THE ROLE OF DIAGNOSTICS IN THE

ANTIMICROBIAL RESISTANCE
RESPONSE LONDON SCHOOL OF
HYGIENE & TROPICAL MEDICINE

WEEK 1 INTRODUCTION TO THE ROLE OF DIAGNOSTICS IN THE

RESPONSE TO AMR
STEP 1.11 ROLE OF DIAGNOSTICS IN THE AMR RESPONSE - FEVER AS AN
EXAMPLE

DAVID MABEY: The syndromic approach to patient management, a patient

comes with a syndrome. So it might be a fever or a fever and cough or a
urethral discharge. So a syndrome basically is a collection of symptoms and
signs. And the syndromic approach to managing that patient is you try and
treat them for all the possible causes of that syndrome.

Well, we think it's not malaria, but there are lots of other causes of fever. So
we better treat for all of them. So people are giving out lots of antibiotics. And
obviously, the more antibiotics are given out, the more likely you are to select
for resistant bacteria. And the same is true for the other syndromes. So a man
came complaining of pain passing urine and urethral discharge. Is it
gonorrhoea, is it chlamydia, is it something else? We don't know. So we'll treat
for all of them. So we'll give at least two different antibiotics. So clearly, the
more antibiotics you give, the more you're promoting antimicrobial resistance.
Then that helps you choose your antibiotics.

But what you really need is better diagnostics to say which infection they
have.

What is the role of diagnostic tests? So ideally, you have a diagnostic test that
says this patient has pneumonia due to strep pneumoniae that is sensitive to
penicillin. But we don't often have that luxury. So you see a child with a fever,
he is most likely to have a viral infection, which doesn't need antibiotics. But
you don't know that. So could you have a diagnostic test that will say this child
probably has a viral infection whereas this child probably has a bacterial
infection? So it doesn't give you the exact cause. But it helps you decide
whether to give an antibiotic, which again will limit the prescription of
antibiotics, which is obviously a good thing

FutureLearn 1
MANUEL DEWEZ: Fever is a common problem in paediatrics. On average,
children under five years of age present three episodes of fever annually.

So in both high and low in low income countries, fever is a common symptom
and common reason to seek for health care and also of hospital admission.
Most of the cases of fever are related to infections. And the causes of
infections in paediatrics vary. It depends on the age, on the geographical
location, also on the season, on the presence of other comorbidities, like HIV,
and other factors like the use of vaccines.

Infections are also an important cause of mortality. It's estimated that they
cause 32% of under five mortality globally. However, most of children with
fever present only trivial and incidental infections.

For example, severe bacterial infections, which can be fatal and cause severe
complications, only represent less than 1% of the cases of fever in children
treated in community and 7% to 15% of those presenting to hospitals. So fever
infections are common, but most children with fever have trivial infections. A
typical conundrum in clinical practice is to identify which children can be safely
sent home without the specific treatment and which needs antimicrobials and
hospital admission.

This uncertainty of not knowing which kid can be safely sent home will lead to
the overprescription of antimicrobials. This is because at primary care level,
patients often have little access to diagnostic tests. But they know that there
is a small risk of severe infection, and they fear missing that. And to avoid
missing cases of severe infection, they prescribe antimicrobials. At hospital
level, often, clinicians would admit the child to rule out a severe infection. And
the child will be hospitalised. He will undergo several tests, if they are
available. These tests could be blood tests, urine tests, X-ray, sometimes
invasive tests like lumbar puncture to obtain cerebrospinal fluid. And in parallel
to that, children will receive antimicrobials, including both spectrum
antibiotics.

Most of these children actually have not a bacterial infection. And this
approach represents an important source of anxiety and discomfort for
children and their families. It's also a burden for health services. And the
overuse of antimicrobials can contribute to the spread of antimicrobial
resistance. So in this context, diagnostic tests that are able to help clinicians
identify which children with fever need antimicrobials among the vast majority
of children who don't need them may improve antimicrobial stewardship. And
in fact, the WHO recommends the introduction of rapid point of care tests in
clinical practice to improve the use of antimicrobials.

FutureLearn 2
SHUNMAY YEUNG: Broadly, in terms of diagnostic tests available for managing
children with fever, there are tests that look for specific pathogens, bacteria,
viruses, fungi or the host immune response to infection. So in the first
category, so looking for pathogens, we have to get samples from the site of
infection, for example, the back of the throat, urine, cerebrospinal spinal and
then look for the pathogen in that specimen. Traditionally, that has meant
doing microscopy and then culturing the specimen to see if any bacteria or
fungus grows. That can take about 24 hours. And that's a problem in terms of
altering your acute management of a child with fever. There have been newer
tests, which are based on the antigen, or PCR.

And the advantage is they also are positive even if the bacteria is dead, which
will happen if the patient has taken an antibiotic. The trouble is for PCR,
although it's very sensitive, it requires a large laboratory infrastructure, and it's
not cheap. So it's not widely available in peripheral health settings or low
income settings. Antigen tests are a little more readily available. And in fact,
some of them are available as rapid diagnostic tests-- for example, the rapid
group A strep test and malaria rapid diagnostic tests. For the latter, the
malaria rapid diagnostic test, it's very, very widely used in malaria endemic
countries. One of the problems with these types of tests is, one, in children,
they often don't localised infections.

So you don't know where they've got an infection. So you can't act. And
secondly, one of those common cause infection is low respiratory tract
infections. And you can't get specimens from the lungs. Young children tend
not to cough up sputum. The other problem is even the presence of a pathogen
somewhere-- for example, a virus in the back of the throat-- might not be the
problem might not be the cause of infection. So lots of us are carrying bacteria
or viruses, and it's not causing any problems. So just the detection of the
pathogen itself may not be helpful.

So that's where the second group of tests comes into its own, so tests which
detect a human host response to an infection, because it implies that the
presence of the pathogen is causing the immune system to mount a response.
Traditionally, we've been using white cell count and the differential between
neutrophils, which tend to fight bacteria, and lymphocytes, which tend to fight
viruses, as an indication of whether there's a bacterial infection. But again, it's
not that sensitive or not that specific. More recently, we've been using
something called C reactive protein, or CRP and procalcitonin, which are acute
inflammatory markers. They go up when there's inflammation.

They tend to go up higher in bacterial infections and also when infection is

severe rather than mild. So they can be useful, and they are available as rapid
point of care diagnostics. But we know that they can be useful in terms of
reducing unnecessary antibiotic use in adults, but the evidence is still lacking
in children.

FutureLearn 3
We do need better tests to help us manage children with fever. And ideally,
rather than just differentiating between bacteria and viruses, they should also
indicate the likelihood the child is going to develop a severe infection. Just
identifying whether it's a bacterial or viral infection may not be enough if it's
just a specific organism, because it doesn't tell us what the actual cause is. For
example, when malaria rapid diagnostic tests were introduced, they were great
because they did identify malaria very quickly. But if they were negative, we
didn't know what was the cause of the child's fever. So in many cases, people
continued to use antimalarials.

Or if they didn't use antimalarials, they used antibiotics, because they weren't
sure of the cause of fever was a bacterial infection. So ideally, you'd have a
test which actually tells you what the causes of the infection is. But just
because the bug is there, as I said earlier, it doesn't mean it's going to
necessarily need to be treated with an antibiotic. And so ideally, you would
have an indicator of if it's actually causing a problem, so there's a host
response. For example, lots of bacterial otitis media, ear infections, don't need
to be treated with antibiotics. They would get better on their own.

So really, we need we need a test which shows you whether it's a bacterial or
viral infection, ideally what type and what treatment would be appropriate and
the likelihood of developing severity. Tests that are under development include
ones that are detecting RNA transcriptomes. So this is early messengers in the
bloodstream indicating that the immune system is mounting response to
specific types of infection and tests that are based on proteins, which are also
produced-- like CRP-- in response to an acute infection. Any test like that, for it
to be useful in the places where the vast majority of children are seen-- i.e.

In the GP practices or community paediatricians in high income settings or rural

health centres in low and middle income settings-- means that the test needs
to be available at the bedside. It needs to give a quick result so that the
resources available to affect the management of the child there and then. It
needs to be affordable. And critically, it has to be actually taken up by the
nurse or the doctor to whom it's made available.

FutureLearn 4