Objective: To evaluate the effect of mild hypothermia on the incidence of post-stroke infection and explore the relationship between mild hypothermia and outcome of stroke patients by using propensity score matching.
Methods: Patients hospitalized in department of intensive care unit (ICU), neurology and neurosurgery in the First Affiliated Hospital of Guangxi Medical University due to stroke from March 2012 to April 2018 were retrospectively analyzed. According to whether or not mild hypothermia was provided, they were divided into the normal thermic group (NT group) and mild hypothermia treatment group (MHT group). The MHT group patients were matched with the NT group patients by the propensity score matching method at a ratio of 1:1. The observation period was within the first 7 days after admission. Baseline characteristics including age, gender, type of stroke, comorbidities, acute physiology and chronic health evaluation II (APACHE II) score and Glasgow coma score (GCS) on admission, surgical operation, dysphagia, invasive procedures and outcomes of these patients had been analyzed. The primary outcome was incidence of post-stroke infection, and the secondary outcomes included the time of initial infection (TII, the duration from stroke to initial infection), hospital mortality, sequential organ failure assessment (SOFA) at discharge, incidence of complications such as arrhythmia, coagulation dysfunction and multiple organ dysfunction syndrome (MODS).
Results: 201 stroke patients were enrolled, 41.8% (84/201) of whom underwent mild hypothermia. Comparison with NT group before matching, there were more males in MHT group (71.4% vs. 56.4%), the proportion of surgical operation, mechanical ventilation, deep vein catheterization and gastric catheterization were higher (78.6% vs. 54.7%, 84.5% vs. 39.3%, 90.5% vs. 37.6%, 98.8% vs. 70.9%), and so as incidence of infection (90.5% vs. 72.6%), in-hospital mortality (27.4% vs. 12.8%) and TII [hours: 62.00 (35.25, 93.00) vs. 42.00 (28.50, 69.50)]. All the differences were statistically significant (all P < 0.05). Fifty-three patients in the MHT group were matched with 53 patients in the NT group. After matching, there was no significant difference in 15 baseline characteristics between two groups. Significant differences in infection and hospital mortality between the MHT group and NT groups disappeared (92.5% vs. 88.7%, 22.6% vs. 26.4%, both P > 0.05), while TII of MHT group was longer than that of the NT group [hours: 62.00 (40.75, 92.25) vs. 40.00 (28.00, 63.00), P = 0.000]. There were no statistically significant differences in SOFA score or complications between the two groups either before or after matching.
Conclusions: Mild hypothermia had no significant effect on the incidence of post-stroke infection and hospital mortality, it could delay the occurrence of infection and provide longer duration of treatment.