ERS syllabus for postgraduate training in respiratory infections: a guide for comprehensive training

Aliberti, Stefano; Farr, Amy; Tabin, Nathalie; Blasi, Francesco; Torres, Antoni; Woodhead, Mark; Migliori, Giovanni Battista; Sotgiu, Giovanni; Dimopoulos, George; Chalmers, James D.; Ringshausen, Felix C.; Loebinger, Michael R.; Read, Robert; Rohde, Gernot

The Delphi process was phased in three rounds (figure 1). Members of ERS Assembly 10 (respiratory infections) and national experts were asked to complete an online questionnaire, which was then processed by the ERS office. The results were presented to the task force for more detailed discussion. The decisions derived from these consultations were integrated into the next Delphi round (Delphi 1). It is to be noted that although the same respondents were contacted in each round, there was a drop off in the number of responses between rounds. Respondents were asked to rate in terms of agreement whether sections should be included (mandatory or optional) or excluded. A Likert scale from 1 (strongly disagree) to 5 (strongly agree) was chosen to ensure measurement reliability over the three different Delphi rounds. Agreement was operationalised through a majority of responses in the top two points of the scale (measured as the sum of frequencies of agree and strongly agree responses). In Delphi studies consensus is considered a necessary, but not sufficient, condition for agreement concerning the inclusion of items [6] . The iteration of rounds was also required to establish the stability dimension for agreement. Stability was defined as "the consistency of responses between successive rounds of a study" [7] . Both consensus and stability dimensions were investigated in the study.