Mohammed Salameh, MD
Assistant Professor
Childrens Hospital of San Antonio
San Antonio, Texas
Disclosure information not submitted.
Jennifer Joiner, PNP
Assistant Professor
Baylor College of Medicine, United States
Disclosure information not submitted.
Michael Scheurer, PhD
Professor
Baylor College of Medicine, United States
Disclosure information not submitted.
Utpal Bhalala, MD, , FCCM
Pediatric Intensivist, Associate Professor, Research Advisor
Driscoll Children's Hospital
Corpus Christi
Disclosure information not submitted.
Title: Development of An Innovative Fluid Assessment Scoring Tool in the Pediatric Critical Care Unit
Introduction: Fluid overload is associated with poor outcomes in the critically ill. Monitoring fluid balance (FB) is important yet challenging in the pediatric intensive care unit (PICU). Currently used, conventional methods of fluid assessment are either not feasible and/or unreliable. Also, there is a lack of composite score or clinical algorithm which uses different parameters of fluid status. We aim to develop a fluid assessment scoring tool (FAST) to objectively assess the fluid status of children in the PICU.
Methods: We included critically ill patients aged 0-21 years presenting to our PICU between December 2020 to March 2021 and requiring invasive or non-invasive mechanical ventilation, inotropes, diuretics and/or renal replacement therapy. We assigned score 0-3 each for fontanelles, eyes, liver exam, and daily weight changes, fluid balance, urine output, blood urea nitrogen (BUN), creatinine, BUN/creatinine ratio, and chest x-ray findings. We also assigned score 0-3 for thoracic fluid content (TFC) as measured by noninvasive ICON monitor®. First, a nonparametric Receiver Operating Curve (ROC) analysis was used to determine potential cut points for the FAST using the TFC as a “gold-standard”. Cut points were selected based on a balanced sensitivity vs. specificity. Next, a 10-fold cross-validation area under the curve (AUC) for ROC analysis was performed to estimate the best performing cut point.
Results: In our study, 32 patients contributed 118 encounters. Median (IQR) age was 78.7 (9.75-84) months and 62.5% patients were males. The Median (IQR) FAST score was 4 (2-6). The median (IQR) TFC was 44.5 (29-57.25). Of the 118 encounters, 78 (66%) exhibited abnormal TFC. The two best potential FAST cut points were ≥3 (sensitivity=92%; specificity=68%) and ≥4 (sensitivity=80%; specificity=80%). On cross-validation, FAST≥3 (AUC=0.79) performed slightly better than FAST≥4 (AUC=0.78).
Conclusion: FAST is a reliable tool for assessing the overall fluid status of children in the PICU. FAST≥3 and ≥4 cut points correlated with fluid overload in the PICU.