Evaluation of the effect of disease severity predictor scales on hospitalization and determination of the severity level of the disease in patients brought to pediatric emergency department with fever in childhood

Objective: Fever constitutes a significant proportion of emergency department admissions in pediatric patients. In our study, we aimed to retrospectively evaluate the severity and hospitalization decisions of patients admitted with fever with scoring systems. Materials and Methods: Between January and July 2020, 602 patients who presented to the pediatric emergency department at Çukurova University Faculty of Medicine with the complaint of fever and in whom at least one laboratory test or radiologic imaging examination was requested were included in the study. The information of the patients included in the study was collected and analyzed retrospectively. Demographic characteristics, laboratory and radiology test results, clinical findings, Pediatric Early Warning Score (PEUS), Modified Pediatric Warning Score (mPEUS), Pediatric Hospitalization Risk Scoring II (PRISA II) and Liverpool Rapid Sequential Organ Failure Score (LqSOFA) were calculated retrospectively. Medical treatments, length of hospitalization, intensive care unit stays and mortality were analyzed. Results: A total of 602 patients, 258 (42.9%) female and 344 (57.1%) male, were included in our study. The mean age of the patients was 67.7±54.7 months. The mean duration of fever was 2.5±3.6 days and the mean temperature measured at home was 38.4±0.6°C. The most common accompanying complaints in patients presenting with fever were oral intake disorder in 47% and cough in 41%. The most common complaint in hospitalized patients was respiratory distress. In our study, no significant correlation was found between high fever and hospitalization. 377 (62.6%) of the patients received outpatient treatment, while 225 (37.4%) were hospitalized. Intravenous fluid support was the most common treatment in 454 (75.4%) patients. Hospitalization was required within 3 days in 231 (38.4%) of the patients. Among the hospitalized patients, 185 (82.2%) were admitted to the ward and 40 (17.8%) were admitted to the intensive care unit. PEUS, mPEUS, LqSOFA and PRISA II scores were significantly higher in hospitalized patients compared to non-hospitalized patients. The sensitivities of LqSOFA, PRISA II and mPEUS scores were 30.22%, 65.78% and 82.22%, respectively, and the best diagnostic test performance was PEUS score with a sensitivity of 93.33%. The cut-off values of PEUS, LqSOFA, PRISA II and MPEUS scores for hospitalization findings were determined as 6-5-0.5-8 and 5.5, respectively, while the cut-off values for intensive care unit hospitalizations were determined as 6-1-12-8, respectively. When the relationship between PEUS, LqSOFA, PRISA II and MPEUS scores and intensive care hospitalization was evaluated, the sensitivity was determined as 85%, 52.5%, 95% and 57.5%, respectively. When the treatment outcomes of the patients were examined, it was observed that 4 (0.7%) patients were exited, 587 (97.5%) patients were discharged, and 11 (1.8%) patients were discharged due to refusal of hospitalization. Conclusion: In recent years, with the increase in emergency department visits and the necessity of widespread use of triage systems, the effectiveness of scoring systems used in the management of children with fever, one of the most common reasons for visits to pediatric emergency departments, was investigated. The results obtained show that especially PEUS and MPEUS scoring systems make significant contributions to the emergency department management of patients. However, LqSOFA and PRISA II scoring systems were also found to be useful to some extent. There is a need for prospective multicenter studies to determine the hospitalization and intensive care needs and severity of patients admitted to the pediatric emergency department. Key words: Pediatric emergency, fever, pediatric early warning score