The identification of risk factors for critically ill patients with acute fever and formulation of activation criteria to alert outpatient clinic doctors

Rationale, aims and objectives  Acute fever is the most common early clinical symptom of many critical illnesses with a high mortality rate. It is necessary to identify patients with severe acute fever early and accurately. The aim of this study is to identify risk factors for critically ill outpatients with acute fever and formulate activation criteria of adult fever state score (AFSS) to alert outpatient clinic doctors. Methods  Retrospectively, 357 adult patients with acute fever were divided into two groups: 180 patients with a severe state and 177 patients with a mild state. Logistic regression was used to determine risk factors for the severe state. Risk factors were weighted and an AFSS was formulated. A receiver operating characteristic (ROC) analysis of weighted cumulative scores was performed to evaluate the diagnostic accuracy of AFSS, and the kappa test was used to confirm diagnostic reliability. A χ2-test for trend was applied to determine the relevance between AFSS and admission rate and in-hospital mortality. A Kruskal–Wallis test was used to examine the relationship between AFSS and length of stay. Results  Risk factors for state included: old age, long fever course, past medical history, abnormal temperature, abnormal respiratory rate, abnormal heart rate, abnormal mean arterial pressure and abnormal peripheral white blood cell count. The area under the ROC curve of AFSS was 0.964 and ≥8 points predicted severe state; the Kappa value was 0.801. With an increase in score, there was an increase in admission rate, mortality rate and length of stay. The forecast performance of AFSS was superior to the modified early warning score. Conclusions  The AFSS has high diagnostic accuracy and reliability for the early identification of patients with severe acute fever.