Table 2.

Cross-site comparison – contexts and implementation

	Site A: large city, South East England	Site B: major city, Northern Ireland	Site C: major city, East Midlands	Site D: rural, South West England	Site E: major city, North West England	Site F: urban area, South East England	Site G: major town, South England
Existing service prior to SUPPORT-HF2 trial	Consultant-led clinics and inpatient management Standard primary care diagnosis and ongoing management
	Hospital HFSN team, community HFSN teams serving HFrEF only	Hospital and community HFSN teams	Hospital and community HFSN teams	Hospital and community HFSN teams	Community HFSN team in one part of CCG only	Hospital and community HFSN teams	No community HFSN team
Staff involved in SUPPORT-HF2 trial	Chief investigator, secondary care HF nurses, SUPPORT HF trial team	Local PI, 2 research nurses	Local PI, secondary care-based lead HFSN and 2 communities	Local PI, secondary care HFSNs, 2 research practitioners	Local PI, research administrator	Local PI, secondary care-based HFSN	Local PI
Setting for recruitment	Study lead nurse recruited from wards or clinics, community HF nurses in the early phase only	Research nurse recruited from CCU and other wards or by letters sent out post-discharge	Recruited by lead HFSN then later by research nurse and community HFSNs from the clinic	2 research practitioners, secondary care HFSNs and local PI recruited from wards and hospital clinics	By local PI in hospital clinic	By secondary care nurse in hospital clinic	By local PI in hospital clinic
Clinical profile of participants in this site	As per protocol, no specific distinction described.	As per protocol, no specific distinction described.	Patients were recruited only when ready for discharge from the hospital clinic or HFSN service.	As per protocol, no specific distinction described.	Patients were targeted if they were particularly unwell and deemed in need of monitoring between 6 monthly clinic visits.	In time, with appreciation of RCT design, staff avoided recruitment of any patients deemed too unstable for the control arm.	As per protocol, no specific distinction described.
How technology use by patients was supported during the trial	By CCM team	By local research nurse, supported by the CCM team	By local research nurse, supported by the CCM team	By local research nurse, supported by the CCM team	By CCM team	By CCM team	By CCM team
Extent to which the model was integrated into clinical pathways	Initially integrated, later separated due to role overlap	Limited: resistance from both hospital and community HFSNs and GPs	Partial: intention was for participants to be looked after by the CCM team, but sometimes, the hospital HFSN accessed patient data.	Partial: integrated with secondary care only as community HFSN service was resistant	Not integrated: following recruitment, participants were looked after by the CCM team.	Not integrated: following recruitment, participants were looked after by the CCM team.	Not integrated: following recruitment, participants were looked after by the CCM team.