Longitudinal study assessing antimicrobial usage in broiler production cycles in northern Vietnam in 2022 to 2023

Antimicrobial drugs (AMDs) are commonly used in chicken production in Vietnam for both preventive (prophylactic) and therapeutic purposes. Characterizing how the frequency of antimicrobial usage (AMU) changes from flock placement to harvest, varies by purpose, and may be influenced by production stage and mortality levels, can guide antimicrobial stewardship interventions to mitigate the risk of antimicrobial resistance (AMR) in broiler production. We investigated daily AMU and mortality patterns in production cycles of slow-growing broiler chicken flocks raised on 60 farms in two provinces (Ha Noi and Bac Ninh) in northern Vietnam at the end of 2022 and the beginning of 2023. The association between the daily probability of AMU and mortality risk was evaluated using generalized additive mixed models (GAMMs). The dynamic process of initiating a new treatment, and the expected duration of that episode, were further assessed using a two-state semi-Markov process (SMP) within a Bayesian framework. A total of 569 treatment events (70.5% preventive vs. 29.5% therapeutic) were recorded across all 60 monitored cycles. A median value of 186.1 mg/Population Correction Unit (PCU) AMDs was used for prophylaxis, while 116.7 mg/PCU was used for therapeutic treatment per production cycle. The most frequently administered AMDs for preventive treatments were colistin (23.2%), amoxicillin (19.2%), and doxycycline (11.7%). Therapeutic treatments mostly involved doxycycline (26.8%), florfenicol (17.9%), and amoxicillin (13.7%). The main effect of within-flock mortality was not significant, but the interaction between mortality and broiler age was associated with AMU both in the case of preventive and therapeutic AMU (p-value <0.05), suggesting that the effect of mortality on AMU may have varied depending on the broiler age. Broiler age was strongly and non-linearly associated with AMU (p<0.05), with distinct temporal patterns for preventive and therapeutic use. At the start of a production cycle, initiation of preventive treatment was substantially more likely than initiation of therapeutic treatment (day 0: risk ratio [RR] = 8.6), and prophylactic treatment remained dominant during the early cycle (RR > 2 before day 22). These findings indicate that preventive treatment is the primary driver of AMU during the early cycle, highlighting the need for stewardship interventions at this stage, focused on reducing non-essential prophylactic use and promoting lower-risk alternatives.