The subject of this lecture are the investigations
that we use for assessing patients who are
presenting with potential lung disease.
And I’m going to concentrate this lecture
on the interpretation of how we use lung-function
tests to assess patients with respiratory
Why do we use diagnostic tests? Well, in effect
we don’t use diagnostic tests if it’s
only mild disease and the clinical assessment
makes it clear what the problem might be.
But with more complex disease, more severe
disease we will need diagnostic tests. And
we need those for different reasons. One is
to confirm the suspected diagnosis. Two is
to make sure we’re not missing an alternative
diagnosis we hadn’t thought about. The third
is to assess how severe that diagnostic is,
how badly affected the patient might be by
that disease. And fourthly, we use diagnostic
testing to establish the baseline data so
we can monitor the progression of the disease
There are some important things that a clinician
needs to understand about diagnostic testing
before using them. The first is that you need,
in your own mind, to define the question that
you’re asking the test to answer.
So, for example, somebody presenting with
pneumonia – well I think it’s clinically
pneumonia. I do a chest x-ray to see whether
there is consolidation present and therefore
confirm that they have pneumonia.
And the choice of which test you do is clearly
going to be dictated by the clinical scenario.
So patients presenting with suspected airways
disease, it’s the lung-function tests which
are important. But if somebody’s got cancer
or potential cancer, it comes down to imaging
and obtaining a biopsy of the abnormal area
to see whether there is a cancer or not.
The third factor we need to remember is that
interpretation of a test result is dependent
on the clinical assessment.
So this is an x-ray that shows right mid-zone
shadowing. It looks like alveolar shadowing
so that is potentially consolidation. And
so if somebody is representing with a three-day
history of cough, fever and is hypoxic, this
x-ray will confirm that they have pneumonia.
However, if this patient has a several-month
history of cough, no fever and not much in
the way of hypoxial breathlessness, then this
isn’t a pneumonia. This is some other problem.
And that requires a further different range
of investigations than if the patient has
There are two parameters which we use to describe
how good a test is. One is sensitivity and
the other is specificity.
So sensitivity is the proportion of people
with the disease who will test positive. That
is how good is that test at identifying somebody
with that disease.
Specificity is the proportion of people with
positive results who actually have the disease.
So many tests, you do them and they reflect
somebody who has a disease but, actually,
they can also be abnormal in people who don’t
have the disease.
So specificity is how often a test, when it’s
positive, means the patient has that disease.
The likelihood the test has actually identified
somebody with the disease.
And, in general, no test is 100% sensitive
or 100% specific. And this needs to be borne
in mind when interpreting the results.
And this needs to be combined with the pre-test
probability of the disease being present.
And what I mean by that is that: is it likely
from the clinical assessment that that patient
has that disease? Because then that makes
the test much more useful.
So, for example, if we have a test which has
95% specificity and 90% sensitivity, that
means every time you use the test, for twenty
people there’ll be one false positive and
there will be two people with the disease
where the test is negative. So 10% of patients
with the disease will have false-negative
Now, if you apply this test – and you wouldn’t
as you can see from the data that’s coming
up – as a screening test for a disease in
a population of otherwise normal people, the
disease may only be present in 1 person in
every 10,000. So you test 10,000 people and
1 person will be positive. They’re likely
to come up positive with the result because
the test will identify them. However, there
will be 9,999 × 0.05 false-positive results
– people who give a positive result who
don’t have the disease. So essentially for
every person you identify with the disease,
there’ll be nearly 500 people who have a
positive results who don’t have the disease.
That test is essentially useless as a screening
Right. Now the tests that are available for
respiratory disease can do various things.
One is that we can measure physiological function.
And that’s basically the lung-function test
but it also includes blood gases, sleep studies
and some of the radioisotope scans are really
functional assays for the patient. And we
often use echocardiograms in patients with
lung disease to make sure they don’t have
cardiac disease as an alternative diagnosis
and to assess for the presence of cor pulmonale
Two: we can assess the lung anatomy and that’s
basically radiology. X-rays, CT scans, ultrasounds,
MRI but it also includes the bronchoscopy,
which is where we pass a flexible tube with
an endoscope on the end down through the windpipe
down into the lungs and actually visualise
the airways directly.
Three: we can obtain samples for histological
and cytological examination in the laboratory.
And those samples can be attained by different
methods and will give us an opportunity to
confirm exactly what the diagnosis is. And
these tests are particularly important for
patients with lung cancer.
Four: if somebody’s got an infection, we
can look for the pathogen. We do microbiological
tests: sputum, blood, pleural fluid can all
be cultured to see whether there is a bug
Then we do have various tests for immunological
status. Those are normally blood tests or
skin-prick testing. And then there are other
blood tests for general health. And importantly
we can actually measure the level of inflammation
present in somebody presenting with potential
lung disease using a C-reactive protein, white
cell count and the erythrocyte sedimentation
Sometimes we use genetic testing. That’s
for cystic fibrosis, ciliary disorders, alpha-1-antitrypsin
deficiency. And actually, increasingly that’s
being used directly on tumour samples to get
the tumour genetic status. And that’s relevant
now before the more specific treatments we
have for some of the more unusual tumour types.
And lastly, we just have to remember that
lung disease is only one cause of breathlessness.
And therefore you may need to have other tests
for people presenting with potential lung
disease: a blood count to make sure they’re
not anaemic; ECG, echocardiogram to assess
the cardiac function.