10 Apr 2019
Written by Christine Carson, Research Microbiologist
The antimicrobial resistance (AMR) maelstrom has been gathering momentum since Fleming, Florey, Chain and others first invented antibiotics in the 1940s. From the moment we started extracting, purifying and using antibiotics in industrial quantities, bacterial strategies to survive exposure to their toxic effects have multiplied and spread. This “perfect storm” has been allowed to gather because, for too long, we have underestimated our opponent and not invested enough in collecting accurate, comprehensive and timely information. We have settled for merely adequate testing and levels of surveillance of AMR. This has to change if we are to survive AMR. We must decide that these levels of information are no longer enough.
For patients with serious bacterial infections, a cornerstone of individual patient care is performing an antimicrobial susceptibility test (AST) – collecting the bacteria causing the infection and testing its susceptibility to a range of antibiotics using standardised, reliable laboratory methods. Results from these tests are used to guide antibiotic treatment decisions. It’s personal. And it matters – to patients and their doctors. It means the doctor makes an antibiotic treatment choice based on high quality information and the patient stands the best chance of getting better.
”We have settled for merely adequate testing and levels of surveillance of AMR. This has to change if we are to survive AMR. We must decide that these levels of information are no longer enough.”
The best evidence for guiding antibiotic choices comes from tests called broth microdilution tests. They are fiddly, fussy, time-consuming tests requiring lots of manual-handling and they are expensive and inconvenient for labs to perform – so they don’t. Instead, simplified variations of these tests are performed. The trade-off is some loss of accuracy, but still, these tests are widely used and generally regarded as adequate.
Substituting adequate tests for the best tests leads to adequate answers rather than the best answers. That is ok as long as the adequate test continues to fairly reliably identify the right antibiotic for treatment and everyone gets to go home happy, most importantly the patient.
It helps that the adequate, patient-specific AST is underpinned by a raft of additional data that’s collected and aggregated to give us the big picture about which antibiotic is most likely to work against particular species of bacteria and infection types. For decades the surveillance work done in the background by microbiology laboratories, health departments, and reference centres, has made it possible to have a pretty good feel for which antibiotics will successfully treat which infections. Until recently, levels of resistance to antibiotics were generally low, and in combination with background surveillance data, the adequate ASTs gave outcomes that were acceptable. Overall, any negative consequences of using cheaper, more convenient adequate ASTs rather than the more expensive, inconvenient best test were infrequent or minimal.
Since most bacteria were susceptible to antibiotics and AMR was relatively uncommon, we could get away with using an adequate AST instead of the best AST. The adequate tests might occasionally miss a resistant organism, but since resistant bacteria weren’t that common, it was a calculated risk worth taking. Labs could save time and money and still give the right answer nearly all of the time. But with AMR becoming more common, adequate tests are at increased risk of mistakenly recommending an antibiotic that won’t work.
While the small picture is changing at the patient level, the big picture is changing too. The increased prevalence of resistant bacteria makes the big picture more complex, difficult and time-consuming to discern. Maintaining useful background AMR surveillance data gets harder because bacteria are developing and spreading resistance to antibiotics faster than public health surveillance efforts can gather information on the patterns. For decades, many public health efforts to conduct AMR surveillance were woefully under-funded or under-resourced. As long as most people with bacterial infections were getting better most of the time, it was very hard for health professionals to argue the system wasn’t adequate and to get any traction for improvements. The system continued to mostly work because adequate ASTs for individual patients combined with adequate surveillance gave us acceptable outcomes. But AMR changes the small and big pictures, and not in our favour. AMR is dynamic and agile, evolving or moving from one bacterium to another without pause. Adequate ASTs and surveillance systems are no longer a match for the creativity, versatility and speed of AMR. We are, quite frankly, outgunned and outnumbered.
AMR is a hard problem to address in its own right, let alone in concert with all the other factors that feed into addressing and managing it. Anything less than a fully orchestrated, coordinated approach to managing AMR will fail. There is work to be done in many places, far from the patient’s bedside.
”The point is, with all this stuff going on (or not going on), AMR doesn’t care. AMR can be likened to an automaton engaged in an endless quest for survival – a quest that human interventions occasionally disrupt.”
For example, when AMR was a minor issue, public health budgets and systems only just accommodated the cost and complexity of adequate surveillance systems. In recent years, down-sized workforces and scaled-down surveillance, curtailed the depth and breadth of data gathered in public health sectors. Now that AMR is a big problem, the impacts of antimicrobial resistant infections need to be managed at the same time as the quality and quantity of data available from existing surveillance systems struggle to keep up.
Purveyors of ASTs, gripped by inertia, developed, manufactured and sold endless permutations of existing AST approaches. Truly novel, disruptive approaches and the possibility of real improvements in speed and accuracy didn’t get a look in, either at the research and development end of things, let alone the sales end. The market simply wasn’t demanding better ASTs, so it didn’t get them.
Regulators share responsibility for some of this inertia too. There were incremental changes to the complex regulatory requirements for ASTs, mostly prescriptive. There was endless organizational inertia, an almost complete lack of harmonisation with other regulatory frameworks connected to AST development and plenty of lip service to wanting to “do things better” but little action. Certainly, there were few signs that genuine change was possible.
These are just a few of the high-level factors that contributed to getting us to where we are now – with ASTs that just barely do the job at the patient level and too little background information on the big picture that often comes too late. But the point is, with all this stuff going on (or not going on), AMR doesn’t care. AMR can be likened to an automaton engaged in an endless quest for survival – a quest that human interventions occasionally disrupt.
In grappling with the magnitude of the AMR problem, a light has been shone on the backstage area and it’s not pretty. We have had decades of propagating and maintaining adequate-but-not-the-best ASTs to direct doctors’ choice of antibiotics for patients. Combined with the inability of surveillance to keep up with the magnitude of AMR change, we are left vulnerable to the impacts of AMR, a formidable and wily opponent.
It seems our lacklustre responses to the small and big pictures is finally back to bite us on the proverbial. And it will bite hard unless we do something about the inadequate systems that sit behind how we choose and use antibiotics. It will cost money. No doubt. But so too will doing nothing – that will cost money and lives. We get to choose. Is it really that hard a choice?
Dr Christine Carson is a research microbiologist living and working in Perth, Western Australia. She is interested in all things bacterial, particularly antimicrobial resistance. Christine is part of a research group at The University of Western Australia developing faster, more accurate tests to guide antibiotic therapy choices.
Twitter: @ccarsonuwa
