Introduction to screening

Question 1

Screening

Answer 1

Screening is the application of a test to people who are as yet asymptomatic for the purpose of classifying them with respect to their likelihood of having a particular disease.
aka
Screening involves testing people who don’t have symptoms yet to see if they might have a certain disease.”

Question 2

UK national screening committee

Answer 2

People from a specific group, who might not realise they could get a disease or its complications, are asked questions or offered tests. This helps find out who would benefit more than be harmed by additional tests or treatment.”.

Question 3

Why do we screen?

Answer 3

To reduce morbidity or mortality from the disease among the people screened

Break the chain of transmission and development of new cases

Early detection can be cost effective

Not intended to be diagnostic

Results of screening trigger diagnostic work-up and preventive interventions

Question 5

3 types of screening

Answer 5

mass screening:
directed at a whole population or subgroup irrespective of disease (i.e newborn hearing checks, screening in pregnancy etc)

selective screening :
focused on groups at high-risk of developing the disease (i.e gestational DM screening in pregnancy, annual mammography for BRCA carriers)

case finding/ opportunistic screening:
when consult for other reason (i.e check BP when come for flu jab, oral cancer screening by dentist)

Question 6

Eff

Question 7

Effects of screening

Answer 7

lead time: time by which the diagnosis is early due to screening

early treatment: resulting in postponing or avoiding the disease

Question 8

Disadvantages of screening

Answer 8

over- diagnosis (i.e mammogram)

mis-diagnosis (False positives (e.g. Down’s syndrome, psychological harm)
False negatives – (eg false reassurance → late diagnosis in cancer)

psychological harm

Question 9

criteria for screening (3 considerations)

Answer 9

The disease
- the disease should be serious i.e causes death, disability or discomfort)
- the natural history should be understood
- the disease must have a recognisable latent period before symptoms appear (or early symptomatic stage)
- The latent period between first signs and overt disease should be long enough that screening significantly advances the detection of disease
- prevalence of the disease (screening not rewarding for extremely rare diseases) but there are exceptions (i.e PKU)

Diagnosis and treatment
- Facilities need to be available
- There is an available, effective, acceptable, and safe treatment
- Treatment at early stage better

The screening test
- Should be valid (accurate)
- Reliable
- Simple and cheap
- Safe and acceptable (benefit outweighs harm)

Question 10

Feasibility of screening

Answer 10

• Screening procedures: convenient and free of discomfort or risk; attractive to target population
• Efficiently and economically
• Treatment available
• High level of case detection (sensitivity) and reasonably low level of false-positive tests (specificity)

Question 11

Criteria for screening

Answer 11

Common disease

Substantial morbidity/mortality

A preclinical detectable phase

An effective treatment

Improvement in prognosis by treatment in PCDP

A test with good performance

…and which is tolerable (side-effects, acceptable)
…and cost-effective

Question 12

Measures of screening test performance

Answer 12

Validity
Sensitivity

Question 13

Validity

Answer 13

The ability of a screening test to accurately identify individuals with and without disease
Validity is measured by sensitivity & specificity

Question 14

Sensitivity

Answer 14

Ability of the test to identify those with disease correctly
Increase in sensitivity will increase number of true positives.

Sensitivity = true positive / (true positive + false negatives)

Increase in sensitivity will increase false positive (i.e decrease specificity)

Question 15

Specificity

Answer 15

Ability of test to identify those without disease correctly
Determines whether or not the frequency of false positives will be low enough for a screening programme to be feasible.
Determines the number of false positives.

A test with high specificity will have few false positives
An ideal screening test would have 100% sensitive and 100% specific (ie no false + or false-), but in practice these are usually inversely related. It is possible to vary the sensitivity by changing the specificity

Specificity: true negative/ true negative + false positives

Increased specificity means increase in false negatives (i.e decrease sensitivity)

Question 16

Receiver Operator Characteristic curves (ROC)

Answer 16

Evaluates test over range of cut-off points
–> performance of a screening test
–> method for comparing screening tests

Plots sensitivity against false-positive rate (1 – specificity) and illustrates the trade-off that exists between them

Question 17

AUC

Answer 17

Area under the curve is a measure of the test’s accuracy:
AUC= 1 indicates a perfect test
AUC = o.5 indicates no better than chance alone

Question 18

Summary sensitivity and specificity

Answer 18

Calculated in different populations
Both are needed to understand test performance
Do not add up to one – but gain in one is loss in the other
They are not fixed

Question 19

Performance (predictive value)

Answer 19

Provides information on how sure we are about a test result (how predictive is the test result itself?)

PPV & NPV

Question 20

Positive predictive value

Answer 20

Proportion of people with a positive test who have the disease in question.

A high PPV indicates that a reasonably high proportion of the costs of a programme are being expended for the detection in the preclinical phase.

A low PPV indicates that a high proportion of the costs are being wasted on the detection and diagnostic evaluation of false positives.

PPV = TP/(TP+FP)

Question 21

Negative predictive value

Answer 21

Proportion of people with a negative test who do not have the disease in question.

An NPV close to 1 indicates that testing negative is reassuring as to the absence of disease and that re-screening may not be worthwhile.

NPV = TN/(TN+FN)

Question 22

Summary PPV and NPV

Answer 22

Determinants of the PPV and NPV for a given screening test are the specificity and sensitivity of the test and the prevalence of disease in the target population

High prevalence = higher the PPV (the more likely a positive result predicts presence of disease)

Low prevalence = PPV will also be low, even when using a test with high sensitivity and specificity  wasted resources

Question 23

Evaluating a screening programme: essential considerations

Answer 23

Study design

• Observational studies (case-control & cohort studies) – commonly used but subject to bias
• Randomised Controlled Trials (RCTs) – ‘gold standard’

Study population & outcome and exposure assessment are also important

Question 24

Selection bias

Answer 24

People who choose to participate in screening tend to be healthier, have healthier lifestyles, and their outcomes tend to be better because of this

‘Healthy volunteer’ effect, ‘worried-well’

Question 25

Length time bias

Answer 25

The seemingly improved survival results from finding many slow-growing tumours in diseases that take a long time to develop symptoms.

Question 25

Lead time bias

Answer 25

The apparent better survival observed for those screened because diagnosis is being made at earlier point in natural history of disease.

Question 26

RCTs

Answer 26

Main strength of randomization is that the distribution of measured and unmeasured factors that could impact disease outcome is randomly distributed among the screened and unscreened groups