Successful Implementation of a Pediatric Early Warning Score in a Resource-Limited Pediatric Oncology Hospital in Guatemala

25 Background: Hospitalized pediatric oncology patients are at high risk for clinical decline and mortality, particularly in resource-limited settings. Pediatric Early Warning Scores (PEWS) are commonly used to aid with early identification of clinical deterioration; however, these scores have never been studied in oncology patients in low-resource settings. We describe the successful implementation of a modified PEWS at Unidad Nacional de Oncología Pediátrica (UNOP), a national pediatric oncology hospital in Guatemala. Methods: The PEWS used at Boston Children's Hospital (BCH) was modified through key informant meetings at UNOP, adjusting for practice variations between the two hospitals. After an initial pilot of the tool, the PEWS was implemented in all non-ICU inpatient areas at UNOP (60 beds with about 2,000 admissions/year). During implementation, systems were created to monitor errors in calculating PEWS, patient transfers to a higher level of care, and high PEWS scores for ongoing quality improvement. Results: Hospital-wide implementation occurred over 6 months, when 113 nurses were trained in the PEWS tool and algorithm. Compliance with PEWS performance and documentation was 100% by the end of the implementation period, with 300 to 400 PEWS measured daily and less than 10% errors. Monitoring of PEWS results reports an average of 5 high PEWS per week with 30% transferring to a higher level of care. Among patients requiring ICU transfer, 86% had an abnormal PEWS prior to transfer, which is similar to results at BCH (90%). Staff surveys showed a high degree of satisfaction with PEWS (4.6/5) and minimal difficulty using the score (2.3/5) (n=67). Conclusions: We describe the successful implementation of a PEWS in a pediatric oncology hospital in Guatemala. This work demonstrates that PEWS is a feasible, well-accepted, and low-cost quality improvement measure in this resource-limited setting. We now plan to evaluate the effects of this implementation on patient care and outcomes. AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST: No COIs from the authors.