Consensus process

The conference was organized by the Department of Anaesthesiology of Greifswald University Medicine under the auspices of the German Resuscitation Council (GRC) and the German Association of Anaesthesiologists (BDA). To find consensus an adaptation of the RAND-UCLA Appropriateness Method [27] was used employing a three-step approach: (i) presentations and discussions on the current state of practice and research, (ii) open break out workshops, where faculty and attendees produced consensus statements, (iii) which were subsequently adopted or rejected following the RAND-UCLA method. RAND-UCLA method was developed to evaluate whether the expected benefits of a medical intervention outweighs the expected harm to such a degree, that implementation should be recommended [27]. In RAND-UCLA Appropriate Method a literature review lays the foundation for experts rating the benefit-to-harm ratio for different aspects of the new intervention on a 9-point scale. A point of 1 assesses the intervention as unsafe as expected harms greatly outweigh expected benefits; while a point of 9 strongly recommends the intervention, because expected benefits greatly exceed expected harm [27].

Presentation of the current state of research

The conference started with presentations from national and international speakers reporting on their experience and research in the field of SBA of CFR. CFR systems in Denmark, the Netherlands, Switzerland, Austria, and Germany were presented and discussed with a focus on their differences and similarities. To facilitate comparison of the systems all speakers were asked to address a set of topics listed in Table 1.

Comparison between different CFR systems in Europe

Additionally, all major CFR app developers were invited to present their systems and attendees of the conference had the opportunity to compare different products. For probity reasons this session was held as a separate industry forum.

Both research and product presentation served as basis for the discussion and generating consensus statements during the open break out workshop sessions.

Open break out workshops

The subsequent workshop session lasted 40 mins and was held in an open space allowing for six workshops simultaneously addressing different topics, see Table 2. All attendees were invited to take part in the workshops and actively contribute to the creation of the consensus statements. Each workshop was led by two faculty members, who facilitated the discussion, took notes on a whiteboard and ensured that the views of all attendees were reflected in the consensus statements. The attendees could choose freely which workshop to join. There was no set number of attendees per workshop. The attendees were also encouraged to move between workshops during the session to augment exchange of expert knowledge and stakeholder views on the different topics. However, one limitation might be that discussions were interrupted by panel members joining or leaving the workshop.

Workshop topics

The consensus statements were presented to all participants of the conference and put to vote during stage three of the modified RAND-UCLA process, which started immediately following the break out workshops. To enable anonymous real-time voting, the mobile phone software PINGO (University of Paderborn, Germany) was used. The consensus statements were presented sequentially to all participants of the conference the consensus panel on a screen and on the voting tool. After the statement was presented and read out, it was directly voted on. To allow the participants to familiarize themselves with the technique, voting time for the first and second statement were 1 min respectively. Thereafter voting times for the remaining statements were reduced to 30 s. All participants had the same vote independent from experience and background. In line with the RAND-UCLA Appropriateness Method [27] all statements were assessed with a nine-point scale, with “1” meaning “rejection”, “5” “uncertain” and “9” “approval”. The voting results were later grouped into approval/rejection and consensus was determined based on the criteria listed in Table 3.

Modified RAND-UCLA classification criteria