Developing Consensus Using the Delphi Method

The Delphi method is a technique for obtaining consensus from a panel of experts on a particular topic or set of decisions, such as diagnostic or treatment algorithms, medical education competencies, or curricular components. This article aims to introduce researchers new to consensus methodology to the Delphi method. A classic Delphi process is characterized by anonymity of panelists, who complete iterative surveys to rank items and suggest changes. After each survey round, researchers amend the items, seeking a predetermined cutoff of consensus. The amended survey is sent to panelists for reconsideration along with quantitative and qualitative analysis of prior results. Modified Delphi methods may include having in-person components and basing initial surveys on literature rather than panelist input. Rigor is enhanced by clear reasoning for selection and representation of the expert panel, evidence for items in the initial survey, a clear definition of consensus, response rates and results after each round, how iterative feedback was provided, and reasoning for steps researchers took along the way to move toward consensus. Qualitative analysis of comments can yield insights into the topic, including items for which consensus is not achieved. Thoughtful Delphi design can lead to trustworthy results with broader community acceptance.