Evaluation of the antimicrobial use in pigs in Japan using dosage-based indicators

The use of antimicrobial agents in food-producing animals may lead to the emergence and spread of antimicrobial resistance in bacteria of animal origin. The use of antimicrobial agents in pigs in 2018 in Japan was evaluated in terms of the weight of active ingredient and number of defined daily doses (DDD), using annual sales data of veterinary antimicrobials sold for use in pigs. In addition, the use of antimicrobial agents in the Japanese pig sector in 2008 to 2017 was evaluated to determine whether or not there were any differences in temporal change pattern by use of different metrics. In 2018, 447 metric tons of active ingredient, corresponding to 77,379 × 106 kg-days (Japanese DDD) and 34,903 × 106 kg-days (European DDD) were sold. The proportion of the sales amount of sulfonamides, trimethoprim and lincosamides to the total sales amount was significantly different depending on the metric used. For most antimicrobial classes, the number of Japanese DDDs was greater than the number of European DDDs. These results indicate that the DDD-based metric, which is more reflective of the selective pressure of antimicrobials, is recommended for use in monitoring the antimicrobial use in pigs in Japan. The differences in the number of Japanese DDDs and European DDDs appear to confirm the need for Japanese DDDs.


Introduction
Increased antimicrobial resistance in bacteria that cause infections in humans is a threat to public health. The use of antimicrobials in food-producing animals in the form of veterinary medicine and feed additives might lead to the emergence and spread of antimicrobial resistance in bacteria of animal origin. Currently 700,000 people die of resistant infections every year. If no proactive solutions are taken to reduce the rise of drug resistance, by 2050, some 10 million lives per year could be at risk from drug resistant infections [1]. Bacterial resistance arises through complex mechanisms, normally by mutation and selection, or by the acquisition of genetic information that encodes resistance from other bacteria [2]. Therefore, diminishing the selection pressure by reducing antimicrobial use is considered to be one of the important strategies to prevent and control the emergence and spread of antimicrobial resistance [2].
As in European countries, over half of the veterinary antimicrobials purchased in Japan are used in pigs [3][4][5][6][7][8][9]. Therefore, reducing the use of antimicrobials and the promotion of prudent use in pig production are important strategies to reduce selection pressure and thus to lower resistance rates. There is no global consensus on the collection of antimicrobial use, data and reporting methods but many activities in this field are in progress [10]. Under the European Surveillance for Veterinary Antimicrobial Consumption (ESVAC) project of the European Medicines Agency (EMA), European countries routinely report total quantities of antimicrobials sold for use in food-producing animals as mg of active ingredient, adjusted by animal biomass (population correction unit: PCU) [11]. The authors have previously investigated the use of antimicrobial agents in food-producing animals in Japan in terms of mg of active ingredient sold per kg of biomass [8,9]. This metric is simple to calculate and easy to understand. However, use of this metric might encourage favouring high potency antimicrobials given their lower mg quantity per dose [12].
In Denmark, the Netherlands and some other European countries and Canada, dosagebased indicators are used to monitor antimicrobial usage at the farm level [13][14][15]. Dose-based indicators have the advantage of making it possible to correct dosage differences between active ingredients and formulations and to measure developments over time, despite changes in which active ingredients are used [16]. In 2016, the EMA published the average defined daily dose (DDDvet) values for antimicrobial agents used in food-producing animals as a tool to facilitate the standardised collection and presentation of antimicrobial use among EU member states [17]. These values were defined by calculating the mean dose of antimicrobial products registered in nine EU member states.
To establish a monitoring system using an indicator based on daily dosage, the authors have previously assigned DDD values for 354 veterinary antimicrobial products approved and marketed for use in pigs in Japan [18].
The aim of this study was to assign Japanese DDD (DDDjp) values for each antimicrobial agent (active ingredient) based on the DDD values assigned to the products. Using these DDDjp values and DDDvet values, we evaluated the sales of antimicrobials agents destined for use in pigs in Japan in 2018 in terms of the number of Japanese and European DDDs. The use of antimicrobial agents in pigs in Japan from 2008 to 2017 was also evaluated to determine whether or not there have been differences in temporal change patterns when using these metrics.

Antimicrobial sales data collection and calculation of the weight of active ingredient
Manufacturers and importers of veterinary antimicrobials in Japan are required, under the Regulations for Control of Veterinary Pharmaceutical Products (Ministerial Order No.3, 1961), to submit details of the sales quantity of veterinary antimicrobials to the Minister of Agriculture, Forestry and Fisheries each year. The data submitted must include the names of antimicrobial products, routes of administration, concentrations of the active ingredient in each product and the target animal species for which the products are used [19]. Annual antimicrobial sales data submitted in this way are compiled into a database by the National Veterinary Assay Laboratory of the Ministry of Agriculture, Forestry and Fisheries, which is available from their website [19]. We used the sales data from 2008 to 2018 and calculated the sales quantity of active ingredient sold for use in pigs by antimicrobial class and administration route.

Assignment of Japanese Defined Daily Dose values for antimicrobial agents (DDDjp)
The DDDjp values were calculated using the DDD values that we previously assigned for 354 veterinary antimicrobial products approved and marketed for use in pigs in Japan [18]. The DDDjp values were calculated by averaging the DDD values of products if there were two or more products containing the same antimicrobial agent. For those antimicrobial agents that are used as active ingredient in products both for injection and oral administration, DDD values were assigned separately for each administration route. Likewise, for those that are used both in single substance and combination products, DDD values were assigned by averaging dosages of both the single substance and combination products. In other words, the average (arithmetic mean) of all DDD values of products for each combination of antimicrobial agent and administration route was used to assign DDDjp-e.g. benzylpenicillin/parenteral.

Calculation of the number of defined daily doses
To calculate the number of DDDjps and DDDvets of each antimicrobial active ingredient, the amount of antimicrobial active ingredient sold each year from 2008 to 2018 was divided by the DDDjp and DDDvet of the corresponding antimicrobial active ingredient. The DDDvet values were available from the EMA website [17]. In calculating the number of DDDvets using formula 2, the DDDjp value was used for those antimicrobial ingredients for which DDDvet was not available.

Number of
The weight of active ingredient and the corresponding number of DDDjps and DDDvets were calculated in total, for the different administration routes (parenteral and oral) and for all antimicrobial classes.

Classification of antimicrobial agents
The antimicrobial agents were classified into 13 groups based on the Anatomical Therapeutic Chemical classification system for veterinary medicinal products (ATCvet) proposed by the World Health Organization (WHO) [20]: tetracyclines; amphenicols; penicillins; sulfonamides; macrolides; lincosamides; aminoglycosides; pleuromutilins; cephalosporins; trimethoprim; polymyxins; quinolones; and others. The specific classification of antimicrobial agents and their DDDjp values used are presented in Table 1.

Statistical analysis
The correlation between the number of DDDjps and the number of DDDvets for different antimicrobial classes was investigated using Spearman's Rho test. Statistical analysis was conducted using Excel 2010 (Microsoft Corporation) and BellCurve for Excel ver. 3.00 (Social Survey Research Information Co., Ltd.) added to Excel.

Antimicrobial sales amount for use in pigs in 2018
The antimicrobial agents sold for use in pigs in Japan was calculated to be 447 tons of active ingredients and 77,379 million DDDjps using Japanese DDD values, indicating that theoretically a total of 77,378 million kg-days of biomass were treated with antimicrobials in 2018. The number of DDDvets was 34,903 million, indicating that the number of DDDs was more than twice as large when calculated using DDDjp than when calculated using DDDvet (Table 2). When investigating the different administration routes by the number of DDDs, the number of DDDs using the oral route represented the largest proportion regardless of the metrics used.

Comparison between the number of Japanese and European defined daily doses
The number of DDDjps of antimicrobials sold for parenteral use was 1.34 times greater than that of DDDvets. The number of DDDjps of antimicrobials sold for oral use was 2.25 times greater than that of DDDvets (Table 2 and Fig 2). With regard to the number of DDDs of antimicrobial agents sold for parenteral use, the number of DDDjps was larger than DDDvets for most of the antimicrobial classes except for sulfonamides, trimethoprim and amynoglicodes (Fig 3: top). Spearman's Rho test revealed that these two variables were significantly correlated (r = 0.978, p<0.001). The number of DDDjps sold for oral use was larger than DDDvets for all antimicrobial classes (Fig 3: bottom). Again, Spearman's Rho test showed these two variables to be significantly correlated (r = 0.736, p<0.041).

Temporal change of antimicrobial sales amount using different metrics
The evolution of antimicrobial sales from 2008 to 2018 in terms of the weight of active ingredient, the number of DDDjps and the number of DDDvets are presented in Fig 4. The temporal changes between years saw the same trend regardless of the metrics used, except for between

Discussion
This study is the first attempt to evaluate the national antimicrobial sales amount in Japan using the number of DDDs. Dosage-based indicators have been used mainly to measure the antimicrobial use at farm level [13,14,[21][22][23], except in France where a dosage-based indicator, ALEA (animal level of exposure for antimicrobials) was developed to monitor the antimicrobial use using national sales data [7].

Effect of using a dosage-based indicator
The relative distribution by antimicrobial class differed depending on the metrics used. As a result, the temporal change pattern was reversed for certain years depending on the metrics used (between 2008 and 2009, the weight of active ingredient sold for parenteral use increased while the corresponding number of DDDjps decreased; between 2011 and 2012, the weight of active ingredient sold for oral use decreased while the numbers of DDDjps and DDDvets

PLOS ONE
increased). In particular, most tetracyclines and sulfonamides have a dosage larger than other antimicrobial agents, resulting in that the relative distribution of these classes was large when measured by the weight of active ingredient but is lower when a dosage-based metric was used. On the contrary, high potency antimicrobials such as cephalosporins, macrolides and quinolons (for parenteral use) and amphenicols, macrolides, lincosamides and trimethoprims (for oral use) presented a larger relative distribution of these classes when a dosage-based metric was used. This change caused by the use of dosage-based indicator instead of using an indicator based on the weight of active ingredient has been highlighted in other studies [24][25][26]. This illustrates that the weight of active ingredient does not reflect treatment intensity and risk of development of antimicrobial resistance. Thus, the use of dosage-based metric is recommended for monitoring of antimicrobial use in pigs in Japan.

Comparison between the number of DDDjps and the number DDDvets
This study shows that evaluating antimicrobial use at national level leads to a significant difference in the number of DDDs depending on whether the DDDjp or DDDvet values are used. The number of DDDjps was greater than the number of DDDvets for most antimicrobial classes. This was attributed to the fact that DDDjp values are lower than DDDvet values for most antimicrobial agents. Despite the fact that the number of DDDjps and the number of DDDvets  Canada also found that in developing their country-specific DDD values, the majority of their DDD values were lower than their corresponding DDDvet values [27]. There are many reasons for the difference observed between DDDvet and DDDjp or DDD values in other non-European countries. One reason is that the EMA might have had a wider range of

PLOS ONE
antimicrobial doses to work with due to the collection of antimicrobial agent doses from nine European countries [28,29]. The different labelling regulations, different treatment indications and different husbandry practices might all contribute to the variations in DDDvet and DDDjp values. However, fully elucidating the reasons for these differences is beyond the scope of this study.

Need for the use of Japanese Defined Daily Doses (DDDjp)
This study revealed that despite the large difference in the number of DDDjps and the number of DDDvets, a possible national level antimicrobial use monitoring system will provide similar conclusions regardless of whether the Japanese or European DDD value is used. Furthermore, this study showed that DDDvets did not cover all the antimicrobial agents used in veterinary medicine in Japan. Although drawing conclusions from differences between assigned DDDjp and DDDvet values is difficult as discussed previously, the differences in the number of DDDjps and DDDvets appear to confirm the need for Japanese DDDs, which are based on national approvals in comparison to the average EMA definitions collected from nine EU members and they better reflect antimicrobial selection pressure in the Japanese context.

Limitations
Given the fact that the present study used calculations from the national sales data and DDD values based on national approvals, one should keep in mind that the exact amount of biomass subjected to antimicrobial treatment in terms of kg-days cannot be assessed because both over-dosing and under-dosing could alter the results. The calculation presented in this study only allows a statistical estimation of probable antimicrobial use but provides a consistent and transparent technical method for adjusting weight-based measures by dose. To avoid over-or under-estimation of antimicrobial use, the use of used daily doses (UDDs) might be a solution, but to assign the UDDs, additional information (such as the number of animals treated and the treatment duration) are required [30,31].