TY - JOUR T1 - Progress of GP clusters 2 years after their introduction in Scotland: findings from the Scottish School of Primary Care national GP survey JF - BJGP Open JO - BJGP Open DO - 10.3399/bjgpopen20X101112 VL - 4 IS - 5 SP - bjgpopen20X101112 AU - Stewart Mercer AU - John Gillies AU - Bridie Fitzpatrick Y1 - 2020/12/01 UR - http://bjgpopen.org/content/4/5/bjgpopen20X101112.abstract N2 - Background The concept of GP clusters is derived from 'quality circles' in general practice in Europe and Canada. GP clusters commenced across Scotland in 2016 to improve the quality of care of local populations.Aim To determine GPs' views on clusters, and the robustness of bespoke questions about them.Design & setting A cross-sectional national survey of work satisfaction of GPs in Scotland took place, which was conducted in July 2018–October 2018.Method An analysis of bespoke questions on GP clusters was undertaken. The questions were completed by quality leads (QLs) and all other GPs in a nationally representative sample of GPs.Results In total, 2456 responses were received from 4371 GPs (56.4%). QLs reported that clusters were meeting regularly, and were friendly and well organised but not always productive. Support for cluster activity (data, health intelligence, analysis, quality improvement methods, advice, leadership, and evaluation) was suboptimal. Factor analysis identified two separate constructs (cluster meetings [CMs] and cluster support [CS]), which were minimally influenced (<2%) by GP and practice characteristics. Non-QLs (75% of all GPs) were generally satisfied with the two-way communication with the cluster QLs, but the great majority (>70%) reported no positive changes in various aspects of quality improvement. Factor analysis of these items indicated two constructs (cluster knowledge and engagement [CKE] and cluster quality improvement [CQI]), which were minimally affected by GP and practice characteristics.Conclusion GP clusters are ‘up and running’ in Scotland but are at an early stage in terms of perceived impact and appear to be in need of more support in order to improve quality of care. The bespoke questions developed on clusters have robust construct validity, suitable for future surveys. ER -