Analysis
All 14 transcribed interviews were coded and categorised
inductively using qualitative content analysis [25-27],
primarily inspired by Graneheim and Lundman [26],
and supported by the qualitative data analysis software
NVivo 10. We focused primarily on the manifest content
of data [26,27]. Transcribed interviews were first
read in their entirety to obtain a sense of whole [25,26].
The text was then divided into units of meaning that were
condensed and grouped together to create categories and
sub-categories [26]. During this process, the content of
sub-categories and categories was continuously compared
to data in other emerging categories to ensure
that the data in categories belonged together and not to
other categories [25]. The analysis resulted in four categories
with a total of 12 sub-categories.