Analysis
Thematic analysis of the transcribed interviews was performed through an
iterative process of inductive and deductive coding.23Inductive codes were identified through topics emerging from
transcripts, while deductive codes were identified a priori from prior
research and the interview guide.8 Each transcript was
independently coded by two researchers (AW/MS) and memos were made to
identify emerging themes and data interpretation.(Appendix 2) All coding
discrepancies were resolved via consensus discussions between the coders
and final coding assignments were made using NVivo
12.19 Although thematic saturation was achieved at 12
interviews, we continued conducting interviews because we had a finite
number of participants, all of whom had unique clinical experiences with
UTx and whose insights, we believed, were additive to the study results.
Quantitative data analysis entailed descriptive statistics (e.g.,
frequencies, medians, interquartile ranges, and means) for patient
demographic and outcomes data. In addition, qualitative codes were
transformed into categorical variables (e.g., perceived high or low risk
of UTx, information needs, and expectations of UTx). Bivariate analysis,
performed using R , tested associations between transformed
qualitative data and quantitative demographic and outcomes
data.24 Statistical significance was defined as
p<0.05.