Fig 1.
The PRISMA flow chart provides an overview of the article selection process, detailing the steps of identification, screening, assessment for eligibility, and final inclusion of studies.
It highlights the number of records excluded at each stage and the reasons for exclusion, ensuring transparency and adherence to PRISMA guidelines.
Fig 2.
Forest plot depicting the overall proportion of Hemophilus isolates resistant to antibiotics (erythromycin, clarithromycin, azithromycin), calculated using a random-effects model.
Each antibiotic’s resistance proportion is represented by a box plot, with error bars indicating 95% confidence intervals. The individual study results are shown as red points within each category, while the diamond shape at the bottom of each section represents the overall pooled estimate. The plot also includes heterogeneity statistics (I² and p-values), providing insights into variability across the included studies. This visualization highlights the variability in resistance rates across studies while providing an aggregated estimate for each antibiotic.
Fig 3.
Funnel plots assessing publication bias for resistance proportions of (A) erythromycin, (B) clarithromycin, and (C) azithromycin.
Each plot visualizes the relationship between the standard error (y-axis) and the proportion of resistance (x-axis) for individual studies included in the meta-analysis. The white region in each plot’s center represents the symmetry area expected in the absence of publication bias, while the shaded areas indicate potential asymmetry. Symmetrical distributions suggest a low likelihood of publication bias, whereas asymmetry may indicate potential bias or heterogeneity in the included studies. These plots visually assess the reliability and robustness of the pooled estimates.
Fig 4.
Global prevalence of macrolide-resistant Haemophilus spp., illustrating resistance levels to (A) clarithromycin and (B) azithromycin across different regions.
The maps display the proportional resistance rates reported in each country, with colors ranging from green (low resistance) to orange (high resistance) to indicate the severity of resistance. Geographic disparities in resistance prevalence are evident, with certain regions, such as Asia, exhibiting higher resistance levels compared to North America and Europe. These maps provide a visual summary of the global distribution of macrolide resistance and underscore the need for region-specific antimicrobial stewardship and resistance surveillance efforts. Global map visualization was created using OpenStreetMap data, available under the Open Database License (ODbL).
Fig 5.
Subgroup analysis results of macrolide-resistant Haemophilus spp., broken down by various factors.
(A) Comparison of prevalence by continent, illustrating regional differences in resistance levels. (B) Comparison of prevalence by AST methods, highlighting how different testing methodologies may influence resistance rates. (C) Comparison of prevalence by AST guidelines, showing how adherence to specific guidelines can affect reported resistance levels. (D) Comparison of prevalence among Haemophilus species, emphasizing potential species-specific differences in resistance. (E) Comparison of prevalence before and after 2020, illustrating trends over time and the possible impact of recent public health interventions or shifts in antibiotic use. Each panel provides a detailed breakdown of the prevalence of resistance across these subgroups, offering insights into the factors driving the macrolide resistance in Haemophilus spp.
Fig 6.
Meta-regression analysis of macrolide resistance in Haemophilus spp. from 2015 to 2023.
(A) A scatter plot shows the trend of erythromycin-resistant isolates, demonstrating a stable resistance rate over time with a correlation coefficient of −0.087 and a non-significant p-value of 0.849. (B) A scatter plot depicts a significant upward trend in clarithromycin resistance, with a positive correlation coefficient of 0.639 and a statistically significant p-value of 0.03, indicating an increase in resistance over the years. (C) A scatter plot shows the trend of azithromycin-resistant isolates, with a stable resistance rate over time reflected by a correlation coefficient of 0.144 and a non-significant p-value of 0.464. Each data point represents a study, with the size of the circle indicating the study’s weight in the analysis. Solid blue lines represent fitted regression lines, and dashed red lines represent the 95% confidence intervals. These plots provide insights into temporal changes in macrolide resistance, highlighting differences in trends across the three antibiotics.