OBJECTIVETo evaluate the therapeutic effects of antibacterial agents, glucocorticoid and intravenous immunoglobulin (IVIG) in treating Mycoplasma pneumoniae(MP) infections.

METHODThe literature was screened by the inclusion and exclusion criteria after searching at Cochrane Library, Pubmed, Wanfang, CNKI, and Weipu databases. According to JADAD evaluation system, the relevant information in each included report from the literature was evaluated. The evidence-based analysis was performed for the therapeutic effects of macrolides, glucocorticoid, and IVIG in treating MP infections. Meta-analysis was conducted on the suitable literature by RevMan 5.3 software supplied by Cochrane collaboration. Descriptive analysis was conducted on the literature unsuitable for meta-analysis.

RESULT(1) Seven foreign RCT reports and 7 domestic RCT reports were included in the analysis of the therapeutic effect of macrolides. There was a high heterogeneity among the 7 foreign reports. Five of these reports showed no significant difference in clinical effects between macrolides and non-macrolide antibacterial agents. The forest plot analysis of antipyretic timing and cough duration in the domestic literature with complete indicators suggested that for azithromycin sequential therapy vs. erythromycin intravenous therapy, the mean difference of antipyretic timing was-1.10 (95% CI: -1.60,-0.60) and the mean difference of cough duration was-1.56 (95% CI: -2.10,-1.03). (2) Three foreign RCT reports and 5 domestic RCT were included in the analysis of glucocorticoid therapy. The JADAD scores of all the reports were 1. The basic therapy drug was macrolides. The results of sub-group analysis suggested that for the patients who used glucocorticoid early vs. the patients who used non-glucocorticoid therapy, the mean difference of antipyretic time was-1.77(95% CI: -2.44,-1.10) and the mean difference of cough duration was-2.47 (95% CI: -2.86,-2.08); for the patients treated with glucocorticoid at 10 days after onset of diseases vs. the patients received non-glucocorticoid therapy, the mean difference of antipyretic time was-3.41 (95% CI: -4.10,-2.73) and the mean difference of cough duration was-2.25 (95%CI: -4.38,-0.12). (3) Regarding IVIG, all the included reports were case study or case report. Most of the literature focused on severe Mycoplasma pneumoniae infection and those with extrapulmonary complications. The limited results suggested a trend of the shortening of disease process and improvement of clinical symptoms by IVIG.

CONCLUSIONThere was no exact evidence of the therapeutic effects of antibacterial agents in Mycoplasma pneumoniae infections. A trend of better therapeutic effect was inferred in macrolide antibiotics, especially azithromycin. The improvement of clinical symptoms was suggested with the usage of glucocorticoid as adjuvant therapy. IVIG as an adjuvant therapy is at an exploration stage.