© 2018 Elsevier Inc. Background: Early Warning Scores were introduced into acute hospitals in 2000. 99% of acute hospitals employ a EWS to monitor deteriorating patients with 97.9% of these linked to a referral protocol. Despite this high level of adoption, there has been little improvement in the recognition and response to deteriorating patients over the last decade. Objective: To explore the patterns of compliance with Early Warning Track and Trigger Tools. Design: A narrative review. Data sources: Electronic databases (Medline, CIHAHL, EmBase, the Cochrane library, the Centre for Reviews and Dissemination (CRD) and PROSPERO) were searched from 1 January 2000 to 5 July 2018. Titles, abstracts and full text papers were screened (two independent reviewers) against inclusion criteria and seven papers were included in the review. Data were extracted by one reviewer and checked by a second reviewer using a bespoke data collection sheet. Review methods: All papers were quantitative in design but demonstrated clinical and methodological heterogeneity therefore a meta-analysis was not possible. A qualitative approach was undertaken to synthesise findings using a framework analysis and narrative synthesis. Themes were identified, named, defined and reported according to outcome measure. Results: 7/27 papers representing over 3000 patients and 963,000 data points were included in the analysis. Reported studies were conducted in the United Kingdom (n = 4), Denmark (n = 2) and Amsterdam (n = 1). Three key themes were identified, early warning score calculation accuracy, monitoring frequency and clinical response. This review identifies poor compliance with the Early Warning Score (EWS) protocol in all three themes. There is significant scoring inaccuracy with omitted EWS, missing elements of the EWS and incorrectly calculated EWS. Adherence to monitoring frequency is poor with a higher EWS being associated with reduced compliance with the escalation protocol. There is inadequate compliance with the escalation element of the EWS protocol with concerning extended delays to clinical review. There is evidence of worsening clinical response with increasing EWS. Although significant improvement is demonstrated in clinical response with the use of electronic EWS protocols, non-compliance still occurs at all EWS stages. Conclusion: Compliance with EWS is poor but the cause is unidentified. Outcomes can only improve if staff complete the EWS fully, calculate the score accurately, monitor according to protocol and escalate according to clinical response. Social, environmental and professional behaviours that affect effective use of track and trigger tools should be explored to improve our understanding of suboptimal management of the deteriorating patient.