A mixed-methods investigation of the extent to which routinely collected information can help evaluate the implementaion of screening and brief alcohol interventions in primary health care

Abstract

Background: UK health policy has sought to encourage alcohol screening and brief intervention (ASBI) delivery in primary care, including via pay-for-performance (P4P) schemes. To measure the impact of such policies, a range of data exist, including General Practitioner (GP) Read codes, which record all clinical activity. However, previous studies have highlighted the difficulties of using Read code data for evaluation purposes, with concerns around the distorting effect of P4P on healthcare recording. Against this background, this research investigated whether Read code data can be used to provide a meaningful measure of ASBI implementation in primary care. Methods: Sequential mixed methods design, comprising: (1) systematic literature review to identify what factors influence the recording of routine clinical data by UK primary care physicians; (2) analysis of ASBI Read code data from 16 GP practices in North East England; (3) 14 GP interviews to explore the barriers and facilitators affecting their ASBI recording. Results: (1) Multiple factors shape primary care physicians’ recording of routine data, including structural influencers (such as the design and resourcing of the coding system), and psychosocial factors (including patient characteristics and physicians’ perspectives on their role as care-givers). (2) 287 Read codes exist to record alcohol- related activity however only a small minority are used regularly, generally relating to the identification of alcohol use disorders. Whilst many unused Read codes are associated with relatively rare alcohol conditions, a significant number relate to duplicate or outmoded terminology. Overall, practices associated with higher recorded rates of key ASBI service indicators were signed up to P4P schemes. (3) GP interviews suggested that across all practices, nurse-administered ASBI components were most likely to be provided and coded consistently, with GP-delivery and recording activity far more ad hoc. Conclusion: Whilst routine data may be a valid indicator of more successfully embedded ASBI activity in UK primary healthcare following the introduction of P4P schemes, measuring the impact on delivery at GP level remains challenging due to the deficiency of the available Read code data across a number of quality dimensions.