Abstract
Introduction
Pediatric Early Warning Scores (PEWS) facilitate the identification of non-ICU pediatric patients at risk for deterioration. Limited studies exist to describe the utility of Burn specific PEWS (bPEWS) in the early identification of clinical vulnerability. The purpose of this retrospective chart review was to associate a bPEWS assessment value with a need for elevation to an ICU level of care for the burn-injured child.
Methods
A retrospective chart review of all non-mechanically ventilated (nMV) pediatric patients admitted to the burn service from July 2013 to May 2016 (n=709). Data included bPEWS scores, age, total body surface area (TBSA) burn and hospital length of stay (LOS). A pediatric level of care (PLC) designation, instituted as a model of care delivery following this study, utilizes bPEWS to categorize patients by acuity. Patients are grouped by their highest bPEWS: a) unstable (bPEWS > 8; b) watchers (bPEWS 5–7); and c) stable (bPEWS < 5). This study retroactively utilizes this framework.
Results
709 patient charts yielded 12,642 bPEWS data points. 37 patients (0.53%) scored > 8 during their hospital stay. Patient age was not statistically significant, a=2.4 years, b=3.25, and c=2.3(F=0.64, p = 0.53). More secondary diagnoses were present in the a) unstable cohort (59%) than either the b) (26. 6%) or c) (20%); (c2 = 6.3, p = 0.02 and p < 0.01). There was a statistically significant difference in the number of patients in the unstable cohort versus the watcher and stable cohorts combined (c2 = 13.21, p < 0.01). Patient transfer to the pediatric ICU (PICU) occurred in 10.8% of the a) unstable group and none of the watcher or the stable cohorts (b, c), (p=0.02). Pediatric critical intensivist consults occurred in 19% of the a) unstable patients but not in either the watcher or the stable patients (p < 0.01). The average LOS was 18.1 days in the a) unstable group, 9.41 days in the b) watcher group, and 6.06 days for the c) stable group, (F=19.20, p < 0.01). TBSA burn was larger for the unstable group (12.5%), versus 5.74% for watcher patients, and 2.5 % for stable patients, (F=9.70, p < 0.01). On average, the peak bPEWS scores occurred hospital day 3.27 in the unstable group, 2.58 in the watcher group, and day 1.88 in the stable status group, (NS) (F=0.88, p = 0.42). There were no mortalities.
Conclusions
This retrospective review captures the infrequent experience of significant clinical deterioration in the nMV pediatric burn population reflected through the bPEWS lens. There appears to be a relationship between high bPEWS scores, burn size, presence of a secondary diagnosis, and increasing LOS. This study supports the designation of these patients to higher levels of care. This bPEWS driven paradigm presently results in adjusted nursing staffing ratios, frequency of assessment, and mandated collaborative medical practice patterns.
Applicability of Research to Practice
Directly Applicable.