Up to one and a half million people in Britain are involved in
caring for a relative or friend with mental illness or dementia.
Recent government policy has emphasised the needs of this
particular group of carers, but effective support can be
implemented only if managers and practitioners know what works.
Researchers from the Social Policy Research Unit (SPRU) at the
University of York, working with consultancy Acton Shapiro,
recently looked at services for these carers. There were two
strands: a review of the literature and a consultation
The literature review revealed that there was little evidence to
show that the support services recommended by government as crucial
to carers, such as short breaks, were effective.
A range of approaches to evaluating effectiveness of
interventions also emerged. Nearly three-quarters of the studies
reviewed used standard outcome measures, such as carer burden,
stress, physical health, emotional well-being, depression and
knowledge levels to determine effectiveness. Changes in the scores
before and after an intervention were taken as indicators.
However, there were instances when scores from these outcome
measures did not tally with carers’ views – for instance when a
satisfaction survey showed that carers rated the service in
question highly while standard outcome measures suggested it was
In contrast to researchers’ often narrow approach towards
effectiveness, the professionals and carers participating in the
consultation took a more holistic view, considering benefits for
carers, benefits for the person supported, benefits for the family
as a whole, impacts on service usage and long-term outcomes for
Some contributors believed that the number of referrals or
self-referrals would indicate how effective the service was. But a
telephone helpline, for example, might be effective in reassuring
and supporting carers even if the number using it was low.
Carers felt that the effectiveness of services could be judged
through satisfaction surveys, client evaluation forms,
testimonials, letters or meetings. They emphasised that carers
should be involved in designing the questions to be asked and also
that providers had to respond to carers’ views and modify services
Contributors acknowledged that evaluating the effectiveness of
mental health carer support services could be difficult and
time-consuming. They noted that evaluation was not always a
priority for service providers, who were busy delivering the
service, and that the more integrated a service the more difficult
it was to identify those components which work.
Despite these potential problems, many local authorities are
enthusiastic. One had commissioned an evaluation that included
postal surveys, client interviews and the shadowing of support
workers. Another was compiling a mailing list of carers willing to
be involved in evaluation, or to be trained to interview other
A key message from the study was the importance of using diverse
research methods to increase the depth and breadth of data
collected; and to seek the perspectives of a range of groups,
including carers or families, the cared-for person and
professionals. Other approaches to evaluating services need to be
used with the standard outcome measures, and there is a case for
drawing on the expertise of carers themselves.
Adopting more innovative and inclusive approaches towards
evaluation stands to give policy makers and practitioners a broader
picture of what services and interventions work for carers of
people with mental health problems – and why.
HilraryHilary Arksey is a research fellow, Social Policy
Research Unit, University of York.
H Arksey, L O’Malley, S Baldwin, J Harris, E Newbronner,
P Hare and A Mason, Services to Support Carers of People with
Mental Health Problems: Overview Report/Literature Review
Report, York, 2002, from Social Policy Research Unit,
University of York
E Newbronner and P Hare, Services to Support Carers
of People with Mental Health Problems: Consultation Report,
York, 2002, Social Policy Research Unit, University of
The above documents can be downloaded at