In recent years, there has been increasing interest in integrating evidence using network meta-analysis. Unlike conventional meta-analysis that integrates evidence from direct comparisons, network meta-analysis is a methodology that enables indirect comparison using a network of direct comparisons between treatment groups. In this paper, we discuss how the obtained results should be interpreted and the care required based on a network meta-analysis study comparing the onset of immune-related pneumonitis caused by immune checkpoint inhibitors in lung cancer. In particular, we introduce the significance of and evaluation methods for heterogeneity, similarity, consistency, and publication bias, and explain why paying attention to these aspects is critical. Software for application, textbooks for reference, and guidelines will also be introduced.