Systematic review and meta-analysis of the safety of erythromycin compared to clarithromycin in adults and adolescents with pneumonia

Abstract

Macrolides are recommended for the treatment of community-acquired pneumonia (CAP). It is debatable whether erythromycin is associated with more adverse drug reactions (ADRs) compared to clarithromycin, and both are recommended in clinical practice guidelines. This meta-analysis aim is to compare ADRs in CAP patients treated with erythromycin versus clarithromycin. Two investigators independently searched PubMed, EMBASE and Cochrane Library databases through Feb 07, 2019. Randomized-controlled trials (RCTs) comparing ADRs of monotherapy with erythromycin versus with clarithromycin in adults or adolescents with CAP were included. We estimated risk ratios (RRs) with 95% confidence intervals (CIs) using random-effects models and evaluated heterogeneity (I²). Bias risk was assessed using the Cochrane risk of bias tool for RCTs. Five RCTs (total of 693 patients) were included. A significantly higher discontinuation rate due to ADRs was found with erythromycin compared with clarithromycin (RR, 4.347; 95% CI, 2.506–7.539; p < 0.001; I²=0%). Overall, ADRs occurred more significantly with erythromycin compared with clarithromycin (RR, 1.773; 95% CI, 1.423–2.209; p < 0.001; I²=0%). Gastrointestinal (GI) ADRs were higher with erythromycin (RR, 2.678; 95% CI, 1.791–4.006; p < 0.001; I²=5.835%). Restriction of analyses to double-blind RCTs did not change our findings. Based on meta-analysis of RCTs in adults and adolescents with CAP, erythromycin results in more overall ADRs and GI ADRs, as well as a higher rate of discontinuation due to ADRs. Therefore, given that erythromycin is not more effective than clarithromycin, erythromycin should not be selected unless other macrolides cannot be used.

Keywords:

• Erythromycin
• Clarithromycin
• Pneumonia
• Nausea
• Gastrointestinal

Compliance with ethical standards

No funding was received for development of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

KE conceived and designed the study, did the data analysis, and supervised all steps of this meta-analysis. AB, FW, MA, and ZA did the literature screening and data extractions. Completing the quality assessment of studies table and writing and editing of the initial versions of manuscript were completed by KE, AKT, AA, MAA, and BA. AH and AT critically revised the manuscript and discussed the idea generation.