any other patients who would be interested to participate. All participants were invited to be involved in the following step of SONG-PKD which involved completing a Delphi survey.26 Results of this survey will be emailed to all participants. The general public were not involved. Data collection The 2-hour focus group discussions were conducted from June to November 2017 until data saturation. Data saturation was achieved when CL, YC and AT agreed that little or no new concepts were arising from subsequent focus groups. We developed the question guide from the literature and with input from the research team (online supplemental material and methods).25 28 29 Focus groups were convened in a venue external to the hospital and facilitated by one investigator (English—AT (researcher), TG (researcher), YC (academic nephrologist); French— BS (academic nephrologist); Korean—YK (academic nephrologist)). Focus groups were designed with the intent to have a broad range of demographic and clinical characteristics (including patients/caregivers, age). We did not consider severity of symptoms a priori. We did not separate patients from a caregiver as they preferred to participate in the same group. A cofacilitator recorded field notes. All discussions were audio recorded and were transcribed. Data analysis All transcripts were entered into HyperRESEARCH (V.3.7) for analysis and coded line by line, in the original language and then translated for investigator triangulation, by CL (researcher) (English, French) and HK (academic nephrologist) (Korean) using thematic analysis and drawing on principles from grounded theory to identify concepts related to perspectives on testing for on July 8, 2022 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2020-038005 on 10 October 2020. Downloaded from Logeman C, et al. BMJ Open 2020;10:e038005. doi:10.1136/bmjopen-2020-038005 3 Open access ADPKD.30 From grounded theory, we conducted initial coding (memoing) and line-by-line coding of the data, used constant comparison within and across the transcripts and inductively identified concepts and themes. In accordance with thematic analysis, we identified initial concepts and grouped similar concepts into themes. Codes were grouped by similar concepts into themes and subthemes which were discussed and revised with AT/ TG/YC/BS/YK who independently read the translated transcripts. To ensure reliable interpretation of the translated transcripts, CL and HK were available to give more context of the quotes. Investigator triangulation ensured that the analysis captured the full range and breadth of the data.31 RESULTS In total, 154 participants (120 patients, 33 caregivers) participated in 17 focus groups across Australia, France and Korea. The demographics are shown in table 1. Participants’ age ranged from 19 to 78 years (mean age 54.5 years) and 67 (42%) were men. Most patients Table 1 Participant demographic characteristics CharacteristiMissing data from two participants. †Patient only (n=61; n=31; n=24). CKD, chronic kidney disease. on July 8, 2022 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2020-038005 on 10 October 2020. Downloaded from 4 Logeman C, et al. BMJ Open 2020;10:e038005. doi:10.1136/bmjopen-2020-038005 Open access were diagnosed between the ages of 21 and 40 years and the majority of patients were predialysis (n=76, 61%), followed by transplant recipients (n=31, 26%) and those on dialysis (n=19, 13%). The majority of caregivers defined themselves as the spouse or partner of the patient (n=24, 71%), but also included child (n=2, 6%), daughter-in-law (n=1, 3%), parent (n=4, 12%) and sibling (n=1, 3%). Reasons for declining to participate included having other commitments and being too unwell to participate. Five themes were identified with both patients and caregivers contributing to the concept unless otherwise stated: avoiding financial disadvantage, futility in uncertainty, lacking autonomy and support