Lecture #30 - Screening Flashcards
What is a (mass) screening programme?
An organised system using a screening test among asymptomatic people in the population to identify a group who have higher probability of having the disease. This group undergo further diagnostic testing to see whether they do have the disease
So
- Not just randomly walk into clinic - that’s opportunistic screening. It’s organised
- Asymptomatic people look healthy
- Seeing the probability because screening tests aren’t diagnostic tests
- Screening test is not the same as a diagnostic test
Why do we try to detect the disease early? What is this an example of?
Draw that diagram of where screening occurs in the timeline in your head
In simple words, when do you screen and what is lead time?
So we can limit the consequences of disease through early diagnosis and treatment
Example of secondary prevention because have the disease but limit the consequences or make the prognosis better
- So, time on the x axis and there is no y. First we have the first biological onset of the disease (like, a few cells start dividing)
- Then there is a point at which the disease becomes detectable (through medical equipment) and it’s between the first biological onset and the symptoms appearing.
- Now comes the lead time which is when screening takes place and this occurs between the disease becoming detectable and symptoms appearing. This is where possibly better prognosis through early detection
- Then the symptoms appear and this is when people usually present.
- Then the outcome which is well, death
- The pre-clinical stage is all between the first biological onset and the symptoms appearing. The clinical phase is between symptoms appearing and the outcome
So with screening, you screen between when it’s detectable and when you usually seek treatment (aka when you become symptomatic). This is the lead time and it’s the additional time you have to treat people
- What are the 4 components to the screening programme?
- What is screening distinct from?
- What are some examples of mass screening programmes in NZ?
- Invitation and recall system then the screening test and then effective management strategy (follow people up if they are positive and then do diagnostic test and see if really have disease or not). Throughout all this is quality control and monitoring.
- Screening is distinct from opportunistic screening or case-finding. That’s when you just turn up and your doctor says, “Yeah, you’re old - let’s test you.”
- These mass programmes are aimed at entire pop = not just because you went to the doctor
- newborn heel prick (can detect if problem early so can give e.g. good diet from the start to help lead a better life
Fyi, what is prognosis?
The likely course of a medical condition
When should we screen - what 3 things to consider overall and what things to consider within those three?
- Is the disease appropriate?
- consider seriousness of the disease
- consider prevalence of pre-clinical disease
- consider lead time
- consider our ability to alter the course of the disease (look at the point where prognosis can be improved and see if there is effective therapy or treatment available) - Is the test appropriate?
- consider if test accurate
- consider if test acceptable and safe - Would the programme be effective?
- consider: are the resources to implement and cope with the positive people?
- consider: is the programme actually effective?
Is the disease appropriate: seriousness of the disease:
- Screening is ____-_____
- So makes sense to…..
- Resource-intensive
2. …….screen for diseases with potentially severe consequences (e.g. death)
Is the disease appropriate: prevalence of pre-clinical disease
- Screening is _____-____
- So more efficient when………
- But don’t have to look at prevalence when (2)
- Resource intensive
- When high prevalence of pre-clinical disease
- It’s serious disease (in its own right e.g. leads to retardation)
Or like, it’s cheap and easy to administer e.g. the heel prick (doesn’t look for prevalent outcomes but easy to do)
Is the disease appropriate: lead time
- Basically, what’s the length of the lead time got to do with screening?
Longer the lead time = greater than chance of detecting disease early
If short lead time, screening challenging and don’t get much benefit
If long lead time, you get more benefit and therefore better prognosis
1 Is the disease appropriate: ability to alter the course of the disease (critical points)
- What are critical points and where are the three places they could be?
- Where do you ideally want your critical point to be?
- It’s a point after which the disease process is irreversible and treatment will confer little or no benefit and it can happen at three different places:
- Happens between the onset and symptoms. Screening after it becomes detectable won’t be of much benefit because you’re just identifying people who have disease and just telling them before they need to know that they gonna die lol. You can’t alter their prognosis.
-Happens between symptoms appearing and outcome. Again, not much benefit because the whole point of screening is when the person doesn’t already have the treatment. Here, they usually have been diagnosed anyway.
(I think what they mean is like why spend money on screening early when they’ll be diagnosed before the critical point anyway? “if the critical point occurs after the time of usual clinical diagnosis, there is no need to detect the disease any earlier, given that treatment following usual diagnosis will be effective.
- Happens between detection and symptoms. Effective here. “In this situation it may be that picking up the disease early would improve outcomes and this is the aim for the screening programmes.”
2. Would ideally want the critical point later nearing the ‘symptoms appear’ part because you have more time to screen and treat before critical point
2 Is the disease appropriate: ability to alter the course of the disease (changing the course of the disease)
- Has to be some effective……
- Go read slide 15
- Screening has to……
- ……therapy or treatment available
3. Screening has to improve the length and/or quality of someone’s life
Is the rest appropriate: is the test accurate
- What can you measure accuracy with?
- What is the term given to these properties and why?
- Measure accuracy with sensitivity and specificity
- They are intrinsic to the screening test regardless of which pop it’s used in
- If have false positive or negative = their disease status is the opposite and so you’ve got their test result wrong
Is the rest appropriate: is the test accurate (SENSITIVITY)
- What is the definition?
- What is the formula?
Is the rest appropriate: is the test accurate (SPECIFICITY)
- What is the definition?
- What is the formula?
Sensitivity:
- Prop of people with the disease who test positive
- (a/a+c) so across column (looking at true positives)
Specificity:
- Proportion of people without the disease who test negative
- (d/d+b) so across column (looking at true negatives) - correctly reject people who don’t have disease
How to decide which is best sensitivity and specificity (want to maximise ____ but….)
How do you choose which one to maximise?
Want to maximise both and this can be achieved through
- Improving the screening test
- Choice of disease threshold
But there is a trade off between sensitivity and specificity to do with threshold. If you lower threshold, you’ll increase true positives (so increase sensitivity) but also increase false positives so you decrease specificity (specificity is about true negatives and you’re reducing them)
To choose which one to maximise - see which one of false + and false - has more serious consequence. So basically consider consequences of missing cases (false negatives increase i.e. threshold high) versus false alarms (false positives increase i.e. lower threshold)
- Maximise sensitivity when detecting as many cases as possible is important or if the costs or risks of next step (diagnostic) is not too high (like, a saliva test - all good, let heaps through even if they are false positives)
- Maximise specificity when cost or risks of next step are high (e.g. have colonoscopy in diagnostic step - don’t want heaps of people going through that)
What are the other measures to consider if test is accurate? (not the intrinsic properties like sensitivity and specificity)
Predictive values
- Measure test performance in a particular population (some tests work better in some pops)
- Measure what proportion of those that test +/- do/don’t have disease
(sensitivity and specificity were about ‘what prop of people with or without the disease the test correctly classifies’)
Is the test accurate:
- What is the definition of positive predictive value and what is the formula?
- What is the definition of negative predictive value and what is the formula?
- What are these two influenced by?
- Positive predictive value = proportion of people who test positive and have the disease so (a/a+b) and it’s across the row of positives
- Negative predictive value = proportion of people who test negative and don’t have the disease so (d/c+d) and it’s across the negative row
- Both are influence by the disease prevalence unlike specificity and sensitivity (if you increase prevalence of disease, you’ll increase PPV and decrease NPV). It’s like, if there is more people around with the disease and you test them - more likely to have disease if they test positive. Basically, sensitivity and specificity stays the same between high and low prevalence groups and if high prevalence, out of all those who test positive, more will actually have disease (you increase bigger circle and leave the little circle the same size and look at back page)
