Changing the behaviour of healthcare professionals using theory based, computer-delivered interventions

摘要

Non-adherence to clinical guidelines has been identified as a consistent finding in general practice. The purpose of this research was to develop and evaluate theory-informed, computer-delivered interventions to promote the implementation of guidelines in general practice, which GPs viewed as feasible and acceptable. The intervention aimed to promote guideline adherence for antibiotic prescribing in respiratory tract infections, and adherence to recommendations for secondary stroke prevention. An intervention development study involved the creation of computer-delivered prompts using aspects of social cognitive theory, and drawing on nationally recommended standards for clinical content. Prompts were presented to GPs during interviews, and iteratively refined based on feedback. GPs reported being more likely to use prompts if they were perceived as offering support and choice, as opposed to being an enforcement method. The prompts were then entered into a trial (not reported) and two process evaluation studies were conducted with GPs who had taken part in the trial. A qualitative evaluation study involving interviews with GPs, revealed that the prompts were perceived as useful and acceptable in practice, but GPs who had not been informed of the prompts appearance reported being less likely to engage with them. A quantitative evaluation study involved a questionnaire consisting of theory based measures and an intervention evaluation measure. GPs were satisfied with the usability of the prompts, and intervention group GPs reported higher levels of self-efficacy in managing patients according to guidelines compared to control group GPs. Overall the intervention was viewed as feasible and acceptable. A key characteristic of an acceptable computer-delivered intervention appears to be that it should be perceived as a useful tool supporting GP practice. However, conclusions of the evaluation were limited by a small and potentially non-representative sample of trial GPs.