Impact of systemic corticosteroids on the clinical outcomes in patients with severe community-acquired pneumonia: a multi-center retrospective study / 中华危重病急救医学

ABSTRACT

Objective To assess the impact of short-term, low-dose systemic glucorticosteroids treatment on the clinical outcomes in patients with severe community-acquired pneumonia (SCAP). Methods A multi-center retrospective study was conducted. Data of patients hospitalized with SCAP in five teaching hospitals from Beijing, Shandong and Yunnan Provinces from January 1st, 2013 to December 31st, 2015 were reviewed. Patients were divided into steroids group and non-steroids group according to whether treated with glucorticosteroids during the disease course or not. Data of patients were reviewed, including gender, age, underlying disease, blood routine, biochemical examination and radiology findings (the worst value was recorded if there were more than one value), supportive treatment, complications (hyperglycemia needing insulin treatment and gastrointestinal bleeding) and clinical outcomes [early (0-3 days) treatment failure, late (4-14 days) treatment failure and 30-day mortality, treatment failure was defined as one of the followings needing noninvasive or invasive ventilation, needing vasopressor use or death]. Univariate and multivariate Logistic regression was performed to evaluate the impact of short-term, low-dose systemic glucorticosteroids on the clinical outcomes in SCAP patients. Results Overall, 3 561 immunocompetent adult and adolescent patients with community-acquired pneumonia (CAP) were screened, 132 SCAP patients were entered into final analysis, including 24 patients in steroids group and 108 patients in non-steroids group. The patients in steroids group were prescribed with methylprednisolone (0.6±0.1) mg·kg-1·d-1 for (4.0±1.7) days. Compared with patients in non-steroids group, patients in steroids group showed younger age [years old 70.5 (59.0, 75.0) vs. 80.0 (76.0, 85.0)], less frequency of male [41.7% (10/24) vs. 72.2% (78/108)], less comorbidities with cardiovascular [16.7% (4/24) vs. 42.6% (46/108)] and cerebrovascular disease [0% (0/24) vs. 40.7% (44/108)], less confusion [16.7% (4/24) vs. 40.7% (44/108)]; more frequency of chronic obstructive pulmonary disease [COPD, 41.7% (10/24) vs. 13.0% (14/108)], asthma [25.0% (6/24) vs. 1.9% (2/108)], chronic hepatic disease [8.3% (2/24) vs. 0% (0/108)] and respiratory rate≥30 times/min [33.3% (8/24) vs. 9.3% (10/108)] with significant differences (all P < 0.05), the proportion of guideline-based empirical antimicrobial therapy, early needing noninvasive ventilation, late gastrointestinal bleeding, early and late hyperglycemia needing insulin treatment were higher in steroids group than non-steroids group [50.0% (12/24) vs. 21.3% (23/108), 33.3% (8/24) vs. 7.4% (8/108), 20.8% (5/24) vs. 4.6% (5/108), 20.8% (5/24) vs. 1.9% (2/108), 37.5% (9/24) vs. 2.8% (3/108), all P < 0.05]. Adjusted by gender, age, comorbidities and empirical antimicrobial therapy, Logistic regression confirmed short-term, low-dose systemic glucorticosteroids was associated with higher risk for vasopressor usage [odds ratio (OR) = 3.369, 95% confidence interval (95%CI) = 1.369-6.133, P = 0.035], hyperglycaemia needing insulin treatment (OR = 4.738, 95%CI = 1.890-8.652, P = 0.017) in late stage and 30-day mortality (OR = 2.187, 95%CI = 1.265-4.743, P = 0.002). Conclusion Adjunctive treatment with short-term, low-dose systemic glucorticosteroids worsen the clinical outcomes and should not be used to SCAP patients routinely.