Abstract
Objective: A retrospective study was conducted to elucidate the clinical significance of leukemoid reaction (LR) in the general ICU. Methods: After analyzing the data of all ICU admissions during the latest 4-year period, we enrolled the patients for this study whose WBC count in peripheral blood exceeded 30,000 /μl during the ICU stay. The medical records were reviewed for their underlying disease, laboratory findings and 6-month mortality rate. Result: LR occurred in 46 patients, aged 41-93 yr, corresponding to 1.9% of all ICU admissions. They were divided into two groups: a group with infectious diseases (infection group) and another group with non-infectious diseases (non-infection group). There were 31 patients in the infection group including 24 cases by which caused bacteria were revealed, while the non-infection group comprised 15 patients, including six intraabdominal bleeding, four post-cardiopulmonary resuscitation and five other diseases. The maximal WBC (Max. WBC) count, duration of LR, concurrent malignancy and steroid requirement were significantly higher in the infection group than in the non-infection group, whereas neither group showed statistical differences in the gender distribution or catecholamine requirement. A higher percentage of subjects in both groups (87%) revealed a “shift to the left” of neutrophil in the differential WBC. Of the 46 patients with LR, 25 died (16 from the infection group and 9 from the non-infection group), indicating a 6-month mortality rate of more than 50% in each group. No correlation was found between the high mortality rate and the Max. WBC count or duration of LR. Conclusion: The overall incidence of LR in our ICU admissions was approximately 2%. LR was associated with an increased mortality rate (>50%), irrespective of the underlying disease with an infectious or non-infectious cause. Understanding the clinical significance of LR is useful for intensive care management.