We have categorised responses to the treatment frequency with available pharmacological agents as "routine", "occasional", "individualised according to patient" and "never".
We categorised responses in 3 groups: broadly satisfied, broadly dissatisfied and a mixed opinion (i.e. reporting some positive as well as negative aspects).
We categorised responses to the open-ended question (about conditions for which respondents would like POC tests to help them make diagnoses) using the International Classification of Primary Care (ICPC-2-R) 17 system (see online supplementary appendix VII).
Alternatively, some researchers asked participants to identify their ethnic or cultural background when completing a demographic questionnaire but it is not clear if these were open ended or pre categorised responses [ 25].
The second social functioning question (question 6), which assessed physical or emotional interference in normal social activities, was dichotomised by categorising responses of 'not at all' or 'slightly' as not reporting dysfunction and responses of 'moderately'quiteite a bit', or 'extremely' as reporting dysfunction.
For perceived survival we also categorised responses for each cancer according to whether they were accurate using CR-UK survival statistics (Cancer Research UK, 2011) into pessimistic (survival estimates lower than actual survival statistics), accurate (survival approximated actual survival) or optimistic (survival estimates higher than actual).
These parameters were used to categorise brainstem responses as parasympathetic (calming) response, sympathetic (activating) response, arousal (alerting) response and unclear response.
Initially, two raters (KG and DC) independently read and categorised response into categories and subcategories under the general headings of 'Advantages' and 'Disadvantages' of seeking support from friends or family.
It was of interest to categorise these responses to gain an appreciation of the diversity of thinking in the measurement context, but the responses were not considered to be hierarchical in nature.
We also applied the modified Choi criteria, which take into account both the change in size and enhancement, to categorise patient responses (Nathan et al, 2010).
