he prospect of improved clinical outcomes and more efficient health systems has fueled a rapid rise in the development and evaluation of AI systems over the last decade. Because most AI systems within healthcare are complex interventions designed as clinical decision support systems, rather than autonomous agents, the interactions among the AI systems, their users and the implementation environments are defining components of the AI interventions' overall potential effectiveness. Therefore, bringing AI systems from mathematical performance to clinical utility needs an adapted, stepwise implementation and evaluation pathway, addressing the complexity of this collaboration between two independent forms of intelligence, beyond measures of effectiveness alone 1 . Despite indications that some AI-based algorithms now match the accuracy of human experts within preclinical in silico studies 2 , there

In contrast to some other studies, many of our patients were aware of medication risks and were prepared to accept them provided benefits were seen to be high. Existing clinic patients were more aware of risks and benefits, and reported higher compliance levels than new patients, possibly as a result of the hospital education programme. Future studies should evaluate the effects of the programme more systematically.

Three cases of childhood acute lymphatic leukaemia masquerading as juvenile chronic arthritis are presented. All had symptoms and signs for at least 4 months before leukaemia was diagnosed and in two the full blood count was normal at presentation. The importance of a high index of suspicion is emphasized, particularly if the white cell count is low.

No abstract

Rheumatology patients who have not yet experienced nurse prescribing are, in general, positive about nurses adopting this role. It is important that nurses provide appropriate information about the prescribed medicines, in a form that can be understood.