Death predictors in children hospitalized in emergency: A retrospective cohort study

Abstract

Background. Incorporation of death risk factors in case of urgent pathology in children remains a matter of argument. Objective. Our aim was to determine death predictors in children hospitalized in emergency. Methods. In a retrospective cohort study, we analyzed the data of inpatient's medical records (form 003/u) of resuscitation patients aged 0-17 years hospitalized in the Rostov region in 2006-2017. We recorded a favourable outcome of the underlying disease in case of patient discharge from the hospital and death in case of death in the hospital. Results. We studied the data of 151 children (boys-61%), median age 10 (2; 36) months. Among the predictors of death in a hospital (90 out of 151 patients, 60%), we determined the disease duration up to 24 hours (odds ratio, OR, 2.1, 95% confidence interval, CI, 2.1-4.1), self-referral hospitalization (OR 3.0, 95% CI 1.1-10.4), hospitalization in a moderate (OR 19.6, 95% CI 12.6-337.4) or extremely critical condition (OR 103.1, 95% CI 6.2-1, 718.9), sepsis/generalized infection (OR 5.8, 95% CI 2.3-15.0), decrease in diastolic blood pressure (BP) ≥ 20% of the reference value (OR 7.0, 95% CI 1.3-38.4), score ≤ 14 points by the Glasgow coma scale (OR 10.6, 95% CI 4.1-24.6). Predictors of death at the resuscitation phase (72 of 133 patients, 54%) included cerebral dysfunction (OR 3.8, 95% CI 1.5-9.1), shock (OR 3.1, 95% CI 1.3-7.0), decrease of ≥ 20% in the reference value for respiratory rate (OR 24.2, 95% CI 3.1-186.2), heart rate (OR 29.0, 95% CI 1.7-497.2), systolic blood pressure (OR 5.2, 95% CI 1.9-14.7), diastolic blood pressure (OR 5.4, 95% CI 1.9-15.3), increase of ≥ 20% in the reference value for glucose level (OR 2.1, 95% CI 1.1-4.5), and creatinine (OR 4.3, 95% CI 1.5-12.6). Conclusion. We identified the predictors of death in children hospitalized in emergency. They can be used to consistently assess the risk of death in a hospital.