Despite advances in technology and medicine, half the world’s population has no access to essential diagnostics (1). For individuals living in lower- and middle-income countries (LMICs), this diagnostic gap can be life threatening.
But not all hope is lost. Researchers are exploring ways to introduce low-cost biomarker testing into low-resource settings. Here, we turn to two experts exploring the possibilities in the respiratory infections landscape.
Meet the panelists...
Augustine Onwunduba is Lecturer in Clinical Pharmacy and Pharmacy Management, Nnamdi Azikiwe University, Awka, Nigeria.
Yoel Lubell is Professor of Global Health at the University of Oxford, UK, and head of the Economics and Implementation Research Group in the Mahidol Oxford Tropical Medicine Research Unit (MORU) in Bangkok, Thailand.
What gaps exist in current diagnostic strategies for respiratory infections in LMICs, and how do these gaps affect patient care?
Augustine Owunduba: Because rapid diagnostic tools are often unavailable in low-resource areas, primary care providers usually treat suspected respiratory infections based on symptoms alone (2). As a result, they often prescribe antibiotics without confirming the cause. However, most respiratory infections don’t need antibiotics (3), so this approach can lead to unnecessary use. This, in turn, contributes to the spread of antimicrobial resistance (4) (AMR), a problem discussed later in this article.
Yoel Lubell: Agreed, and this issue is made worse by limited access to follow-up care or the ability to escalate treatment if a patient gets worse. As a result, antibiotics are overprescribed, driving AMR and possibly leading to patients receiving the wrong kind of care.
A key issue is that diagnostic strategies don’t always match public health goals. Some focus too little on reducing unnecessary antibiotic use, while others may be too strict, risking the denial of needed treatment.
When developing new diagnostic tools – especially those based on biomarkers – it’s often unclear what the main goal is. Are we trying to tell bacterial and viral infections apart to guide antibiotic use? Or are we trying to spot patients who are likely to get worse, regardless of the cause? Without clear goals, even good diagnostics might not improve patient care or support public health efforts.
What are the biggest challenges to implementing biomarker-based diagnostics in low-resource settings?
AO: There are two key challenges. Firstly, most patients pay for healthcare themselves, so they may not want to spend money on a test. And secondly, community pharmacies – where many respiratory infections are treated – are private businesses that make money from selling antibiotics, so they may be reluctant to offer testing. Any plan to introduce testing must take these issues into account.
YL: Other barriers are cost, weak supply chains, lack of training, and limited infrastructure. Even basic tests like lateral flow tests are hard to implement if they seem expensive compared to the treatment they guide, making them appear less useful. Health workers also need training – not just on how to use the tests, but on how to understand the results, adjust treatment decisions, and clearly explain the results to patients. This can be especially difficult in busy, under-resourced clinics.
What are the key criteria for selecting biomarkers for use in diagnostics in environments with limited resources?
YL: In low-resource settings, a good biomarker must be:
Stable and easy to understand without lab equipment
Affordable
Simple to use with little training
Helpful for guiding antibiotic decisions
Usable with a finger-prick blood sample in a rapid test format
There are already many low-cost C-reactive protein (CRP) point-of-care (POC) tests available that meet these needs, especially for assessing respiratory infections. While new tests are being developed, there’s a strong case for using proven biomarkers and technologies that are already well studied and understood.
AO: Affordability is arguably the main factor – biomarker tests must be affordable for patients or whoever is paying for them. That’s why CRP is a good option in these settings: low-cost CRP test kits, like semi-quantitative lateral flow tests, are already available.
How do you ensure the sensitivity and specificity of biomarker-based tests when used outside highly controlled laboratory environments?
YL: You don’t – at least, not at the same levels of accuracy that you’d find in a well-equipped lab in a high-income setting. Instead, you trade a bit of accuracy for practicality. For example, lateral flow CRP tests aren’t as precise as lab tests, but they’re still accurate enough to guide treatment – especially with cut-offs like 10-40 mg/L, which help distinguish viral from bacterial infections. Our real-world studies in Vietnam, Thailand, and Myanmar show these tests work well, even in tough field conditions.
AO: As Yoel says, CRP tests aren’t perfectly accurate – they don’t always detect every case or rule out others. That’s why clinical judgment and other test results should be used alongside CRP results when deciding whether to give antibiotics. For example, if a patient has a high CRP level, a malaria test might help confirm if malaria is the cause, which could reduce the need for antibiotics. On the other hand, clinical judgement alone could inform antibiotic use, despite normal or low CRP.
Why is there a low CRP uptake in LMICs – what are the benefits and limitations of these tests?
AO: CRP testing helps improve responsible antibiotic use for suspected respiratory infections in both public and private primary care in low-resource settings – and it’s cost-effective. However, its accuracy isn’t perfect.
In our trial in Nigeria (5), CRP tests were used in only 21.4 percent of the visits where we expected them. There were a few reasons for this low uptake. Some patients refused the test because they worried they couldn’t afford both the test and antibiotics if needed. Also, some pharmacy staff, lacking a clear understanding that most respiratory infections don’t need antibiotics, began to distrust the test kits when results often pointed away from antibiotic use. To improve uptake, we recommend making the test more affordable and training pharmacy staff more thoroughly.
YL: CRP testing is still used infrequently, and the reasons vary by setting. Some common challenges include the fact that it disrupts long-standing habits and adds complexity to already busy clinical routines. Also, avoiding antibiotics – even when appropriate – may not seem important to healthcare workers or patients, especially when there’s pressure to give treatment quickly.
While it's helpful to highlight how overusing antibiotics contributes to drug resistance on a population level, it's just as important to explain the personal risks – like side effects, harm to the gut microbiome, and longer-term health issues from unnecessary antibiotic use.
Local factors also affect uptake. In a recent trial in Vietnam (6), even with strong support for CRP testing, few patients received the test because family members often collected treatment instead. Still, when the test was used, it significantly reduced antibiotic use without harming patients.
CRP tests, like all biomarker tools, have some drawbacks. Inflammation from non-bacterial causes can lead to false positives, and some early or unusual bacterial infections may not be detected. But overall, CRP testing improves how antibiotics are used – especially in places where overprescribing is common. Making CRP testing routine with today’s tools can also help us learn more and prepare for the rollout of new diagnostic tests in the future.
Can you briefly discuss your study and how CRP testing affected results compared to standard processes?
AO: In Nigeria, community pharmacies give antibiotics to 81 percent of patients with suspected respiratory infections – often without a prescription or diagnosis. In our trial, when we gave these pharmacies CRP test kits and trained them to use the tests to guide antibiotic use, the chances of them giving antibiotics without a prescription dropped by 72 percent.
YL: In our large trial in Vietnam, we aimed to test CRP use in real-world conditions. To keep things realistic, we didn’t place research staff at the clinics, and ethics committees allowed us to skip individual patient consent because the test was already well studied. This helped avoid disrupting routine care and showed how the intervention might work in everyday practice.
However, the study took place during the COVID-19 pandemic, when many people avoided visiting clinics. This likely limited how many patients were reached and reduced the full impact of the intervention.
Still, the results were encouraging. Antibiotic use dropped from 98 percent to 93 percent overall, and to 71 percent among patients who actually received a CRP test. There were no negative effects on patient health or recovery. But the low use of testing shows that simply offering a new tool isn’t enough. To create lasting change, broader support is needed – such as staff training, better integration into clinic routines, and incentives to encourage appropriate antibiotic use.
How does CRP fit into symptom agnostic screening?
YL: CRP isn’t truly symptom-agnostic – it performs best when a bacterial infection is suspected, helping to rule out the need for antibiotics. Still, in cases where the cause of illness isn’t clear, CRP provides a practical way to assess risk and guide more targeted treatment. It’s especially helpful when symptoms alone don’t give enough information to make a clear decision.
If not CRP, are there any other POC tests that can support infectious disease diagnostics in LMICs?
YL: New multiplex biomarker tests are being developed to improve how respiratory and febrile illnesses are managed in LMICs. Many of these tests still include CRP, combined with one or two other markers to improve accuracy – since CRP alone has limitations.
However, this improved accuracy comes at a higher cost. These newer tests are more expensive, and it’s hard to judge whether the extra precision is worth the price, especially since current economic models don’t always factor in the long-term savings from avoiding AMR.
Some of these tests may also help identify patients at risk of becoming seriously ill, regardless of whether the illness is viral or bacterial. This could be especially useful in remote areas by helping decide when a patient needs to be referred to a higher-level facility. CRP alone isn’t ideal for this purpose. Other markers, like sTREM-1 and ANG-2, may be better at predicting severe illness. We’re currently working with industry partners to develop POC tests for these markers, to give frontline healthcare workers better tools in low-resource settings.
Do you believe POC testing will play a big part in diagnostics in LMICs moving forward?
AO: Absolutely – relevant stakeholders in low-resource settings are encouraged to consider implementing CRP testing intervention for suspected respiratory infection in primary care.
YL: The pandemic showed that large-scale testing for respiratory illnesses is possible – not just in hospitals, but even in communities. Along with the long-standing success of malaria rapid tests in remote areas, this proves that POC testing can work well in LMICs.
Looking ahead, POC testing will likely become even more important. New technologies – like advanced tests that measure both host biomarkers and specific pathogens – are being developed. When paired with mobile apps, electronic decision tools, and AI, these innovations could help fill gaps in lab services and make up for the shortage of highly trained health workers in low-resource settings.
What are the challenges that AMR presents in LMICs?
AO: AMR occurs when bacteria stop responding to antibiotics that once worked. It’s a serious health problem, especially in low-resource settings. In 2019, the death rate from bacterial AMR (7) was higher in Sub-Saharan Africa (23.7 deaths per 100,000 people) than in high-income countries (13.0 per 100,000).
AMR also makes infections more expensive to treat. In low-resource settings, treating someone with a resistant infection can cost about $12,442 (8) more than treating someone with a non-resistant infection.
YL: We’re just beginning to understand all the complex factors that drive AMR – and many go beyond antibiotic use in healthcare. Things like farming practices, climate change, conflict, poverty, and inequality all play a part, especially in LMICs. At the same time, many people in these settings still don’t have enough access to the antibiotics they truly need. So, any plan to fight AMR must carefully balance reducing misuse without limiting access to life-saving treatment.
That said, overuse of antibiotics in human health – especially in clinics and communities without proper diagnostic tools – is a major and avoidable cause of AMR. In LMICs, this is clear when it comes to treating respiratory illnesses, where antibiotics are often given without a confirmed need or bought over the counter without regulation. While it’s important to look at the bigger picture of AMR, cutting unnecessary antibiotic use in primary care – especially for respiratory infections – is one of the quickest and most achievable steps we can take. Widening access to affordable, effective diagnostic tests is key to making that happen.
Are there ongoing trials or pilot programs that are testing new biomarker-based diagnostic tools in the field?
YL: We recently ran a trial in Cambodia (9) that combined CRP testing and pulse oximetry with an electronic decision support system to help manage acute fever cases. While final data is still being analyzed, we’ve already gathered feedback through focus groups and interviews with healthcare workers and policymakers. Overall, the response has been positive. People saw the value of these tools in helping make better clinical decisions, especially where lab resources are limited.
However, one key takeaway was clear: for these tools to work well, they must be co-developed with the people who will use them. Involving frontline workers in the design process is essential to make sure the tools are practical, user-friendly, and likely to be adopted long-term.
Newsletters
Receive the latest analytical scientist news, personalities, education, and career development – weekly to your inbox.
References
- The Global Alliance for Diagnostics (2022). Available at: https://cdn.who.int/media/docs/default-source/blue-print/session-3_dan-bausch_diagnostics_pathogenx-meeting_aug2022.pdf?sfvrsn=c6e58800_3.
- NV Nguyen et al., Lancet Reg Health West Pac, 30 (2023). PMID: 36419738.
- M Leven et al., Clin Microbiol Infect, 24, 11 (2018). PMID: 29447989.
- S Riedel et al., Eur J Clin Microbiol Infect Dis, 26 (2007). PMID: 17551759.
- A Onwunduba et al., Int J Infect Dis, 127 (2023). PMID: 36509332.
- Y Lubell et al., Antimicrob Resist Infect Control, 7, 119 (2018). PMID: 30323922.
- Antimicrobial Resistance Collaborators, Lancet, 399, 10325 (2022). PMID: 35065702.
- K Allel et al., PLoS Med (2023). PMID: 37347726.
- R Chew et al., BMJ Open, 14, 10 (2024). PMID: 39424394.
