Dynamics of Azithromycin Resistance in Pediatric Haemophilus Influenzae Isolates in Wuhan (2015-2024)

To investigate the 10-year dynamics of azithromycin resistance in Haemophilus influenzae isolates from pediatric inpatients in Wuhan, and to assess the temporal association with major public health events, including COVID-19 containment and the 2023 Mycoplasma pneumoniae epidemic.

A total of 815 non-duplicate H. influenzae isolates were collected from pediatric inpatients with respiratory tract infections (2015-2024). Azithromycin susceptibility was determined by CLSI-recommended disk diffusion. Annual resistance rates, mean inhibition zone diameters, and isolate volumes were analyzed. Statistical methods included the Jonckheere-Terpstra trend test, χ² tests, linear regression, and segmented regression with pre-specified calendar breakpoints at 2020 and 2022; an exploratory Spearman correlation with annual MRMP positivity (2018-2024) was also performed.

The azithromycin resistance rate increased significantly from 20.0% in 2015 to 61.7% in 2024 (Z = -6.837, P < 0.001). From 2020 to 2024, susceptibility declined sharply from 69.7% to 38.3%, averaging a 7.85% annual decrease. A transient, non-significant improvement in susceptibility was observed in 2020 (P = 0.518), temporally coincident with the initial COVID-19 containment period. In 2023, susceptibility dropped by 20.57% (P = 0.004), the largest single-year decline, with resistance surpassing 40% for the first time. This resistance surge paralleled a 57.7% increase in isolate volume and a city-wide MRMP epidemic with high macrolide resistance reported. Mean inhibition zone diameter strongly correlated with susceptibility (R² = 0.90, P < 0.001), decreasing to 11.91 mm in 2024, near the CLSI resistance breakpoint. In segmented regression with pre-specified breakpoints at 2020 and 2022, we observed a non-significant level drop at 2020 (-0.484 log-odds; p = 0.270) and a non-significant post-2022 slope increase (Δslope +0.777 log-odds/year; p = 0.126). In a complementary three-phase model, the 2023-2024 slope was significant (+1.065 log-odds/year; p < 0.001).

Early-2020 COVID-19-related containment may have coincided with a temporary slowdown in azithromycin resistance, whereas the 2023 MRMP epidemic was associated with a marked surge-consistent with a level drop at 2020 and a non-significant post-2022 slope increase in segmented regression. Inhibition zone diameter appears to provide an early phenotypic warning of resistance shifts. These findings support reassessing empirical treatment choices and strengthening regional resistance surveillance.