Development of an optimal antimicrobial resistance surveillance system in Viet Nam
Antimicrobial resistance (AMR) is a major global public health concern. The Viet Nam National Action Plan on AMR recognised surveillance as one of critical components for control. However, the current AMR surveillance system (AMRSS) in Viet Nam is likely to be over-representing severe and hospital acquired infections (HAI), potentially resulting in an overestimation of resistance among community acquired infection (CAI). This thesis aims to evaluate the AMRSS in Viet Nam and to make suggestions to optimize the AMRSS effectiveness in providing accurate and representative AMR data for CAI patients in this setting.A systematic litterature review was conducted to generate an overview of the AMRSSs that have been implemented globally and any evaluations of such systems. There is no standardized framework or guidelines for conducting evaluation of AMRSS. Less than 10% of the systems reported some system evaluation, focusing on few attributes such as representativeness, timeliness, bias, cost, coverage, and sensitivity. This review highlighted the need for systematic evaluation to assess AMRSS performance and for developing specific methods, building on current evaluation guidelines, with additional attributes specific for AMR surveillance.An evaluation of the hospital-based VINARES (Viet Nam Resistance) AMRSS in Viet Nam in two time periods, 2012-2013 and 2016-2017, was carried out. The sensitivity of the AMRSS was in the 2-5% range and remained similar between the two periods. There was a delay in data submission from the hospitals, which affected surveillance timeliness. No evaluation of the surveillance system was carried out to identify problems and implement prompt resolutions. Data from these two periods showed increasing trends of resistance among key pathogen – antimicrobial combinations, and a lack of discrimination in resistance results between CAI and HAI patients.Optimization through modelling of the hospital based AMRSS was then carried out focusing on carbapenem resistant Klebsiella pneumoniae using baseline data from VINARES and employing model-based methodologies examining key attributes including accuracy, sensitivity, coverage, and representativeness with two assumptions: (1) hospitals of a same type (national, specialized and provincial) were similar; (2) resistant proportions were similar by type of hospitals for CAI while they varied for HAI. Overall, the results showed that the accuracy of AMR data is enhanced when the number of hospitals increases (0.6% decrease in mean squared error for one additional hospital (CI 0.6% - 0.7%)). For a given amount of budget, the optimal numbers of hospitals by type can be determined using this modelling approach to identify a system with the best values for each performance attribute.The results indicate that the current AMRSS can increase the proportions of specialized and provincial hospitals to increase accuracy of data and system representativeness. The models were based on VINARES data, therefore the results are likely to be valid for an AMRSS with similar organizational structures and data collection protocols. The amount of budget that the government and foreign development partners are willing to spend on AMR surveillance is also an important factor in identifying the optimal hospital combination for the AMRSS.
| Main Author: | |
|---|---|
| Format: | thesis biblioteca |
| Language: | eng |
| Published: |
Université de Montpellier
|
| Subjects: | S50 - Santé humaine, population humaine, politique sanitaire, surveillance épidémiologique, système de surveillance, pathologie humaine, http://aims.fao.org/aos/agrovoc/c_3683, http://aims.fao.org/aos/agrovoc/c_32496, http://aims.fao.org/aos/agrovoc/c_16411, http://aims.fao.org/aos/agrovoc/c_31861, http://aims.fao.org/aos/agrovoc/c_11634, http://aims.fao.org/aos/agrovoc/c_8227, |
| Online Access: | http://agritrop.cirad.fr/596699/ http://agritrop.cirad.fr/596699/1/ID596699.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|