Antimicrobial susceptibility testing results of VINARES project in two periods: 2012- 2013 and 2016-2017

Get Complete Project Material File(s) Now! »

AMR surveillance systems in humans

Among the 75 AMRSS in humans or both humans and animals, 40 and 35 were systems on a national and international scale, respectively. AMC data was collected in 16 systems.
Thirteen international AMRSS collected existing data; 21 acquired new data and one used data from a literature review. The systems that acquired new data used reference laboratories to analyse isolates sent from the participating medical centres or hospitals. Western Europe and Northern America were the two most often represented regions in the surveillance programs. Out of the 40 national systems focusing on humans, 29 systems were based in high-income countries and 11 systems in LMICs (classifications based on world-bank stratifications) [75].

Hospital-based AMRSS

As the overall objective of this thesis focuses on the hospital-based AMRSS, the main analysis focused on 71 such AMRSS (including laboratory and medical centres) (table 1.1; from S01 to S71). These included 34 active, 33 passive and 4 sentinel surveillance systems (table 1.2). The United States were the leading country with 7 national-scaled AMRSS, followed by France (4 AMRSS) (table 1.1). Twenty AMRSS were deployed in more than one continent (including 6 active and 14 passive surveillance systems). Supplementary figures 1a to 1i represent location of countries participating in the largest international systems (S02, S12, S21, S05, S10, S09 and S14).
The most frequently described objectives were monitoring trends in infection and resistance (43 systems), followed by studying the susceptibility pattern of targeted antimicrobials (20 systems) and early alert for novel resistant strains (6 systems) (figure 1.3).

Evaluation approach and framework

SurvTool [152] was used to establish the framework to evaluate AMR surveillance system in Viet Nam. A number of approaches are available for evaluating surveillance systems [60,158,168,169], however, most of these had the limitation that they lacked the level of detail and flexibility needed for practical implementation [62]. SurvTool was developed to provide a practical approach for evaluation of surveillance systems. It was built within the RISKSUR project, an EU-funded project from 2012 to 2015 delivered by the RISKSUR consortium ( Within this project, decision support tools (software which supports analysts and decision makers in making better and faster decisions) were developed that can be used to evaluate surveillance systems, based on an interdisciplinary approach. The development of SurvTool was built on existing evaluation frameworks, methods and tools available in the literature [62] and incorporating inputs from experts. The main objective of SurvTool was to guide users in practically planning and implementing an integrated evaluation of surveillance systems [152].
Specifically, SurvTool helps the user to establish an evaluation context, including surveillance description, evaluation questions and suggestion of assessment methods [170]. This tool was used to: 1) describe the surveillance system and its components under evaluation; 2) identify the most relevant evaluation questions based on the specific context and decision-makers requirement and 3) identify the most relevant evaluation attributes to assess and the method to do so. An evaluation attribute is a criterion that measures a specific functionality or effectiveness of the surveillance system.
To date, SurvTool has been applied mostly in animal health surveillance [171,172]. However, SurvTool can support evaluation of surveillance systems with multiple objectives such as prevalence estimation and case finding, or demonstration of freedom from disease or early case detection. It could be applied on endemic, sporadic or emerging diseases and also to AMR surveillance (
We evaluated the antimicrobial susceptibility testing (AST) data submitted by participating hospitals in VINARES 1 and VINARES 2. For denominator data, we used data from the annual health statistics of Viet Nam to calculate the effectiveness of AMR surveillance system in Viet Nam in the timeframe between 2012-2013 and 2016-2017. The proportions of resistant priority pathogens against relevant antibiotics were summarized by region and hospital type, separately. The antimicrobial susceptibility data come from microbiological laboratories of VINARES 1 and 2. We received the data in WHONET format [173] and imported into the R program for the descriptive analysis. The cost data was extracted from official budget documents of VINARES project stored on the OUCRU server. These data were available for OUCRU staff only. Cost analysis was done in Microsoft Excel.

Qualitative assessment of the surveillance system using OASIS tool

To assess the evaluation attributes, we used OASIS (Outil d’Analyse de Systèmes d’Information en Santé – a French acronym that translates as ‘analysis tool for surveillance systems in human and animal health’); a qualitative assessment tool for assessment of strengths and weaknesses of surveillance systems based on 78 criteria describing the situation and operation of a surveillance system [174].
OASIS was developed by ten epidemiologists, developers and users of assessment methods and surveillance system managers from the French Agency for Food, Environmental and Occupational Health Safety (ANSES) [175]. OASIS combines three methods, including surveillance network assessment tools, critical control points assessment method and the US CDC Guidelines for Evaluating Surveillance Systems to develop a complete and standardized surveillance system assessment tool [175]. It standardizes the results and gives a better comparison of surveillance systems. OASIS was initially developed as an assessment tool of surveillance systems on zoonoses and animal diseases, but it can also be applied for general health surveillance systems. The OASIS questionnaire includes 78 questions representing the functional parts of a surveillance system, and is used to collect information to support the scoring of the assessment criteria. These assessment criteria are categorized in ten sections according to the structure and activities of a surveillance system. Based on the information collected, each criterion is scored from 0 to 3 according to the level of completion of the system and scores are added up for each section; a higher score reflects better performance. These sections include:
· Objectives and scope of surveillance: assess the relevance of surveillance objectives, level of detail, accuracy.
· Central organization: assess the structure and operation of central unit.
· Hospital organization: assess the structure and operation of participant hospitals.
· Laboratory: assess the integration of laboratory in the surveillance system, the quality assessment of laboratory and accuracy of test.
· Surveillance tools: assess the standardization of surveillance protocols, measurement tools and data collection.
· Surveillance procedure: assess the active/passive component of surveillance system.
· Data management: assess the adequacy of the data management system.
· Training: assess the frequency and quality of training.
· Communication: assess the communication between the central unit and hospitals, technical support of central unit and the publication of results.
· Evaluation: assess the evaluation of performance of system and the implementation of improvement.


Quantitative assessment of the effectiveness attributes

The results of SurvTool indicated the effectiveness attributes that the evaluation should focus on. We selected only five effectiveness attributes that are relevant for the surveillance system under evaluation in answering our research questions and based on expert opinion. These include sensitivity, coverage, representativeness, timeliness and cost. The description and assessment method for each of these five attributes are described in table 2.2.
These assessments will be carried out separately for the following pathogen – antimicrobial combinations: A. baumannii – imipenem; P. aeruginosa – imipenem; E. coli – imipenem; E. coli – ESBL; K. pneumoniae – imipenem; K. pneumoniae – ESBL; S. aureus – MRSA. These combinations are suggested by WHO [177] and were the most common in VINARES.

Descriptive analysis of AMR surveillance system in Viet Nam

VINARES 1 and VINARES 2 were two passive surveillance components based on the same hospital network, which aimed to collect routine AST results of selected hospitals across the entire country. All pathogens in all types of samples were collected and the laboratory results were submitted to the system. Both components had similar objectives: (1) to detect the proportion of resistant isolates in order to give recommendations for empiric therapy and (2) to monitor the prevalence of these resistant isolates to generate knowledge on the AMR situation in hospitals to trigger control actions on treatment used or on antibiotic use such as changes of practices and/or behaviours. These two components also used a similar surveillance protocol as described in table 2.4.
The components varied in the number, type and location of the hospitals included in the surveillance activities: 16 for VINARES 1 and 13 for VINARES 2 (Figure 2.1). Three northern hospitals participated in VINARES 1 but did not participate in VINARES 2. The provincial hospital type was the largest group (7 hospitals); followed by specialized hospital (5 hospitals in 2012-13 and 3 hospitals in 2016-17) (table 2.5). Two specialized hospitals covered the Southern region in two periods; while three were in the Northern region in first period and only one in second period (Supplementary table S2.1).

Influence of the system design on its performance

The second output of OASIS indicates the overall influence that the system design and organization have on the ten performance attributes. The percentage associated with a performance attribute represents its level of functionality corresponding to the quality of the organization and design of the system (figure 2.3). The current system organization was found to impact most its flexibility (35%); representativeness (42%), timeliness (57%) and sensitivity (65%); specificity acceptability and reliability were also impacted (Table 2.7). The system was relatively strong in terms of stability, utility, and simplicity. Table 2.7 describes the problems identified in the system that have affected its performance attributes.

Table of contents :

Chapter 1 A systematic review of antimicrobial resistance surveillance systems
1.1. Introduction
1.2. Methods
1.2.1. Literature sources and search strategy
1.2.2. Article selection
1.2.3. Data selection and analysis
1.3. Results
1.3.1. Article selection
1.3.2. AMR surveillance systems in general
1.3.3. AMR surveillance systems in humans
1.3.4. Hospital-based AMRSS
1.4. Discussion
1.5. Conclusion
Chapter 2 Application of SurvTool and OASIS to evaluate the VINARES network for surveillance of antimicrobial resistance
2.1. Introduction
2.2. Methods
2.2.1. Evaluation approach and framework
2.2.2. Qualitative assessment of the surveillance system using OASIS tool
2.2.3. Quantitative assessment of the effectiveness attributes
2.2.4. Evaluation protocol
2.3. Results
2.3.1. Descriptive analysis of AMR surveillance system in Viet Nam
2.3.2. Coverage and Representativeness
2.3.3. Timeliness
2.3.4. Sensitivity
2.3.5. System costs
2.4. Discussion
2.5. Conclusion
Chapter 3 Antimicrobial susceptibility testing results of VINARES project in two periods: 2012- 2013 and 2016-2017
3.1. Antimicrobial susceptibility testing and antibiotic consumption results from 16 hospitals in Viet Nam- the VINARES project, 2012-2013.
3.1.1. Introduction
3.1.2. Methods
3.1.3. Results
3.1.4. Discussion
3.1.5. Conclusion
3.2. Antimicrobial susceptibility testing results from 13 hospitals in Viet Nam – the VINARES project, 2016-2017
3.2.1. Introduction
3.2.2. Methods
3.2.3. Results
3.2.4. Discussion
3.2.5. Conclusion
Chapter 4 Optimizing cost and effectiveness of AMRSS
4.1. Introduction
4.2. Methods
4.2.1. Classification model for CAI/HAI status
4.2.2. Calculate hospital parameters
4.2.3. Effectiveness attribute formulas
4.2.4. Cost assessment
4.3. Results
4.3.1. Proportions of VINARES patients by origin of infection
4.3.2. Parameters for hospitals participating in surveillance network
4.3.3. Number of hospitals
4.3.5. Cost and effectiveness assessment
4.4. Discussion
General discussion, future directions and conclusion


Related Posts