To establish whether the DK responses would bias the results we performed a sensitivity analysis using three different methods: complete case analysis, directional coding, and multiple imputation, as suggested by Young and Kroh [ 29, 30].
Only 5 of the 103 scenarios evaluated resulted in an ICER comparing BOC/RGT to PR48 that was more than $5000 different from the base case analysis ($16,792/QALY).
Similarly, only 18 of the 103 scenarios evaluated resulted in an ICER comparing BOC/PR48 to PR48 that was more than $5,000 different from the base case analysis ($55,162/QALY).
Regarding drop-out and loss to follow-up, different analytical approaches (complete case analysis, last observation carried forward, multiple imputation, best/worst outcome) will be followed in order to assess their impact on the study results.
This paper combines frame analysis, context analysis and ecosocial analysis to three different case studies.
If the best-worst and worst-best case analyses result in a different conclusion than the complete case analysis, multiple imputations will be performed.
When the percentage of missing data is low or intra-cluster correlation coefficient is small, different imputation strategies or complete case analysis approach generate quite similar results.
In the subsequent axial coding step, the data from the different interviews were compared (cross case analysis) and grouped into clusters based on the main emerging (sub themes.
Our results show that, first, when the percentage of missing data is low or intra-cluster correlation coefficient is small, different imputation strategies or complete case analysis approach generate quite similar results.
Through the analysis of different case studies, we will investigate different topics from the perspective of ethnomusicology, cultural anthropology, and performance studies.
Strings were transformed to uppercase to make sure that different case did not affect our analysis.
