Introduction Early Warning Scores are associated with an increase in early detection of illness (true positive), high scores without clinical deterioration (false positive) and clinician over-ride or alarm fatigue that leads to missed detection (false negative).

Aim: To establish the prevalence of false positive and false negative alerts generated by the Paediatric Early Warning Score (PEWS)

Methods Data collection occurred on three different days selected purposefully for when the hospital would have the maximum number of patients and a representative case mix of medical and surgical patients. The PEWS score and physiological parameters for the patients were entered on a secure database.

For patients with high PEWS (>/= 9), notes were reviewed and nursing staff were contacted for clarification of the action taken and bedside management and escalation plan. This is similar to routine Trust wide safety audits.

Results: 248 patients’ PEWS charts from 300 inpatient beds were reviewed over 3 days. 87% of the patients had observations in their PEWS chart, with a physiological value against their PEW score. PEWS were scored more >9 in 2%, 5–8 in 13% and 1–4 in 85% of patients. Majority of the patients with high PEWS were from the cardiac, respiratory and post-operative units.

In 46 and 57% of the patients with PEWS >/= 9 and 4–8 respectively, inappropriate response was taken. During this period 4 of the patients were admitted to the PICU from the wards.

Conclusion 2% of patients had score >/= 9 and so even if these are false positives it is a small burden.

Clinicians do override the PEWs recommendations. Is this due to poor calibration of the score or clinician disbelief?