Risk assessment for antibiotic resistance in South East AsiaBMJ 2017; 358 doi: https://doi.org/10.1136/bmj.j3393 (Published 05 September 2017) Cite this as: BMJ 2017;358:j3393
- Fanny Chereau, technical officer, health emergency information and risk assessment1,
- Lulla Opatowski, consultant1 2,
- Mathieu Tourdjman, consultant3,
- Sirenda Vong, programme area manager, health emergency information and risk assessment 1
- 1World Health Organization Regional Office for South East Asia, New Delhi, India
- 2Université de Versailles Saint Quentin, Paris, France
- 3Saint-Louis University Hospital, Paris, France
- Correspondence to: S Vong
South East Asia is at high risk of the emergence and spread of antibiotic resistance in humans
The risk assessment framework can help countries identify interventions for maximum impact, although isolated interventions will be inadequate
A comprehensive strategy using the One Health approach is needed to contain antibiotic resistance in South East Asia
The indiscriminate use of antibiotics in human and veterinary medicine and food production and the release of antibiotics into the environment have led to the emergence of antibiotic resistant bacteria and genes. The effectiveness of antibiotics is compromised by the growing resistance, and controlling antibiotic resistance is a priority for the World Health Organization.1 Knowledge is lacking about the biological determinants of the emergence and spread of antibiotic resistance. Evaluating these determinants is important to identify effective interventions to contain antibiotic resistance.
Weak surveillance systems and a lack of data make it difficult to estimate the extent of antibiotic resistance in the WHO South East Asia region. We carried out a qualitative risk assessment to evaluate the relative effects of the main determinants of antibiotic resistance and to estimate the risk of the emergence and spread of antibiotic resistance among humans in the WHO South East Asia region.
We used the risk assessment approach outlined in box 1 to characterise the level of risk—that is, the likelihood of the emergence and spread of antibiotic resistance among humans—in the WHO South East Asia region. The region includes 11 countries: Bangladesh, Bhutan, Democratic People’s Republic of Korea, India, Indonesia, Maldives, Myanmar, Nepal, Sri Lanka, Thailand, and Timor-Leste.
Box 1: Risk assessment approach and definitions
Risk: Likelihood of the occurrence of a health event
Risk assessment: A systematic process to gather, assess, and document information to assign a level of risk.2 It provides the basis for taking action to manage the negative consequences of public health risks. The level of risk assigned to an event depends on a specific hazard, exposure to this hazard, and the context in which the event is occurring. Risk assessment includes three components:
Hazard assessment: identification of the hazard(s) causing the event of interest
Exposure assessment: evaluation of how, and how much, a person, or a population is exposed to the hazard(s)
Context assessment: evaluation of the environment in which the event is taking place. Context factors include socioeconomic, ecological, or programmatic factors
Risk characterisation: It combines the findings from the hazard, exposure, and context assessments to assign a level of risk of the occurrence of the event of interest
We first assessed the specific hazards by identifying the important antibiotic resistant bacteria and associated antibiotic resistant genes, such as those with high attributable mortality and high potential for transmission.
WHO has identified seven bacteria with high levels of antibiotic resistance (box 2).3 Of these, extended spectrum β-lactamase and carbapenemase producing Enterobacteriaceae and meticillin resistant Staphylococcus aureus (MRSA) are of greatest concern, as infections have high mortality.4 These pathogens are also commensal bacteria in humans and animals (healthy carriage) and in the environment (water and soil). In the South East Asia region healthy carriage of and infections by extended spectrum β-lactamase producers and MRSA are common.3567
Box 2: Bacteria with high levels of antibiotic resistance3
Gram negative bacteria—Neisseria gonorrhoeae with decreased susceptibility to third generation cephalosporins, and Escherichia coli, Klebsiella pneumoniae, Shigella spp, and non-typhoidal salmonella with resistance to third generation cephalosporins (including resistance conferred by extended spectrum β-lactamases) and fluoroquinolones or carbapenems
Gram positive bacteria—meticillin resistant Staphylococcus aureus and penicillin resistant Streptococcus pneumoniae
We then described the processes that lead to the acquisition, selection, and spread of the resistant bacteria and genes in humans. The exposure assessment identified the reservoirs (human, animal, and environment) from which antibiotic resistance can emerge, the transmission routes both within and between reservoirs, and the biological determinants of the emergence and transmission of antibiotic resistance. The transmission routes for human acquisition of the antibiotic resistant pathogens are shown in fig 1⇓.
The widespread use of antibiotics in human and veterinary medicine for therapeutic or prophylactic purposes is the main driver of the acquisition and selection of antibiotic resistant bacteria.891011 The resulting release of active or unmetabolised antibiotics from human and animal waste, from aquaculture, and from pharmaceutical companies into the environment further increases the risk of selection of antibiotic resistance.1213
Transmission of resistant bacteria and genes to humans is caused by ingestion of contaminated food or water and by direct contact with contaminated animals, soil, or water.813141516 Spread within the human population depends on the capacity for sustained human-to-human transmission. This capacity is still poorly understood.17 In humans, spread of resistance can occur both in the community and in healthcare settings.1819 High levels of antibiotic use, high density of at-risk individuals, and poor hand hygiene in healthcare settings lead to increased transmission of resistant strains from patients or healthcare workers to other patients by direct contact or through contaminated surfaces.112021 The use of antibiotics in the community is thought to be even greater than in healthcare settings and further increases human-to-human transmission of antibiotic resistance.11
MRSA and extended spectrum β-lactamase strains from healthcare settings or community outbreaks often differ from livestock associated strains.22 This could suggest a limited transmission capacity of strains from animal or environmental origin in humans.1723 These species barriers are not well understood, and this lack of knowledge impedes the risk assessment process, particularly about the acquisition of resistant bacterial strains from another reservoir.
We assessed the factors specific to the WHO South East Asia region that could affect the causes of the emergence and transmission of resistant strains. The factors were examined at the policy level (scope of policies and guidelines to contain antibiotic resistance in the region), system level (implementation of policies, healthcare system management, wastewater management, and agriculture and livestock management), and individual level (human behaviour and beliefs).
We gathered information from peer reviewed articles, press articles, published documents from WHO, the Food and Agriculture Organization, and Unicef and from unpublished WHO reports. Because of the lack of data available at the country level, the context assessment considered the region as a whole.
Several context factors in the region increase the risk of acquisition, selection, and spread of antibiotic resistance (table 1⇓). Awareness of the drivers of antibiotic resistance, particularly inappropriate antibiotic use, is low among the public and health professionals in most countries of the region.42 Unregulated antibiotic distribution and poor implementation of antibiotic stewardship programmes in humans 27 lead to overuse and misuse of antibiotics. Wastewater management is inadequate. About 75% of wastewater is not treated in the countries of the region and is directly released into the environment, thus contaminating natural drinking water sources where access to piped or improved water supplies is limited.37 Even when water is treated, the effectiveness of removal of contaminants is limited.33
The lack of antibiotic stewardship programmes results in the inappropriate use of antibiotics in livestock and aquaculture. Furthermore, poor infection control practices in farms lead to the selection and spread of resistant strains.31 Antibiotic use in livestock is expected to rise because of increased production in response to a growing demand for food from animal sources.32 International food safety and quality standards have reduced antibiotic use in animal products intended for export; however, the same is not true for livestock products for domestic consumption. Other regional challenges include the limited ability to detect and measure resistance in the human population. Our recent analysis of national programmes to tackle antimicrobial resistance43 showed limited capacity for microbiology and antibiotic susceptibility testing in most public hospitals in the region; only Thailand has data on the national trends for extended spectrum β-lactamase and carbapenemase producing Enterobacteriaceae. Early diagnosis and active surveillance are lacking thus delaying implementation of infection control measures. Also, efforts to identify animal and environmental reservoirs and to detect antibiotic resistance in the food chain and water systems are limited across the region.31
Risk characterisation for the WHO South East Asia region
To characterise the level of risk of the emergence and spread of antibiotic resistance among humans in the WHO South East Asia region, we integrated the risk assessment findings described above in a two layer model (table 1⇑). The first layer is based on the exposure assessment of reservoirs, transmission routes, and determinants. The second layer considers the context factors in the region.
The risk—that is, likelihood of occurrence of each event (acquisition, selection, transmission of antibiotic resistance)—was rated using a qualitative approach as:
Negligible—the event occurs under exceptional circumstances
Low—the event occurs some of the time
Moderate—the event occurs regularly
High—the event occurs in most circumstances.
Based on this approach, we first rated the level of risk for each event arising from the main causes of the emergence and spread of antibiotic resistance (independent of the context). We then estimated the effect of the context factors on the strength of the main causes.
Several studies argue that the emergence of antibiotic resistance in humans is primarily driven by inappropriate and overuse of antibiotics among humans and animals, compared with environmental antibiotic contamination. We therefore considered that antibiotic use leads to a high risk of emergence of antibiotic resistance in both human and animal reservoirs.1444 Human-to-human transmission in the community, which has long been underestimated, could have the same effect on the spread of antibiotic resistance as transmission in healthcare settings,1819 and we estimated these risks as moderate. Transmission of resistant bacteria to human populations through foodborne, or waterborne routes combines a high exposure rate to potentially contaminated food or water with a very low probability of transmission and acquisition of resistance in humans. We therefore considered the risk caused by foodborne and waterborne exposures to be low. Direct contact with contaminated animals or environmental elements is rarely reported, but it is possible that a limited number of people are exposed; we thus considered the risk of transmission to be negligible.
Based on our judgment, and on available documentation, a context factor was classified as decreasing or increasing the strength of a cause of the emergence and spread of antibiotic resistance, and thus decreasing or increasing the likelihood of the occurrence of an event or as having a negligible effect (table 1⇑).
We determined the overall risk of each event occurring in the region by combining the effect of the context factors with the risk from the factors causing the emergence, and spread of antibiotic resistance (fig 1⇑).
For example, we estimated that discharge of antibiotics and antibiotic resistant bacteria and genes into the environment was a moderate risk for selection of resistance and that the low level of wastewater management increased the risk, making it high (fig 1⇑).
We chronologically integrated all the events in the chain leading to the transmission of antibiotic resistance in the human population. To calculate the risk from two consecutive and dependent events in the model, we applied a combination matrix (fig 1⇑).45 When several independent events contributed to the estimation of the risk, the event with the highest risk was used. For example, the transmission of antibiotic resistance from the environment to humans is dependent on the emergence and selection of resistance in the environment. Using the combination matrix,45 we estimated that the emergence and selection of resistance in the environment, which we rated as a high risk in the region, followed by the environment-to-human transmission of resistance, which we rated as a negligible risk in the region, led to a low level risk of acquisition of resistance among humans (fig 1⇑).
Given the size of the population, the high burden of bacterial infections, the high rate of antibiotic resistance, and the contribution of context factors identified in this risk assessment, we believe that the overall risk of the emergence and spread of antibiotic resistance among humans in the WHO South East Asia region is among the highest of the WHO regions. The risk of transmission of antibiotic resistance from animal or environmental reservoirs to humans is high in the context of this region and is mediated by the high risks of foodborne and waterborne transmission. We estimated a high risk of transmission of resistance in healthcare settings and in the community, leading to an overall high risk of transmission of antibiotic resistance among humans in the region.
Although the conclusion could have been reached intuitively, our model considered the main biological causes of antibiotic resistance events and factors specific to the WHO South East Asia region and provides an overall picture of the factors affecting antibiotic resistance. Our assessment highlights the limited benefits of isolated interventions in a specific sector and the need for coordinated interventions to reduce the risk of the emergence and spread of antibiotic resistance. The One Health 46 approach recommends linking human health and animal and environmental health in planning interventions to tackle antibiotic resistance.
This study has some limitations. First, we focused on the pathogens of international concern. The risks of each event cannot be generalised to pathogens not included here. Second, we did not do a systematic review of the literature to quantify risk estimates. The effects of some determinants on the occurrence of specific events are not reported (contribution of antibiotic resistance in the environment to resistance among humans) or are controversial (contribution of antibiotic use in livestock to resistance among humans17). These knowledge gaps hindered our estimates of risk. In view of these gaps, listed in box 3, we considered a higher level of risk when uncertainty was high.
Box 3: Knowledge gaps hindering assessment of the risk of emergence or transmission of antibiotic resistance
Acquisition and selection in humans
Rates of horizontal gene transfer from non-pathogenic commensal bacteria
Effect of dose and duration of antibiotic course on antibiotic resistance
Reversibility of resistance after withdrawal of antimicrobial
Acquisition and selection in livestock
Reversibility of antimicrobial resistance after withdrawal of antimicrobial
Acquisition and selection in the environment
Persistence of antibiotics in soil, sediment, or water
Persistence of antibiotic resistant bacteria/antibiotic resistant genes in soil, sediment, or water
Risks from non-environmental bacteria released in environment
Rates of horizontal gene transfer from non-environmental to environmental bacteria
Minimum environmental concentration necessary to select resistance for different antibiotic classes
Dose-response and risk factors for persistence of antibiotic resistant bacteria from animal origin in human flora following ingestion
Routes, risk factors, and rates of transmission of animal bacteria to human flora other than through ingestion
Risk factors and rates of horizontal gene transfer from animal bacteria to human bacteria in different human flora (intestinal, skin, nasopharyngeal)
Efficiency for sustained human-to-human transmission of bacteria or genes from animal origin
Dose-response, and risk factors for persistence of antibiotic resistant bacteria from aquatic origin in human flora following ingestion
Routes, risk factors, and rates of transmission of environmental bacteria to human flora other than through ingestion
Rates of horizontal gene transfer from environmental to human bacteria in the different human flora (intestinal, skin, nasopharyngeal)
Efficiency of wastewater treatment on elimination of antibiotics, and antibiotic resistant genes
Transmission in healthcare settings
Routes, risk factors, and rates of human-to-human transmission of resistant bacteria in healthcare settings
Efficiency of community acquired strains to spread in healthcare settings
Transmission in the community
Routes, risk factors, and rates of human-to-human transmission of resistant bacteria in the community
Efficiency of hospital acquired strains to spread in the community
This qualitative risk assessment framework is in the early stages of development. Each step depends on more than one main biological determinant, and all steps and associated determinants need to be defined to develop a systematic risk model—semi-qualitative or quantitative—in the future. With such a model, it might be possible to get accurate estimates of risk and to measure the effectiveness of interventions on defined outcomes.
In conclusion, despite uncertainties, risk assessment models are needed to characterise the level of risk and to identify important interventions to reduce this risk and to reduce knowledge and system gaps while analysing processes and outcomes. We believe such a model with defined parameters will be of interest at the country level to identify system weaknesses, and assess the effect of measures to control antibiotic resistance.
Web Extra Extra material supplied by the author
Contributors and sources: FC and SV wrote the article, LO and MT reviewed the article, SV designed the study, LO, FC, and SV designed the framework model, and FC, LO, and MT searched the literature. All authors reviewed, and approved the final manuscript. SV is the guarantor. The opinions expressed in this paper are those of the authors, and do not necessarily represent those of the institutions with which the authors are affiliated.
Competing interests: We have read and understood BMJ policy on declaration of interests, and declare funding from the UK Department of Health Fleming Fund.
Provenance and peer review: Commissioned; externally peer reviewed.
This article is one of a series commissioned by The BMJ based on an idea from WHO SEARO. The BMJ retained full editorial control over external peer review, editing, and publication. Open access fees are funded by the WHO SEARO.
This is an Open Access article distributed under the terms of the Creative Commons Attribution IGO License (https://creativecommons.org/licenses/by-nc/3.0/igo/), which permits use, distribution, and reproduction for non-commercial purposes in any medium, provided the original work is properly cited.