Supporting surveillance capacity for antimicrobial resistance: Laboratory capacity strengthening for drug resistant infections in low and middle income countries

Abstract

Development of antimicrobial resistance (AMR) threatens our ability to treat common and life threatening infections. Identifying the emergence of AMR requires strengthening of surveillance for AMR, particularly in low and middle-income countries (LMICs) where the burden of infection is highest and health systems are least able to respond. This work aimed, through a combination of desk-based investigation, discussion with colleagues worldwide, and visits to three contrasting countries (Ethiopia, Malawi and Vietnam), to map and compare existing models and surveillance systems for AMR, to examine what worked and what did not work. Current capacity for AMR surveillance varies in LMICs, but and systems in development are focussed on laboratory surveillance. This approach limits understanding of AMR and the extent to which laboratory results can inform local, national and international public health policy. An integrated model, combining clinical, laboratory and demographic surveillance in sentinel sites is more informative and costs for clinical and demographic surveillance are proportionally much lower. The speed and extent to which AMR surveillance can be strengthened depends on the functioning of the health system, and the resources available. Where there is existing laboratory capacity, it may be possible to develop 5-20 sentinel sites with a long term view of establishing comprehensive surveillance; but where health systems are weaker and laboratory infrastructure less developed, available expertise and resources may limit this to 1-2 sentinel sites. Prioritising core functions, such as automated blood cultures, reduces investment at each site. Expertise to support AMR surveillance in LMICs may come from a variety of international, or national, institutions. It is important that these organisations collaborate to support the health systems on which AMR surveillance is built, as well as improving technical capacity specifically relating to AMR surveillance. Strong collaborations, and leadership, drive successful AMR surveillance systems across countries and contexts.

Antimicrobial resistance surveillance in Ethiopia: Implementation experiences and lessons learned

“…2 In addition, AMR puts a financial burden on resource-limited countries. 3 Studies from East Africa have reported high levels of AMR to commonly used antibiotics with Gram-negatives showing 50% – 100% resistance to ampicillin and trimethoprim or sulfamethoxazole (for instance in Uganda, 100% resistance in Escherichia coli has been observed). The extent and burden of this resistance, however, is not being monitored through established ongoing surveillance systems.…”

AMR Surveillance in low and middle-income settings - A roadmap for participation in the Global Antimicrobial Surveillance System (GLASS)

“…These were critically assessed according to the context they were designed for, and their relevance to AMR surveillance in low and middle income countries. Concurrently, we reviewed AMR surveillance systems and existing AMR surveillance capacity 7 .…”

Current accounts of antimicrobial resistance: stabilisation, individualisation and antibiotics as infrastructure

“…At the moment, estimates of the AMR burden in LMICs is derived through a series of extrapolations—for example the proportion of neonatal sepsis deaths attributable to drug resistant infection, estimated as 214,500 in 2012, based on an expected proportion of all neonatal deaths attributable to severe infection, and of those, attributable to a drug resistant infection to the first line treatment (Laxminarayan et al, 2016). Even with planned investment in laboratory capacity for microbiology, from the UK’s Fleming Fund, the US Centres for Disease Control and others, surveillance is unlikely to expand beyond a reference site model in most low resource settings (Seale et al, 2017), leaving population levels of AMR, and clinical impact of drug resistance relatively unknown. The scale of the AMR problem is inferred rather than known, even in relatively high resource settings.…”