What is the Delphi Technique?
The Delphi Method, or Delphi Technique, is a process used to help determine the outcome of a project based on rounds of questionnaires given to authoritative figures in a particular topic, who can offer their expert opinions freely thanks to using anonymous responses.
The Delphi Method seeks to ascertain the most likely forecast or outcome based on the results of agreement with statements in the questionnaire and provide a summary report of findings. The process continues until the panel of experts reach a group consensus, often after a first, or second round and sometimes after further rounds of survey. In a healthcare setting, the Delphi Technique can help influence change and add context to decision making.
It is often used by clinicians, researchers and medical affairs and market access teams from commercial organizations to help foster better outcomes for patients.
When Should We Use Delphi Technique?
The Delphi Technique is most often employed in situations where there is no clear-cut evidence on which to rely for answers. This may include forecasting, estimating future risks, generating new evidence, and policy making.
It is favoured by the healthcare industry when seeking to better define unmet need (e.g. identifying patients who may benefit from a given intervention), optimise healthcare practices (e.g. diagnostic criteria or treatment approach), and support with defining policy (e.g. guidelines).
Conducting research using the Delphi technique
Step One – Define the problem and aim of the project
You will need to clearly define the issue you are trying to overcome, and the goal you are trying to reach. This is best clarified following a review of current evidence and guidelines.
Step Two – Employ a Delphi facilitator
To deliver a Delphi consensus appropriately, an independent facilitator is necessary to ensure correct structure, anonymity of responses and minimisation of any bias. Engaging a third party experienced in the Delphi method like Triducive for support in designing a successful study using the Delphi technique ensures the outcome can be relied upon and any subsequent communication campaign will hold credibility.
Step Three – Develop clear statements
Delphi studies use statements that describe a position that demands an opinion on them and ask the panel to rate their level of agreement according to a Likert scale.
When creating Delphi statements, ensure you have covered the issue you are trying to overcome from a range of angles to better define the consensus opinion of your panel.
Make sure there is only one point to consider in each Delphi statement, to reduce confusion and ensure you receive the most accurate responses.
Step Four – Engage a panel of relevant experts
The panel of experts can comprise of any person with the relevant knowledge and experience on the subject in question. This may include professional experts, consumers, or advocates who would engage with the service or product.
An individual does not need a set number of years of experience or training to participate, but ensure your panel reflect the expertise you need to respond effectively to each statement. The minimum number of panel members needed is open to debate with some suggesting as few as 8 and one study demonstrating that the outcome from expert opinion didn’t change when the results from 23 experts were compared to a much larger sample. For areas where the total population of experts who could have an opinion is significant Triducive recommends aiming for larger number than this.
Step Five – Send the first round of your Delphi Technique statements or questionnaire to your experts
Once you are satisfied with your statements and the questionnaire, send it to your panel of experts to gain insight into their opinions and expectations for the future.
Step Six – Analyse results & consider the need for further questionnaire rounds
The most common definition for consensus is per cent agreement with a given statement (i.e. the number of responders agreeing vs the overall number of responses to that statement). Triducive suggests consensus at a minimum threshold of 75% participant agreement which has been established as an acceptable threshold in many previously published Delphi studies. This threshold needs to have been agreed prior to the questionnaire being shared with the panel.
If any of the statements fail to achieve the pre-agreed consensus threshold then a decision needs to be made to either accept the ‘dissenting’ response and consider how this informs the study aim or to amend those statements and recirculate the questionnaire for another round. During this process the group will converge towards the “correct” answer and the process is stopped using mean scores of the final round.
Step Seven – Provide recommendations based on outcome and consider how best to communicate them to support change
Once your panel have reached consensus, build a summary report of the outcome and their views and use this to identify common themes, and potential new opportunities to drive change and improve practices. Usually, the study can be written up and submitted for peer-review publication. This is strongly encouraged by Triducive as it adds strong credibility to the results and enables citation in other communications and activities.
Examples of Delphi Technique In Use
The Delphi Technique can be used in a range of projects with broad or niche aims. View the case studies below to see how Triducive have employed a modified Delphi Method to support clients and drive positive change.
Triducive create groundswell of credible opinion to positively influence EU antibiotic policy
Triducive develop aligned recommendations for managing hyperkalaemia across the cardiorenal spectrum
Triducive help a leader in respiratory medicine to improve referral pathways in the United Kingdom
Ready to spearhead your Delphi study?
Learn more about the expert team behind Triducive that helps teams working in healthcare all around the world generate consensus-led evidence that gets published and supports change.
We will help you design and deliver a strong consensus study capable of peer-review publication, using our modified Delphi technique.