Doc menu


The Delphi consensus method proved a suitable method for identifying the key research priorities for the radiography profession as determined by practitioners. A total of 133 priority topics were identified that reached consensus agreement across the three round process. Delphi consensus processes often have the potential to lose participants as the rounds progress and this can often limit the impact of the subsequent outcomes. However, in this study the attrition rate between rounds was minimal and the final panel size remained high with 76.6% of the original panel still responding in round three of the Delphi process; much higher than the retention rates from the physiotherapy Delphi process where between 52-64% of panel members remained by the final round across the four panels.While the consensus criteria was set with a coefficient of variation (CV) <30, across all 133 identified topics, the CV was ≤23 showing good agreement for the identified topics.

Tables 7-10 provide a breakdown of the top ten topics for four key sub-specialties, radiotherapy, diagnostic imaging, mammography and ultrasound, and education. Care should be taken when assessing the importance of the ordering of the topics as this is based on the mean score for importance. Higher mean scores may reflect greater consensus within the group of individuals that scored that topic, so lower ordered topics may simply reflect a wider diversity of opinion within those that scored the item. Hence in Table 6 many of the higher ranked top ten topics are radiotherapy or oncology focussed and this may be due to the lower number of respondents ranking these topics. In addition, the differences between mean scores in Table 6 are small and represent very small differences in panel scoring.

Finally, the five key themed areas (see Figure 1) were identified to simplify presentation of the priority areas to a wider audience. It is important to be mindful that these themes reflect a broad overview of the 133 topics that reached consensus and primary reference should be made to the consensus topics themselves. 

Content tools

Accessibility controls

Text size