Introduction: Identification of early hemodynamic changes ensures optimal patient outcomes. We examined if a second-generation, electronic health record-based pediatric hemodynamic software system with semi-automated early warning Situational Awareness Vital Electronic Scout (SAVES) identified actionable warning levels with greater accuracy. Methods: Using a retrospective medical records review, hospitalized nonintensive care children’s encounters were included if they had first-generation warning system Pediatric Early Warning System (PEWS) software data. Clinicians manually entered data and warning levels were automatically created. An analyst applied PEWS data into the SAVES software that automatically included key patient characteristics from health records. Using descriptive analyses, PEWS and SAVES data were assessed in 3 ways: all data points, the first data point, and one randomly selected data point per encounter. Results: In total, 693,962 PEWS data-point rows from 43,505 encounters among 26,131 unique in-patients were included. Of patients, 20% were transferred to intensive care during their hospital stay. Over 90% of the time points of data using PEWS and SAVES were equally accurate based on warning category level; however, in 6.0 to 8.86% of data points, the warning level varied, with SAVES prompting higher warning levels in all 3 assessment methods. Compared with the PEWS assessment scores, SAVES scores were 3.2 times (first record) to 13 times (all data points) more likely to have higher warning levels. Discussion: The SAVES system identified higher early warning levels than PEWS. Conclusions: Accurate data, collected efficiently, facilitates recognition of subtle clinical changes before status deterioration.
Anderson et al. (2026) studied this question.