91 Monitoring the Dutch Solid Start program: developing an indicator set for municipalities agreement was assessed by the IPR-score (interpercentile range, difference between 30th and 70th percentile) and the IPRAS-score (interpercentile range adjusted for symmetry). If the IPRAS is larger than the IPR, there is agreement among experts and if the IPR is larger than the IPRAS, there is disagreement. We planned to 1) accept topics with median score ≥ 7 with agreement, 2) reject topics with median score ≤ 3 with agreement, and 3) discuss all other topics (median score 4 – 6 or without agreement) in Delphi round 2. However, round 1 resulted in a large majority ‘accepted’ topics and well exceeded the number of intended indicators. We therefore decided to prioritize these ‘accepted’ topics in the second Delphi round and rejected all other topics. The experts’ suggestions for new topics were read and discussed by the researchers (JM, IB, JS) until consensus was reached on additional topics. New topics were combined or reformulated if necessary and added to Delphi round 2. Delphi round 2: expert meeting The second Delphi round consisted of expert meetings to prioritize the topics using the cumulative voting method. Meetings were held online due to Dutch COVID-19 policy restrictions and we organized three separate smaller meetings to encourage active participation during the online meetings. The meetings of +- 120 minutes were recorded. Experts were first informed about the results of Delphi round 1. Next, they were encouraged to prioritize topics by dividing 100 points at their own discretion. After the individual prioritization, experts entered their scores into an interactive program to aggregate scores of all participants in the meeting. We encouraged experts to reflect on these aggregated scores. After the discussion, experts were invited to reconsider their earlier individual scores again. This sequence was repeated for the three phases (preconception, pregnancy and after birth). Subsequently, we aggregated all final scores and classified the topics from high to low sum scores. Within every phase (preconception, pregnancy and after birth) we searched for a sudden decline in sum scores as a natural cut-off point for prioritized topics. This led to a draft list of prioritized topics. In addition, we transcribed the expert meetings verbatim and analysed the data using MaxQDA. One of the researchers (JM) coded the data for considerations in the prioritization and requirements for the indicator set. Coding was checked by a second researcher (IB). The researchers (JM, IB, JS) consequently checked the draft list of prioritized topics against the experts’ requirements for the indicator set. We checked whether the requirements were fulfilled or whether we should add lower prioritized topics to fulfil the requirements. At the end of the second Delphi round, we had a final list of prioritized topics. 4
RkJQdWJsaXNoZXIy MTk4NDMw