Methods to evaluate services.

EDITORIAL Methods to evaluate services The health care of elderly people takes place in a variety of settings: primary; secondary; and tertiary care, and now 'intermediate care'. It covers acute and chronic conditions, as well as preventative, curative and restorative and palliative services, which can be simple or complex. These need to be evaluated to help purchasers, providers and patients make informed decisions. Epistemology is the science of the methods of research. Health Services Research (HSR) has an epistemological problem in deciding what sort of knowledge is needed and how to find it out. The randomised controlled trial (RCT) is an evaluative method that is well established for single interventions, and used widely in complex interventions. But this method is criticised for not reflecting the real situations of health care delivery. Defenders of this approach would argue that well designed pragmatic trials obviate much of this criticism. But there are some important interventions that cannot or will not be randomised, such as national health and economic policies. There is growing interest in alternative methods to evaluate services, but no obvious consensus about the best way to do it. It is not just a problem with methods. There are absurd barriers to the evaluation of services. For example, services can be established apparently by political whim without any necessity for them to be evaluated, whereas no such laxity would be permitted for a new drug. At the same time, when evaluation is planned and submitted to an ethics committee, it may find restrictions imposed that distorts the research, such as complex consent procedures for simple surveys. Systematically collecting information (the basic process of research, whatever the method) can appear burdensome in the NHS, which already collects vast amounts of routine information but not in a way that allows it to be used for evaluation. Since the methods for HSR are unclear, it is therefore difficult to obtain funding for research. A preference for systematic reviews means that it is unclear how to consider the findings of studies using methods other than the RCT. As a result, evidence-based decisions are not easy, and so the planning of services for the elderly may suffer. These research problems may threaten the development and standards of services for older people. And if all that wasn't enough, it is also a concern that services, and the climate in which they operate, change over time, therefore there …