Recruitment and Retention

The observed consent rate met the green-light target, but adherence and retention did not. Follow-up retention was very low, with only 18% of participants providing any follow-up data. Consent among approached participants was high at 92%, exceeding the green threshold. The trial recruited only 43% of its target sample. Participant interview data were almost unavailable despite many participants initially consenting to interview. Bianzhi quotas are fixed by served population and do not adequately reflect service scope or regional variation. Contact instability contributed to poor follow-up because many participants had no contact details or disconnected phone numbers. Probation-based follow-up failed because many participants lacked an assigned probation officer, were recalled to prison, or disengaged. Contract-based GPs earn significantly less than Bianzhi staff despite similar responsibilities, contributing to turnover. Probation tracking was weakened by missing probation details and service strain from ECSL and Probation Reset. Recruitment eligibility rules restrict non-local applicants through household registration, degree, and age requirements. General practice was seen as l…