Table 1 Description of features of IMPART cases relevant to study

Wave 1

Woods

• Urban area, very high rates of psychosis

• High ethnic diversity

• Two geographically distinct teams

• Reported average caseload per staff of 28

• Very high staff turnover & absence

• Low morale: half of one team left in the first 6 months of study

Moor

• Large rural area

• Population predominantly white British

• Early Intervention not separate service, embedded in geographically spread community teams

• Lower caseloads (approximately 15, exact figures n/a) but long travel times

Wave 2

Ocean

• Urban area

• Population majority white British

• Three geographically distinct teams which performed quite differently

• One team had very high staff turnover and high levels of sickness absence, in this team carer support was delegated to one carer lead rather than part of all CC’s work.

• Caseloads high (approx. 26)

Seashore

• Primarily urban area

• High ethnic diversity

• Three teams across locality, operating quite differently and independent of each other

• Low staff morale and very high turnover, in one team all CCs and team manager left over a period of three months

• Trust implementing new DHI for service users at the same time as IMPART study

• Caseloads described as ‘high’ but numbers not available due to period of intense change

Wave 3

Lakes

• Largely rural area

• Six teams cover large geographical area, managed in pairs

• First IL a senior psychiatrist who left early in project, succeeded by another psychiatrist

• Led to variable engagement with IMPART study over time

Marsh

• Mainly urban area with rural pockets

• Two separately located teams covered by one IL, one RS

• Very early in project, RS role given to non-clinical staff member in R&D department