WHY WE NEED GUIDELINES FOR GUIDELINES - A STUDY OF THE QUALITY OF CLINICAL-PRACTICE GUIDELINES IN AUSTRALIA

Objective: To appraise the quality of clinical practice guidelines (CP Gs) produced in Australia. Design: Cross-sectional survey. Participant s and setting: 76 organisations, comprising all clinical colleges and faculties, federal and state health departments and national and state non-government agencies involved in health policy in Australia in Oct ober 1993. Outcome measures: Ratings for how CPGs met 18 criteria (''f ully'', ''partially'' or ''not at all'') for validity, reproducibility , applicability, clinical flexibility, clarity, multidisciplinary proc ess, documentation and scheduled review. Results: Response rate was 87 %, with 32 organisations submitting 42 documents; 34 of these were cla ssified as CPGs. None of the 34 CPGs fully met all criteria for qualit y. None fully stated costings or described processes for retrieving an d synthesising evidence; only six (18%) explicitly stated the expected health outcomes; six (18%) described the method to reach consensus; a nd seven (21%) fully stated controversies and potential conflict with other guidelines. However, at least 75% of CPGs fully met all criteria for applicability, flexibility, multidisciplinary input and documenta tion and two out of three criteria for clarity. Zero to 103 references to scientific data were cited per CPG. Conclusion: Australian guideli nes developed before the publication of the NHMRC ''guidelines for gui delines'' in 1995 did not meet internationally recognised criteria for quality. There is a need to apply the recommended NHMRC process and t o monitor future CPG quality.