RESUMO
BACKGROUND: Periodic vital sign monitoring is commonly used to detect clinical deterioration. Incomplete assessment of Early Warning Score (EWS) and poor protocol compliance may negatively impact nurses' responsiveness to critical situations. OBJECTIVE: This study assesses the quality of recorded EWSs and the degree of overall protocol compliance and protocol compliance in adverse event patients, in a centre that is an early adopter of an EWS safety protocol. DESIGN: Retrospective single-centre cohort study. SETTING: General hospital wards of a tertiary referral university medical centre. PARTICIPANTS: 48,864 patients admitted to general wards between 2015 and 2018. METHODS: Vital sign and Modified EWS (MEWS) data were collected from the electronic health record. MEWS completeness was evaluated by determining the proportion of missing values for each vital sign. To evaluate protocol compliance, we assessed whether the elapsed time between consecutive MEWSs was within the protocollary time as dictated by the MEWS protocol. Outcome measures were overall MEWS completeness, and both overall protocol compliance and protocol compliance specifically in the 24 h before adverse events. RESULTS: All required items to calculate a MEWS were present in 76 percent of the 1,663,743 vital sign measurements. Overall protocol compliance was 62 percent. For the low, mid-range and high MEWS groups, protocol compliances were 67, 47 and 30 percent, respectively. Time-to-next-MEWS exceeded protocol-recommended timing by up to 9 h for a substantial amount of MEWSs in the mid- and high-range MEWS groups. In patients with adverse events, compliance to the MEWS protocol during the 24 h prior to the events was 66 percent in aggregate, ranging from 80% in low MEWS groups, 74% in the mid-range, and 49% in the high MEWS groups. In the high MEWS group, improvements in protocol compliance were primarily noted in five hours immediately preceding the event. CONCLUSIONS: EWS assessments were incomplete in one-quarter of measurements. Compliance to a MEWS safety protocol was generally low, specifically when the protocol mandated more frequent MEWS assessments. Although more MEWSs were recorded in patients with adverse events, the increase in vital sign measurements' frequency mostly occurred shortly before the event manifested. This finding suggests missed opportunities to detect clinical deterioration.
