Pediatr Surg Int. 2024 Dec 28;41(1):51. doi: 10.1007/s00383-024-05945-0.
ABSTRACT
BACKGROUND: Burns in children are often complex injuries, leading to prolonged length of stay (LOS) and significant morbidity. LOS in pediatric intensive care units (PICUs) is a key measure for evaluating illness severity, clinical outcomes, and quality of care. Accurate prediction of LOS is vital for improving care planning and resource allocation. There is limited data for predicting LOS in severely burned children in PICU. This study aims to identify significant factors associated with prolonged PICU stays and offer a simple LOS-predicting model.
METHODS: This historical cohort study included all patients < 18 years, admitted for severe burn injuries to Israel’s largest tertiary hospital PICU, from 2015 to 2020. Statistical analyses were conducted to identify factors linked to prolonged LOS and a predictive model was designed.
RESULTS: The study included 39 pediatric burn patients and identified several factors associated with longer PICU stays. Patients with LOS of ˃7 days (i.e.-“long LOS”) had significantly higher %TBSA (33.11 ± 17.87% vs. 16.67 ± 7.98%, p < 0.001. During the first 24 and 48 h, the “long LOS” group had lower minimal systolic blood pressure (SBP) (70.67 ± 17.49 mmHg vs. 84.38 ± 16.73 mmHg, p = 0.015 and 69.39 mmHg ± 16.44 vs. 81.10 mmHg ± 19.67, p = 0.018). Although serum lactate levels were higher in the “long LOS” group, the difference was not significant, and platelet counts in this group were significantly lower during the first 48 h (184 K/µL vs. 264.5 K/µL, p = 0.003). A predictive model based on %TBSA, SBP, lactate, and platelet count was developed, demonstrating 100% specificity and positive predictive value for predicting LOS over 7 days in severely burned children.
CONCLUSIONS: Key clinical indicators at PICU admission in severely burned children were associated with LOS > 7 days. The resulting predictive model, although requiring further validation in multi-site studies, offers a promising tool for enhancing care planning in this population.
PMID:39731634 | DOI:10.1007/s00383-024-05945-0
