The importance of diagnostics in antimicrobial resistance

Antimicrobial resistance (AMR) poses a global health challenge. This article looks at why timely and accurate identification of pathogens through diagnostics is crucial in the successful treatment of infectious diseases.

Infectious diseases, particularly infections carrying some form of drug resistance, are a significant cause of patient morbidity and mortality. Antimicrobial resistance (AMR), which refers to the ability of microorganisms to withstand the effects of antimicrobials, was estimated to cause over 4.95 million deaths globally in 2019 alone.

In 2016, the much cited AMR Review chaired by Lord Jim O’ Neill highlighted the drastic consequences that could arise from AMR if left unaddressed: 10 million global deaths by 2050 and $100 trillon of foregone economic cost. Eight years after that publication, AMR still poses a global health challenge.

• Further reading: Government launches new plan to tackle antimicrobial resistance

In a more recent viewpoint article, leading global experts on AMR highlighted discordant regional priorities on actions to tackle the crisis. They argued that a recalibration of global funding is required to prioritise fundamental AMR interventions, such as provision of clean water, sanitation and infection control, in the global South. Amidst the emphasis on research and development of new antimicrobial technologies in the global North, basic AMR interventions may be overlooked. One key intervention highlighted by the AMR Review is the continued development of rapid diagnostic technologies.

Timely and accurate identification of pathogens, particularly those which are drug resistant, is crucial in the successful treatment of infectious diseases. In the hospital environment, physicians will typically prescribe evidence-based empirical (‘best educated guess’) treatment while awaiting diagnostic information. However, in the case of AMR, this initial therapy may not fully address the patient’s cause of infection and can result in poorer patient outcomes.

At the recent ESCMID Global Congress,  the European Society of Clinical Microbiology and Infectious Diseases welcomed around 16 000 delegates to the five day gathering. Focus was placed on the use of diagnostics to support infection management in healthcare settings and the emergence of rapid diagnostic technologies. Here, we delve into the successes and challenges of diagnostic approaches, shedding light on their potential and future within infection management.

The importance of diagnostics in the AMR era

During an opening address to a think-tank diagnostics symposium, ESCMID president Professor Annelies Zinkernagel emphasised that AMR continues as a “silent pandemic.” She noted that a key ESCMID priority is driving the response to AMR with scientific evidence.

In her symposium presentation entitled “Ineffective Antimicrobial Therapy for Bacteraemia”, Professor Romney Humphries discussed the evolving landscape of empiric antibiotic therapy. She observed that, due to the continuous spread of AMR, the accuracy of empiric therapy has waned over time. In fact, she stated that:

“We’re not just heading into a post-antibiotic world. We’re there today. We see patients all the time with infections for which there is not a single antibiotic left that we can treat with.”

Part of the problem is that antibiotic overuse (a key driver if AMR) is related to the use of empiric therapy where diagnosis of the infecting pathogen is not available. Humphries emphasised that any delay in identifying causative pathogens leads to worse outcomes for patients. She highlighted a study by Rhee and others which described the association between inadequate or overly broad empiric therapy and increased patient mortality.

During a pre-symposium press briefing on tackling AMR through rapid diagnostics, Professor Till Bachmann stated that:

“Diagnostics save lives, diagnostics save money and diagnostics save drugs.”

Bachmann is of course referring to the important place of diagnostic technologies, not only in identifying pathogens, but also in guiding clinicians towards the most effective and safe agents for treatment.

The paradigm shift: rapid diagnostics

Modern rapid diagnostic testing has transformed the way clinicians identify the causes of infections and augment traditional “phenotypic” techniques. Phenotypic microbiological tests (for example Gram staining) are essential tools used to characterise and identify microorganisms based on their observable traits and behaviours. While these tests are well-established and standardised across healthcare settings, they typically require days to yield results. Recent advances in rapid diagnostics such as whole genome sequencing (WGS), matrix-assisted laser desorption/ionisation time-of-flight (MALDI-TOF) and multiplex molecular diagnostic panels have revolutionised the way in which clinical microbiology laboratories operate.

MALDI-TOF technology is seen as a better value alternative to WGS and produces results in a much shorter time. MALDI-TOF has demonstrated benefits in mortality, financial and drug utilisation outcomes.

Multiplex diagnostic panels, in some cases, do not need pre-culturing of samples and have shown benefits in reducing antimicrobial consumption and associated costs as well as shortening time to appropriate therapies

Beyond the clinical microbiology laboratory, rapid diagnostics have made an impact in various settings, including long-term care and primary care environments. In an opening diagnostics session of ESCMID Global, Professor Susan Hopkins, Chief Medical Advisor at the UK Health Security Agency referred to a recent publication demonstrating that laminar flow assays used in community settings were comparable in performance to the PCR reference standard where populations would otherwise not be tested.

Rapid diagnostics: what does the evidence say?

During a special symposium on the rapid detection of AMR, Professor Evelina Tacconelli explored how modern rapid diagnostics support antibiotic stewardship. Her presentation reframed the conversation around rapid diagnostics, calling for increased focus on patient-centric outcomes, such as symptom improvement and quality of life. According to Tacconelli, clinical evidence predominantly supports the use of rapid diagnostics in managing respiratory, bloodstream, and urinary infections, with less evidence for impact on gastrointestinal and central nervous system infections. During her presentation she also reviewed key findings from recent systematic reviews and meta-analyses:

1. Respiratory Viruses

• A 2019 review assessed rapid molecular tests for influenza, respiratory syncytial virus, and other respiratory viruses. These tests demonstrated high accuracy:
  • Sensitivity: 90.9% (95% CI, 88.7%–93.1%)
  • Specificity: 96.1% (95% CI, 94.2%–97.9%)
• Benefits included reduced testing turnaround times, shorter hospital stays, improved oseltamivir use, and potential cost savings. However, no impact on antibiotic prescriptions or hospital admissions was observed.

2. Primary Care Settings

• A 2020 Cochrane review evaluated rapid tests to guide antibiotic prescriptions for sore throat in primary care. The findings indicated:
  • Approximately 25% reduction in antibiotic prescription rates
  • Limited impact on overall antibiotic dispensing

3. Emergency Departments (EDs):

• A further review explored rapid respiratory virus testing in EDs. While it led to increased influenza antiviral use and reduced chest x-rays and blood tests, it did not affect antibiotic use, ED length of stay, or revisit rates.

4. Bloodstream Infections (BSIs):

• A 2021 Cochrane review compared rapid susceptibility testing with standard methods for BSIs. Based on a review of six randomised controlled trials involving 1638 patients, no significant differences were found in patient mortality, time to discharge, or time to appropriate antibiotic therapy.
• The authors attributed these findings to underpowered studies.

Professor Tacconelli consistently emphasised that patient-specific outcomes were often overlooked or studies were inadequately powered to detect changes in them. Consequently, clinical guidelines remain variable. For example, a recent publication from respiratory and infectious disease special interest groups cautiously advised using multiplex PCR testing for lower respiratory tract samples only when non-standard severe community-acquired pneumonia antibiotics are prescribed.

Given the promise and (somewhat limited) evidence to support the use of rapid diagnostics in clinical practice, Professor Humphries noted in her talk that the uptake of some these technologies has been relatively slow.

Rapid diagnostics: challenges to implementation

The adoption of any new technology in healthcare, including rapid diagnostics, presents multifaceted challenges. While rapid diagnostics hold potential, several barriers hinder their seamless integration into clinical practice. Reflecting on the global inequity surrounding AMR action, implementing these technologies in low and middle-income countries (LMICs) faces specific hurdles:

• Need for additional education and training
• May not be suited to existing patient care pathways
• High purchase costs
• Limited access compared to higher income settings

Healthcare institutions and reimbursement agencies also demand robust evidence of value before investing in new diagnostics. Beyond evidence, successful implementation requires addressing contextual factors:

• Integrating new interventions into established workflows and processes will also be difficult. Professor Tacconelli warned that basic antimicrobial stewardship and infection control measures are prerequisites to maximise the cost-effectiveness of rapid diagnostics.
• This point was very much echoed by Professor Céline Pulcini in her talk on clinical decision support within the antimicrobial stewardship paradigm. She noted that rapid diagnostics have the most impact where results are available for clinicians to make prescribing decisions within their existing workflows.
• Adoption of technologies face other challenges which may affect implementation such as: financial constraints, leadership culture, human resources and structural capacities.

Pulcini noted that not including implementation science principles to guide the integration of rapid diagnostics on clinical settings was, in her words “a huge mistake”.

The promise of artificial intelligence

At ESCMID Global 2024, numerous sessions were dedicated to the application of artificial intelligence (AI) in infectious diseases, particularly within the field of diagnostics. Fitzpatrick and others have reviewed the opportunities and challenges associated with AI interfaces with infection treatment prevention and control. In diagnostics, they highlighted the potential of AI:

1. Predicting Infection Risks:

• AI algorithms can analyse vast datasets to predict infection risks associated with specific pathogens. This capability enhances early detection and intervention.

2. Streamlining Rapid Diagnosis:

• AI streamlines diagnostic processes, enabling faster identification of infectious agents. For instance, it can expedite results interpretation, reducing the time from sample collection to treatment decisions.

3. Automating Operator-Dependent Tasks:

• AI can automate tasks such as Gram staining, reducing reliance on manual labour and potentially improving accuracy.

Aligning with similar comments from Evelina Tacconelli, they noted failure within studies to evaluate the impact of AI in real life clinical settings.

In his keynote speech to a packed symposium room, Lord Jim O’ Neill reflected on developments since he chaired the Review on AMD Review on AMR. Acknowledging that he is not scientifically or medically trained, he highlighted that accessibility to rapid, safe and effective diagnostic technology was a key recommendation of the Review.

Waving his mobile phone in front of the audience, he questioned why the vision and success of AI in modern telecommunications cannot be replicated in healthcare settings. He asked:

“Why on Earth, are we not using modern technology to help our medical professionals know…whether an antibiotic is needed or not. And if so, what kind of an antibiotic might be needed!”

He went on to say that the full value and potential of AI in AMR and diagnostics has not yet been fully realised.

The future of rapid diagnostics

Both Susan Hopkins and Romney Humphries touched upon the characteristics that an ideal diagnostic test should embody: rapidity, accuracy, reliability, clear purpose, non-invasiveness, simplicity, and affordability. However, they acknowledged the inherent difficulty in achieving all these aspirations. Hopkins, in particular, urged a shift in our perception of what constitutes a “good enough” diagnostic test—a pragmatic perspective that balances perfection with practicality.

Similar to Evelina Tacconelli’s observations, Hopkins also emphasised that we are not fully harnessing the potential of randomised controlled trials to evaluate diagnostic tests. While evidence has already demonstrated the feasibility of designing effective and safe rapid diagnostic tests, improvement is needed to highlight their impact at the patient level.

Beyond design and efficacy, successful deployment of new diagnostic innovations hinges on integration into the clinical environment. Computerised decision support systems, which seamlessly link electronic patient healthcare records with diagnostic outputs, will likely be pivotal. However, we must not underestimate the importance of implementation science—an approach that proactively addresses barriers to adoption.

Concurrently, continuous education for clinicians is essential. Romney Humphries succinctly captured the challenge: “The reality is that there are so many tests out there; it’s impossible for clinicians to know them all.”

Throughout the ESCMID Global conference, frequent references were made to the upcoming United Nations General Assembly: High-Level Meeting on AMR scheduled for September in New York. This pivotal meeting aims to address global innovation in tackling AMR, among other key objectives. While alignment between the global North and South remains a priority, rapid diagnostics are poised to play a crucial role.

Closing out his keynote speech at ESCMID Global, Lord Jim O’Neill expressed anticipation for returning to ESCMID Global in the future—a time when the full potential of diagnostics within AMR management becomes a reality.