Determining the relevance of themes in thematic analysis can be subjective, and it depends on the research question and objectives. One way to measure the relevance of themes is to consider how frequently they appear in the data and whether they are connected to the research question. Another approach is to prioritize the themes that are most significant to the participants or have the strongest impact on the research objectives. However, it is crucial to maintain transparency in the decision-making process and document the rationale for prioritizing particular themes. Additionally, seeking feedback from other researchers or experts in the field can also be helpful in validating the interpretation of the data.

Relevance of themes in a qualitative approach is mainly determined by the theoretical/conceptual framework you've adopted. Your framework determines the initial assumptions you made which the emerging themes can prove or disprove.

In thematic analysis itself, if you are following Braun and Clarke's version, you should not use frequency as a basis for determining which themes are most meaningful.

But since you are working with data from a survey, you might consider creating a series of dummy (0-1) variables, one for each code. Many of the qualitative data analysis programs (including MAXQDA and NVivo) handle this under the heading of "mixed methods." They thus support export of these variables in spreadsheet form.