Nvivo v 12 pro

Immediately following the interview, the interview moderator drafted a detailed report of participant responses, organized by topic, using the audio recording to include verbatim quotes. The immediate creation of interview summaries is a best practice for collecting iterative feedback. Summaries were imported into analytic software NVivo Pro v. 12.0 (QSR International, Melbourne, Australia), and coded utilizing thematic content analysis using a two-phase approach.^{[23 ]} In the first phase, individual interview summaries were reviewed, and significant statements were coded based on a priori constructs from the interview guide and novel concepts that emerged from the interviews themselves. Then, coded statements were sorted into the three survey domains. Statements from each category, across interviews, were extracted into a single report and re-analyzed to identify major themes and sub-themes. Themes were identified based on frequency and attributes (e.g., interpreting comments in the context of disease-site specialist vs. generalist status). Finally, the coding team (K.L., H.Z.) met with the interviewer (E.C.) and study PI (E.F.G.) to achieve consensus on the major themes.

We performed quantitative analyses using STATA (Windows version 12.0, STATACorp LLC, College Station, TX). We conducted paired t-tests to evaluate pre- and post-treatment change in ESAS scores in each intervention group. We also used linear regression models to evaluate the differences in post-treatment depression scores between music therapy and massage therapy groups after adjusting for pre-treatment depression scores.
We used a thematic content analysis approach to analyze free-text comments with NVivo Pro v.12.0 (QSR-International, Doncaster, Australia).¹² After an initial review of the qualitative data, we constructed a structured codebook to identify manifest and latent content. Using this codebook, we assigned codes to detailed free-text comments, excluding comments with only single-word responses (e.g., “great,” “nice,” “thanks”). We grouped the coded comments into thematic categories (e.g., uplifting, relaxing) via consensus during successive team meetings. We then compared the prevalence of themes between music therapy and massage therapy recipients. We also examined whether different themes were prevalent among patients who reported clinically meaningful pre-post improvements (i.e., ≥ 1 point reduction in ESAS depression score).