Predictors Of Intensive Care Unit Admission For Inpatient With Respiratory Infection: A Retrospective Study

OBJECTIVE Respiratory infection is one of the most commondiseases in childhood, pneumonia is the major cause of death for childrenunder5years old. Severe pneumonia may have poor outcome because itgets worse quickly, the classic performance is covered and so on. For now,the beds in intensive care unit (ICU) are in short supply, and the cost thereis extremely expensive. Therefore, it is important to assess the severity ofthe respiratory infection. However, there are few reports about thepredictors of ICU admission for inpatient with respiratory disease at earlystage. So we conducted a retrospective research to explore factors forinpatient in our hospital who need to be transferred from children’srespiratory department to intensive care unit (ICU).METHODS We conducted a retrospective research of hospitalizedchildren who was transferred to ICU from children’s respiratory departmentas case group, and children in respiratory department as control group fromJune2009to October2013. Data collections included patients’ demographic,clinical characteristics, investigations and pathogens. RESULTS150children are enrolled in the case group, the majordiagnosis for case group is severe pneumonia (77.3%), the case group haspoor outcome than control group (p<0.0001). The children in case groupis younger than the control group, other differences between these twogroup are as follows: patients’ demographic (non-native resident,unvaccinated, ever respiratory infection), clinical characteristics (fever,wheezing, temperature, respiratory rate, heart beat), investigations(Hemoglobin, white blood cell [WBC], neutrophil percentage [N%]，Creactive protein [CRP], hyponatremia) and pathogens (adenovirus [ADV],sputum culture, klebsiella pneumoniae and Haemophilus influenzae arepositive) and complication (bronchial dysplasia). In the multivariablelogistic regression model predicting intensive care unit admission, thesignificant associated factors were: WBC (odds ratio [OR]33.21;95%confidence interval [CI]8.42–131.13), ADV (OR17.17.21;95%CI1.86–159.05), bronchial dysplasia (OR17.20;95%CI2.86–103.30),tachycardia (OR12.68;95%CI1.861-159.053), non-native resident (OR6.49;95%CI2.35–17.93), sputum culture positive (OR3.53;95%CI1.51-8.26), N%(OR1.12;95%CI1.08-1.16), young age (OR0.50;95%CI0.33–0.76).56%children in case group had received mechanicalventilation. In the compare of the poor outcome and the better outcome inthe case group, we find that the need for mechanical ventilation of the poorgroup is1.71times as the better group, and the anemia occurred to the poor group is1.88times as the better group.CONCLUSIONSPredictors of ICU admission for children with respiratory infection areas follows: non-native resident, young age, tachycardia, WBC and N%risein blood routine test, ADV and sputum culture positive, bronchial dysplasia.The patient in ICU has poor outcome if the test of respiratory virus ispositive, anemia is found and mechanical ventilation is needed.