Conclusions
This cohort study demonstrated that introduction of software to obtain reportable data may lead to significant changes in intervention potentially without patient benefit. It suggests that all units should review the software they are using and assess whether the references fit their population or introduce significant bias that alters clinical behaviour and potentially increases clinical harm. The study recommends that for the Plymouth population the UK90 birthweight reference would be the best fit, but it does not currently have the user friendly tools needed for everyday clinical use