2018-11-15
On the first day of the World Antibiotic Awareness Week, WHO released the first global surveillance report on antimicrobial consumption. The surveillance system has been developed in response to the identified need to track antimicrobial consumption as stated in the Global Action Plan on antimicrobial resistance. It is by and large inspired by the European surveillance system on antimicrobial consumption, ESAC-Net and the AMC network. The surveillance system is being rolled out in an early implementation phase, where WHO has informed about the system and also given training to many countries.
The 113 page report provides information from 65 countries based on available records on an aggregate level from 2015 or 2016. Core ATC groups of antimicrobials to be included in data collection are Antibacterials for systemic use (J01), Antibiotics for intestinal tract (A07AA) and Nitroimidazole derivatives (P01AB). Other systemic antimicrobials are optional for data collection. As such, the data has no information about indications, dosages or appropriateness of use. In addition to the consumption data, the report also contains information on regional health status and spending.
In this article, we present four take-aways from the report to provide some overview and help with reading and interpreting the report.
Take-away 1: The surveillance system is making progress
As the surveillance system is modeled upon the European systems ESAC-Net and AMC-network, much of the data in this first report is from Europe and the WHO Euro-region. Thus 46 of the countries included in the report are European, six are from the Americas, another six from the Western pacific region, four are from the Eastern Mediterranean region and another four are from the African region. Three additional countries had reported data but were not included due to issues with data format, coverage or quality. No data was available from the South-East Asian region.
Considering that this is an early implementation report of a surveillance system launched in 2016, the results are encouraging. In addition to the reporting countries, many have received information and training on how to collect and report data. Participation in the system should be encouraged by all stakeholders.
Take-away 2: Differences in data sources and coverage
Many different data sources were used for collection of the national data, depending on availability in the countries. The two major sources of data were sales data and import records, followed by reimbursement data. In countries with local manufacturing, these data play an important role.
The data generally cover a whole sector, such as public healthcare or hospital care, and explicitly only the official markets. No efforts are made to quantify the informal sector, which can be very large in some countries. In some cases the reported data may not cover the whole sector or population, for example in Peru and Jordan. Thus, care should be taken when interpreting data and comparing countries.
Take-away 3: Large differences between countries and regions.
The consumption data reported varies between countries, from 4.4 DDD/1000 inhabitants and day in Burundi to 64.4 DD/1000 inhabitants and day in Mongolia. In Europe, the median consumption was 17.9 DD/1000 inhabitants and day, with consumption ranging from 7.7 DDD/1000 inhabitants and day in Azerbaijan to 38.2 DDD/1000 inhabitants and day in Turkey.
While the report does not comment on the high consumption in many countries, it does comment on the very low consumption in Burundi: The data is from the central medicine store, which mainly serves public hospitals. While Burundi stands out as the country with the seemingly lowest consumption, the lack of comprehensive data makes interpretations between countries difficult. This is not unique for Burundi – in countries where large portions of antibiotic sales (and thus consumption) occurs within informal sectors or through other channels that are not included in the official statistics, the estimates are probably underestimates.
Another potential problem lies within the nature of aggregate, country level data: if there are large differences in consumption between regions within a country, e.g. between rural and urban areas, these differences are not showing in the data. This would efficiently mask local trends of overuse in some places and lack of access to antibiotics in other.
Take-away 4: Too much antibiotics are used globally
Since the data in the report is on consumption rather than use, there is no information about appropriateness of use. It is possible that in many regions, the antibiotic consumption is too low due to lack of access to antibiotics or due to high disease burdens. On the other hand, the large informal sector in many countries will lead to underestimations of use in many countries and regions.
Looking at the range of consumption in high income countries, it appears that the differences in consumption are likely not due to large variations in the prevalence of bacterial infections. In lower income countries however, the prevalence of infections may well be higher due to a multitude of reasons, warranting higher consumption of antibiotics at least in the short term or in specific regions. It is also clear that any higher prevalence of infections in these countries should ideally be addressed on systems levels such as reducing the spread of diseases rather than on an individual level with antibiotics. On the whole, most countries can do much more to improve the use of antibiotics
Further reading:
WHO Report on Surveillance of Antibiotic Consumption, 2018
ReAct Toolbox