Aim Traditional adverse event reporting systems under record the true incidence of system errors. Granular detail is often lost from organisation memory due to the time lag for investigation and feedback, disengaging reporters and dissipating the opportunity to learn.

We describe a simple method to record errors that keeps the clinical team connected, retaining ownership of system safety, allowing rapid local analysis and focus on system improvements.

Method Over three weeks, ward rounds on renal patients on a mixed paediatric subspecialty ward included a consultant-led question about system safety issues for each in-patient. Details of errors were recorded in casenotes under ‘Patient Safety’ then categorised using the mnemonic ‘MR PICO’ (table 1). Interrogation of the system as to why and how each error occurred can be performed at the time. Comparison was made with existing reporting tools.

Results Daily CAPS were reliably recorded in all in-patients casenotes during the test period. 63 system errors leading to 18 harm events were recorded over 101 patient days (table 2). 98% of errors were judged preventable. Every in-patient experienced at least one system error. Nine quality improvement projects were generated by learning from selected errors.

Compared to CAPS, electronic trust reporting system (DATIX) and paediatric global triggering tool (pGTT) captured only 3-11% of errors.

Conclusion Error rates captured by CAPS accurately reflect the true incidence rather than relying on the willingness of staff to report to a remote system. This transparent mechanism secures a daily system safety conversation that we envisage will affect culture change towards an increasingly safer system of care.