Meet the Judges: Dr Direk Limmathurotsakul, Head of Microbiology at Mahidol-Oxford Tropical Medicine Research Unit

AMR is being increasingly prioritised by stakeholders in Thailand, and steps are being taken in line with the country’s national action plan on AMR. 

To my knowledge, Thailand is the first LMIC that officially reports the annual volume of antibiotic consumption in both humans and animals. Nonetheless, monitoring and ultimately reducing the total number of deaths attributable to AMR infections, monitoring appropriate and inappropriate use of antibiotics, reducing inappropriate use of antibiotics, and improving public awareness about AMR are all still critically needed at both a local and national scale.

AMR has long been a main theme of my day-to-day research, and I am committed to  fighting AMR infections in any way I can. As well as conventional clinical research and epidemiological studies, I also work with collaborators to conduct behaviour-related and social-science studies.  

During the past five years, the emergence of other issues demanding an urgent response – including Covid-19 and climate change – has meant that the national and international support and momentum around AMR is at risk of declining.

More needs to be done to save lives from AMR infections. The vision and action to fight against AMR is expanding to also cover behaviour studies and social science studies, which is great. Analysing the international response to Covid-19, as well as the vaccination programme, will also be critical in helping to inform the local, national and international AMR action plan.

Covid-19 has changed the landscape of diagnostic tests in many ways. PCR tests, which are notoriously expensive, have become the gold standard test for Covid-19 in both LMIC and high-income countries (HICs). A C-Reactive protein (CRP) test has been used to support the differentiation between virus infections (low CRP) and bacterial infections (high CRP) and guide the use of antibiotics in some settings. However, many patients with Covid-19 have high CRP levels. As Covid-19 is unlikely to completely disappear, the utility of CRP tests for guiding antibiotic use would need re-evaluation. 

The consideration of antigen detection diagnostic tests for Covid-19 has been a good exercise for every nation, and the world now has a far better understanding of the value of highly accurate diagnostic tests for infectious diseases. The public is also more concerned about whether they can trust the accuracy of diagnostic tests presented by early-phase studies (or by the company) compared to real-world data. The issues of selecting test subjects and solving the problem of imperfect gold standards are being raised more than ever. For example, PCR is not a perfect test for Covid-19 – a proportion of pre-symptomatic and mild Covid-19 patients could show a negative PCR result. 

Evaluating a new diagnostic test by comparing severe Covid-19 cases to healthy controls could show that the test has a very high accuracy. However, this accuracy could be overestimated compared to when deploying the test in real-world settings with a wide range of pre-symptomatic, mild to severe patients. Estimating the accuracy of a new diagnostic test in a wide range of patients, when even PCR (which is considered the gold standard for Covid-19) is not perfect, requires complex study designs and analytical methods. 

The pandemic has also highlighted how different diagnostic tests can play a different role for the same disease. For example the antigen-detection test for screening, PCR for a gold standard confirmation test and antibody-detection test for previous exposure. In some ways, the pandemic has paved the way for future technologies and platforms for diagnostic tests to be developed and tested.

Not yet. It’s improving, but more needs to be done to communicate the value to physicians and patients and more research is needed to understand geographical nuances. 

There needs to be a better understanding of the barriers and facilitators for understanding and using diagnostic tests. Armed with that knowledge, researchers could develop and conduct intervention studies. 

I also believe that people in different communities and with different backgrounds will need different interventions and ways of communicating to fight antibiotic resistance. For example, daily-wage workers in many areas in LMICs may struggle to understand the value of diagnostic tests to detect viral infections vs. bacterial infections for the common cold, or to understand the cost-effectiveness (even if the test is free) of testing and avoiding inappropriate use of antibiotics. This is because they may believe that antibiotics are the best remedy to restore their strength so that they can get back to work quickly. Perhaps their friends are telling them to buy the most expensive antibiotic (i.e. they are following norms and influenced by peers), or they might believe that any recommendation or guidelines telling them not to use antibiotics are from conspiracy theories. 

To overcome this, we should learn from the experiences that countries around the world are having when communicating about Covid-19 vaccines, and adapt our communication plans accordingly – locally, nationally and globally. We must also be conscious of adapting communications for specific audiences, for example policymakers and stakeholders are different from communicating with physicians and patients.   

To my mind, market-size and sales opportunities are the biggest challenges. For diagnostic tests to be developed, reviewed and sold, people must be clear of the need. The issues about technologies, costing and sizes of the machine, in my opinion, are less crucial compared to market-size and sales opportunity. This is because if there is a strong need, such as Covid-19 PCR and antigen-detection diagnostic test, many things can be developed. For example, during the last two years, PCR tests have been relatively easier to perform, and are becoming cheaper and more available in many locations in LMICs.

The other crucial issue is that a test that performs very well in HIC may not perform well in LMICs due to differences in disease epidemiology. For example, a diagnostic test for  bacterial bloodstream infections developed for HICs, may not be able to detect bacteria that are common in LMICs (but rare in HICs). For example, Burkholderia pseudomallei, the cause of an often-fatal disease in the tropics.

This has already occurred with new bacterial identification machines in which the database does not include B. pseudomallei and the machines misidentify B. pseudomallei as non-pathogenic soil bacteria. Therefore, the test developers will need to consider the epidemiology of diseases and people in both HICs and LMICs carefully while developing a new diagnostic test so that it can benefit humanity equally. 

I think the world may need a different financial incentive, purchase or development model for diagnostic tests to be developed, similar to programmes that have been started for vaccines and antibiotics. For example, new funding and market models for developing antibiotics in the US and UK, and GAVI’s partnership model for vaccines could be applied for diagnostic tests. 

There are already some organisations working on diagnostic tests with these concepts, but this needs to increase in both volume and size of organisation. More voices from the public, healthcare providers, researchers, policymakers, stakeholders, and leading international organisations (such as WHO and UN) are also needed.