested in randomised controlled trials. The preoccupation of doctors with disease to the detriment of its social consequences, the failure to take into account the adaptive powers of older people and the tendency to underestimate the burden borne by carers were all identified as major obstacles to progress in developing more effective primary care for older people. Medical and social problems overlapped in ways that were often puzzling to clinicians, screening led to an increase in referrals to other agencies but without clear evidence of benefit in many instances and variations in referral rates were determined as much by the referrer as by the patient's problems. Finally, at-risk groups proved harder to identify than anticipated, for more pathological events occurred outside the expected at-risk groups than in them.

Early trials used very different interventions and outcome measures [3], complicating their interpretation, but they also had similar outcomes, including: a rise in morale in those enrolled in screening programmes; referrals to all agencies, including medical specialists, tended to increase (in some studies); the duration of in-patient stay fell in some studies, possibly through early intervention in disease processes; in-patient rates could increase through a greater use of respite care; and reductions in mortality did occur in some trials, perhaps for the same reason that in-patient stays declined, but not in all. No trial conducted up to 1990 demonstrated an improvement in older people's functional ability, and general practitioner workload only decreased in situations where the study provided a service that by-passed existing primary care.

During the 1990s, evidence began to emerge that functional ability could be improved, in certain circumstances. For example, RCTs in Italy [4], Canada [5] and the USA [6] measured the impact of case management methods on functioning. The Canadian study reported no significant differences in functional status between those receiving case management and those not, possibly because of the weak link between the nurse case manager and the patients' primary physician. The Italian and American studies reported that those receiving case management experienced less functional decline and remained independent for longer. The Italian study reported that hospital admission was lower in the case management group than in the control group, as were overall costs, and that visits to family doctors increased in the control group and declined in the intervention group. This study included an integrated model of care between primary and secondary care providers, with weekly multidisciplinary reviews of each patient's case. The American study reported that the use of the emergency department, hospital admission rates and healthcare costs were higher in the case management group than in the control group, possibly because the nurse and social work care managers implementing the intervention were acting as advocates for older people.

Trials showing positive effects on functioning were not typical. In the UK, the Medical Research Council trial [7] compared universal versus targeted assessments of the 75 and over population, and management by primary care teams versus multidisciplinary geriatric teams. Its findings suggested that population screening did not increase functional ability in this age group, and that primary care teams and multidisciplinary geriatric assessment produced similar outcomes.

A systematic review of 15 trials of preventive home visits carried out up to 2000 showed no clear evidence of functional improvement [8]. Preventive home visits in Denmark organised through local government did appear to improve older people's functional ability [9] as did nurse-led case management in Spain [10]. The England arm of the ProAge trial (2000–02) showed no change in functioning in a less disabled population [11], but the Swiss arm (with a more intense intervention and longer follow-up) did, including reduced mortality [12]. Reviews of more recent studies have demonstrated some potential to promote independent living, but it is unclear which components of the interventions contribute to effectiveness [13–15].

We appear to be slowly developing methods of working with community-dwelling older people which, when tested experimentally, do seem to alter functional ability, although effect sizes are small. Negative experiences still accumulate, however. Like ISCOPE, a recent primary care intervention aimed at frail older people in the Netherlands failed to show a reduction in disability [16].

One conclusion that we could reach from this experience is that there is much still to be done to identify and engage with those most likely to benefit and to refine and strengthen interventions. Researchers can target the frail and pre-frail, those older people who are ‘hard to reach’ or the middle aged who are ready to change their behaviour and, for example, increase their physical activity. Interventions can be built around different options for physical activity, dietary change or increasing self-efficacy. Such studies will be complex and expensive, and measurement of their costs and benefits will be important.

Alternatively, we could speculate that the combination of well-developed primary care and easily accessible, high-quality specialist services—as found in the UK and the Netherlands—already embody problem-based, goal-oriented approaches to patient care. The development of geriatric medicine services in some countries may have reduced the capacity of new interventions to alter outcomes. Comprehensive, complex interventions to delay disablement that are additions to usual clinical practice may work best in countries with less well-developed and integrated services.

Conflicts of interest

None declared.

References