Efficacy of Multiple Interventions for Moderate to Large Traumatic Tympanic Membrane Perforations: A Systematic Review and Network Meta-Analysis.

OBJECTIVE

This study aimed to evaluate the comparative efficacy of various therapeutic interventions for moderate to large traumatic tympanic membrane perforations (TTMP) through a network meta-analysis. Different treatment strategies were compared with respect to tympanic membrane healing rate, healing time, hearing improvement and incidence of otorrhea, providing evidence-based support for clinical decision-making.

METHODS

A systematic search of multiple databases, including PubMed, Web of Science and Embase, was conducted to identify randomised controlled trials (RCTs) involving patients with moderate to large TTMP. A Bayesian hierarchical model was utilised to perform the network meta-analysis, assessing the relative effects of different interventions. Effect sizes, including relative risk (RR) and mean difference (MD), were computed for each intervention, with all analyses reported alongside 95% confidence intervals (CIs).

RESULTS

A total of 13 studies involving 1285 patients were included in the analysis. The findings suggested that fibroblast growth factor-2 (FGF-2) demonstrated the highest efficacy in enhancing tympanic membrane healing rates and reducing healing times, particularly when combined with gelfoam patching. Additionally, ofloxacin otic solution appeared to have positive effects on healing rates. Conversely, edge approximation ranked lowest among all interventions, suggesting limited effectiveness in treating moderate to large perforations.

CONCLUSION

This network meta-analysis provides hypothesis-generating evidence that active intervention may be superior to watchful waiting for moderate to large TTMPs. FGF-2-based strategies, particularly when combined with physical scaffolds, show promise in improving healing outcomes. Additionally, ofloxacin otic solution may offer benefits in controlling local infections and optimising the middle ear microenvironment. In contrast, edge approximation appears less effective in this population. However, given the geographic homogeneity, limited number of studies and methodological constraints, these findings should be considered exploratory and require confirmation in large-scale, multicentre RCTs with standardised protocols and robust blinding procedures.