[Benchmarken is 'work in progress']

In the March edition of this journal a group of professors criticised the methods adopted by the SBG (Benchmark foundation of the Dutch Mental Health Services). The critical comments relate to the alleged objectives ('pay-for-performance' in the Dutch Mental Health Services) and the methods involving confounding, bias and the use of insufficiently sensitive instruments.

To provide more detailed information about the SBG and the role of the scientific council (WR) and about the social context within which rom and benchmarking are developing, and to give our views on the methodological criticisms.

We elucidate the methods of the SBG and the role of the WR and we clarify the ways in which ROM and benchmarking are currently being used.

We recognise that there are methodological problems. The wr too warns against jumping to conclusions on the basis of data whose validity is still under discussion. The Dutch Mental Health Services have practically no experience of rom and benchmarking. The Dutch Mental Health Services need to gain more experience in this field. On the other hand, it is true that rom data and comparisons between institutions can lead to improvements in the standard of care, provided these procedures are carried out carefully. In the SGB methods the aforementioned problems have been removed, at least partially. The SBG is concerned not only with symptom reduction but also with the functioning of the individual and with his or her quality of life.

The WR believes that the Dutch Mental Health Services should operate a scientifically validated benchmark system. Sofar, there has been hardly any research into whether benchmark data really say anything about quality and/or whether comparisons between institutions are valid. A national ROM-system is needed. Current initiatives are creating ways in which these matters can be investigated, preferably in conjunction with scientists and professionals.