A new method of developing expert consensus practice guidelines.

Abstract

To improve the quality of medical care while reducing costs, it is necessary to standardize best practice habits at the most crucial clinical decision points. Because many pertinent questions encountered in everyday practice are not well answered by the available research, expert consensus is a valuable bridge between clinical research and clinical practice. Previous methods of developing expert consensus have been limited by their relative lack of quantification, specificity, representativeness, and implementation. This article describes a new method of developing, documenting, and disseminating expert consensus guidelines that meets these concerns. This method has already been applied to four disorders in psychiatry and could be equally useful for other medical conditions. Leading clinical researchers studying a given disorder complete a survey soliciting their opinions on its most important disease management questions that are not covered well by definitive research. The survey response rates among the experts for the four different psychiatric disorders have each exceeded 85%. The views of the clinical researchers are validated by surveying separately a large group of practicing clinicians to ensure that the guideline recommendations are widely generalizable. All of the suggestions made in the guideline are derived from, and referenced to, the experts' survey responses using criteria that were established a priori for defining first-, second-, and third-line choices. Analysis of survey results suggests that this method of quantifying expert responses achieves a high level of reliability and reproducibility. This survey method is probably the best available means for standardizing practice for decisions points not well covered by research.