In reply: The fundamental purpose of clinical practice guidelines is to improve patient outcomes. Thus, as members of the CARI (Caring for Australians with Renal Insufficiency) Guidelines Steering Committee, we welcome Johnson's letter, which gives us reassurance and renewed enthusiasm to move forward with improving and refining the CARI clinical practice guideline process. The most gratifying revelation in Johnson's survey was the near-90% endorsement of the CARI guidelines as a document providing a useful evidence summary — clearly very reassuring in those areas in which that evidence relates to treatment interventions for patients with renal disease. Of additional interest was the range of responses to questions about matching the recommendations with available evidence. This seems to reflect both an awareness among renal medicine health workers of the importance of evidence-based medicine and a maturing understanding of the need for the evidence to be of high quality.

The CARI guideline process has a relatively short history (just over three years), and before Johnson's survey the CARI Steering Committee had adopted a number of strategies that anticipated some of the issues his survey raises. These strategies included:

• establishing a formal link with the Renal Cochrane Organisation (to produce the best possible search outcomes of all the available evidence);

• adopting the National Health and Medical Research Council (NHMRC) evidence levels I (systematic reviews) and II (randomised controlled trials) as the minimum requirement to justify definitive guidelines (to assess the quality of evidence available and match appropriate guidelines with that evidence);

• adopting a peer-review process to evaluate draft guidelines to complement newly revised requirements for guideline writers' conflict-of-interest declarations (to assess perceptions of guideline validity and the influence of external agencies);

• broadening the multidisciplinary nature of guideline working parties.

In addition, the Australian Kidney Foundation has moved to further disseminate the guidelines, and the CARI guideline process has been reformed with the aim of meeting the standards required to achieve NHMRC endorsement. Furthermore, feedback obtained from legal advisers suggests that the CARI guidelines and the process of establishing them are far more likely to obviate litigation than to promote it. The next important phase for the CARI guidelines will be the development of an implementation process.

As the acceptance of evidence-based medicine increases and the knowledge base among healthcare workers of the nature, quality and relevance of evidence in patient care expands, the CARI guideline process is likely to be enhanced. The results of future surveys of the type carried out by Johnson will be keenly anticipated.