Meta-Analysis of In Vitro Antimicrobial Capacity of Extracts and Essential Oils of Syzygium aromaticum, Citrus L. and Origanum L.: Contrasting the Results of Different Antimicrobial Susceptibility Methods

Diffusion methods, including agar disk-diffusion and agar well-diffusion, as well as dilution methods such as broth and agar dilution, are frequently employed to evaluate the antimicrobial capacity of extracts and essential oils (EOs) derived from Origanum L., Syzygium aromaticum, and Citrus L. The results are reported as inhibition diameters (IDs) and minimum inhibitory concentrations (MICs), respectively. In order to investigate potential sources of variability in antimicrobial susceptibility testing results and to assess whether a correlation exists between ID and MIC measurements, meta-analytical regression models were built using in vitro data obtained through a systematic literature search. The pooled ID models revealed varied bacterial susceptibilities to the extracts and in some cases, the plant species and methodology utilised impacted the measurements obtained (p < 0.05). Lemon and orange extracts were found to be most effective against E. coli (24.4 ± 1.21 and 16.5 ± 0.84 mm, respectively), while oregano extracts exhibited the highest level of effectiveness against B. cereus (22.3 ± 1.73 mm). Clove extracts were observed to be most effective against B. cereus and demonstrated the general trend that the well-diffusion method tends to produce higher ID (20.5 ± 1.36 mm) than the disk-diffusion method (16.3 ± 1.40 mm). Although the plant species had an impact on MIC, there is no evidence to suggest that the methodology employed had an effect on MIC (p > 0.05). The ID–MIC model revealed an inverse correlation (R(2) = 47.7%) and highlighted the fact that the extract dose highly modulated the relationship (p < 0.0001). The findings of this study encourage the use of extracts and EOs derived from Origanum, Syzygium aromaticum, and Citrus to prevent bacterial growth. Additionally, this study underscores several variables that can impact ID and MIC measurements and expose the correlation between the two types of results.