What is integrated care?
Since a robust comparator is essential, but randomisation is not possible, controls are
selected retrospectively from within the large datasets. There has been a tendency for
evaluation of integrated care to focus on `boutique pilots' (indeed, given the current
lack of available evidence, some of the examples within this paper are drawn from such
sources), which has made it difficult to generalise findings to other health and care
settings (Ouwens and others, 2005). The challenge for managers, clinicians and service
users is to propose and support an evaluative component at the outset of any integrated
care initiative. The aim should be to make a robust contribution to the emerging evidence
base for integrated care, encouraging users, planners and decision-makers to become
more directly involved in shaping evaluation, and appreciating where the big gains are
to be made in developing better integrated care.