Page 126 - Shared Guideline Development Experiences in Fertility Care

Chapter 6
underlying the recommendations was classi ed into ve levels: I, systematic review/Randomized Controlled Trials (RCT); II, RCT; III, comparative studies; IV, case studies or expert opinion and level P, patients’ opinion [24]. In collaboration with Freya, the patients’ questionnaire was supplemented with additional laymen’s information to clarify the jargon used. irdly, for the organizational as well as the non-organizational potential indicators, all panel members were asked to give a top ve ranking to promote the discrimination between indicators with a high Likert score [16, 25]. Finally, all panel members were invited to rephrase the proposed indicators and to add comments. Reminders were sent by email to non- responders a er 2 and 4 weeks.
Data analysis: step 2
Campbell’s criteria and the top ve ranking method were used to analyse the results of the questionnaire surveys of step 2 [26]. erefore, per potential indicator, we calculated an overall panel median score and an overall ranking score for each panel. To calculate the overall ranking score, a potential indicator ranked rst, second, third, fourth or last in the top ve received 5,4,3,2 or 1 point respectively. e overall ranking score was expressed as the percentage of the maximum top ve score that could be obtained (0-100%). A potential indicator was associated with face validity and proposed for ‘selection’ in this consensus round if it matched the following criteria: (i) an overall panel median score of 8 or 9 combined with a top ve ranking ≥10th percentile of the percentage of the maximum top ve score and (ii) an agreement between the ratings of the independent panel members. e agreement was reached if 70% or more of ratings within a panel was in the highest tertile (7, 8, 9). A potential indicator was ‘rejected’ if none of these criteria were met. A potential indicator was considered ‘equivocal’ or open to discussion if it matched an overall panel median score of 8 or 9 with agreement, but with low or no top ve ranking score.
Step 3: consensus round
A er rephrasing, the organizational and non-organizational ‘selected’ and ‘equivocal’ indicators were presented separately in a second questionnaire intended for all panel members for commented written approval. In this questionnaire, the ‘selected’ and ‘equivocal’ indicators were presented along with the comments provided by the particular panel (e.g. patients and professionals), the frequency distribution of scores within the panel, the overall median score of the panel as
124