Validation of the Pediatric Early Warning Score in the Pediatric Specialty Population

Introduction: Up to 3% of pediatric inpatients experience a code event with high mortality (Tucker, Brewer, Baker, Demeritt, & Vossmeyer, 2009). Up to 14% of these arrests happen outside the intensive care unit (Demmel, Williams, & Flesch, 2010). Up to 73% of codes or other serious adverse events are frequently preceded by a period of clinical deterioration and physiological instability that if noted, provides opportunity for intervention, with improved patient outcomes. This is known as failure to rescue, with rates of 703/1000 (Miller & Zhan, 2004). Focus shifted towards early recognition of clinical deterioration (Demmel, Williams, & Flesch, 2010). Monaghan (2005) developed an objective Pediatric Early Warning Scoring (PEWS) Tool, which scores the three clinical dimensions of behavior, cardiovascular, and respiratory. Each dimension is scored from 0-3, with a possible total score of 9. A critical PEWS score is a score of 3 in any one dimension or a total score of > 4 and is suggestive of clinical deterioration. An action algorithm with escalating interventions was developed for use in combination with the scoring tool (Demmel, Williams, & Flesch, 2010). This study used a modified version of the PEWS tool. The modified PEWS tool has been validated in the general pediatric population with a sensitivity of 62% and specificity of 89% (Skaletzky, Raszynski, & Totapally, 2011), but never validated in the pediatric specialty population. The purpose of this study was to validate the tool in this population.

Methods: Data was gathered via retrospective chart review using the modified PEWS. Inclusion criteria: charts of any renal, oncology, or hematology patient ages 1 month - 18 years admitted to the Pediatric Special Care Unit with a code, rapid response team (RRT) call, or transfer to the Pediatric Intensive Care Unit. Charts were identified through review of admission logs from the PICU, code logs, and RRT logs from January 1, 2011 - December 31, 2013. Chart reviews conducted via single investigator. The charts were scored every 4 hours using the PEWS tool for 24 hours preceding the code, RRT, or transfer. Inter-rater reliability of the PEWS tool established through having 3 expert pediatric nurses score 10% of the charts. Data points included age, medical service, type of event, reason for event, whether or not there was a critical PEWS, the mean PEWS score, time from first critical PEWS to event, time from first documented intervention to event, and difference between these two times. The study is currently in the data analysis phase.

Implications for practice: The original PEWS tool was shown to identify clinical deterioration in patients over 11 hours prior to a code event (Akre et al., 2010). When surveyed, staff and providers agreed the original PEWS tool and action algorithm enhanced communication among team members, with intervention less dependent on level of experience (Demmel, Williams, & Flesch, 2010). Use of the tool and action algorithm also increased the number of days between code events, from 299 to 1,053. This improvement in patient outcomes was sustained.