Labelling and quality of antimicrobial products used in chicken flocks in the Mekong Delta of Vietnam

Abstract Background The Mekong Delta of Vietnam is a hotspot of antimicrobial use (AMU), but there is no information on the quality of the labelling and strength of antimicrobial products used in poultry production. Methods Based on a large random sample of farms, we identified the 20 most used antimicrobial products in the area, and investigated their antimicrobial active ingredient (AAI) content by UPLC‐MS/MS (91 analytical tests). Results Only 17/59 (28.8%) batches contained all AAIs within 10% of the declared strength. Worryingly, 65.0% products provided in their label preparation guidelines for both therapeutic and prophylactic use. Withdrawal times for both meat and eggs were stated in 8/20 (40%) products. Conclusion Results highlight deficiencies in quality and labelling contents that undermine authorities’ efforts to discourage inappropriate use of antimicrobials.

TA B L E 1 Characteristics of 20 antimicrobial products investigated, sorted by the number of flocks where they were used  was investigated by fitting a null random effects model with product fitted as a random effect and percent content as the outcome using lme4 package and R software.
The 20 products identified were marketed by nine different companies, and all except one (a French company selling product AB008) were Vietnamese (Table 1). All products were formulated for oral administration: Nineteen (95%) were powder-based formulations and one (5%) was a liquid solution. Five (25%) products contained a single antimcrobial and 16 (75%) a mixture of two antimicrobials.
In order to investigate the inter-batch variability, three batches of 19 products and two batches of one product (AB051) were investigated, making a total of 91 analytical tests (Table 1).
In six (30.0%) products the label provided an explicit indication for therapeutic administration only, 13 (65.0%) products provided an indication for both therapeutic and prophylactic use, and one (5.0%) did not include any indication. Withdrawal times for both egg and meat production were provided in the labels of eight (40.0%) products; in 11 (55.0%) products withdrawal times were indicated only for meat (but not for eggs); one product contained no indications with respect to withdrawal time. A total of 11 (55.0%) products contained only one AAI, and the remaining had other substances (including vitamins, mineral supplements and expectorants and analgesic substances). Twenty-eight (30.8%) samples tested were within 10% of the strength declared in the label. Thirty-four (37.4%) contained AAIs above the declared upper limit, and 27 (29.7%) below the declared lower limit. Two extreme values were observed for two AAIs: one (Product AP16) contained oxytetracycline with strength ranging from 10.3% to 11.9% and another (AB09) product had doxycycline strength ranging from 141.5% to 165.0% of the stated value ( Figure 1). products had all batches and all their AAIs within the ±10% range. A total of 24.5% of the variance was attributed to between-batch variation, the remainder being due to between-product variation.
Since our study is based on a random sample of farms, we are confident that these results are representative of antimicrobial products most commonly used by poultry farmers in the Mekong Delta of Vietnam.
Currently there are >10,000 licensed veterinary products in the country, of which about ~50% consist of antibacterial antimicrobial formulations (Anon 2016). This makes quality control monitoring extremely challenging, particularly in a limited-resource setting such as Vietnam.
Quality testing of AAIs is very costly, and there is a lack of unbiased information about this issue in animal production in most countries.
It has been previously estimated that one in 10 medicinal products in low-and middle-income countries is substandard or falsified (Nwokike et al., 2018). Given that the identity of antimicrobials declared in the label was confirmed in all cases, we do not believe that outright falsification is a major issue here. Furthermore, 'legal' antimicrobials are currently very affordable in Vietnam, and two-thirds of the products investigated had an indication for 'prophylactic use' in the label (normally followed by a list of bacterial diseases). This labelling openly conflicts with the animal health authorities' efforts to discourage routine use of antimicrobials for preventing disease (Aidara-Kane et al., 2018;Anon 2013) and sends a 'wrong' message to farmers (the end users), who will not be able unable to discern in the few instances that medication may be required in the absence of disease. This is particularly relevant in the context of small-scale farmers in many low-and middle-income countries. Farmers in these settings often do not have access to veterinary services capable of providing them with unbiased advice on AMU.
Under dosing is expected to result because of either sub-optimal quality of the manufactured product, or inadequate preparation at the point of administration by the farmer. For most products, the guidelines for product preparation (mixing with water) for prophylaxis were about half the strength required for therapeutic purposes.
There is a risk that this may increase the probability of selection of AMR in bacterial populations (Ungemach, Mueller-Bahrdt, & Abraham, 2006). Withdrawal times for egg production were not specified in 60% of the antimicrobial products investigated. This is a concern, since these products are likely to be used both in meat and layer flocks. The observed inter-batch variation in product quality suggests deficiencies in the mixing/packaging process, since in Vietnam most AAIs sold in Vietnam are bulk-imported and then mixed, packaged and distributed within the country.
Based on a representative field survey, we identified the most common antimicrobial products used in poultry farming in the Mekong Delta. Results indicate variable quality results, with only 17 (28.8%) product batches containing AAIs within the acceptable ±10% range. In addition to improving quality control of veterinary medicine products, we strongly advocate for enhancing regulation and inspection of antimicrobial product labelling, crucially removing the indication for prophylactic use. In all cases, products should indicate withdrawal times for meat, eggs and milk (for products aimed at ruminants). It would be desirable to limit the access to antimicrobials of critical importance for human health for veterinary use, and therefore development of policies aiming at this should be a priority.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

E TH I C A L S TATEM ENT
The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to. No ethical approval was required, as this is a retail study, with no direct implications or impact on any particular subject.