Block 10 Flashcards
Name some types of tests =
Biological - Hb, Ca2+, ALT
Images - CXR, mammogram, ultrasound
Questions - CAGE
Examination - tactile vocal fremitus
Tests act to
Shift our understanding on the continuum
A good test maximises what and minimises what
Maximises = true positive and true negatives Minimises = false positives and false negatives
True positive rate is the same as
Sensitivity
True negative rate is the same as
Specificity
Sensitivity definition =
Number of people with a positive test who actually have disease
Specificity definition =
Number of people with negative test who truly don’t hace disease
Sensitivity equation =
No of people with true +ve/all with disease
Specificity equation =
No of people with true -ve/all people without the disease
How can you summarise the impact of a test result?
PPV and NPV
PPV definition =
The change of really having disease with a positive test result
NPV definition =
The chance of really not having disease with a negative test result
PPV equation =
No. of true positives/all those who tested +ve
NPV equation =
No of true -ves/all those who tested -ve
Unlike NPV and PPV, sensitivity and specificity remain
Constant
What can change NPV and PPV?
Prevalence
What changes the prevalence?
Primary vs secondary care
Age
Country
Why are tests used differently in primary and secondary care?
DIfference in PPV and NPV
When prevalence increases …
PPV increases
NPV decreases
When prevalence decreases
PPV decreases
NPV increases
What can tell you how tests are best used in clinical practice?
Likiehood ratio
Each test has how many likelihood ratios?
2 - LR+, LR-
LR+ =
Sensitivity/false +ve rate
LR - =
Specificity/ false -ve rate
Chances of disease after test =
Chances of disease before test x LR
Screening =
Systematic appilation of a test to identify individuals at sufficient risk to warrent further investigation/prevention amongst persons who haven’t sought medical attention on account of symptoms.
Examples of screening
AAA Diabetic eye Blood spot Newborn hearing Newborn physical exam Fetal anomalies Breast, cervical, colorectal cancer
Screening is what type of prevention?
Secondary
When do we screen?
- More definitive tests are risky or unpleasant
- Limited opportunity for primary prevention
- Treatment limited
- Early treatment associated with better outcomes
Screening occurs in what phase of a disease?
Presymptomatic phase
Criteria to decide what is screened for can be broken into what categories?
- Disease
- Test
- Treatment
- Program
Screening: the condition must be
- Identifiable risk factor
- Latent period
- An important problem
- Epidemiology and natural Hx well understood
- All cost-effective primary prevention is exhausted
Screening: the test must be
- Simple, safe, precise
- Acceptable
- Have an identifiable cut off
- Agreed management following positive result
Screening: the treatment must be
- Evidence that early treatment causes better outcomes
- Agreed policies on who gets offered
- Optimised before screening
Screening: the program must be
- Acceptable
- Benefits>risk
- Cost effective
- Opportunity cost
- RCT evidence it reduces mortality and morbidity
Opportunity cost =
The loss of other alternative when one alternative is chosen
What needs to be done before a screening test is rolled out?
Clinical trial
Selection bias in screening RCTs
Healthy people tend to uptake screening
Length bias
What is length bias?
Screening tends to pick up slower progressing diseases and these diseases automatically have better prognosis
How do we get around selection and length time bias
Intention to treat/intention to screen analysis
What is the lead time?
Time between detection of a disease and clinically apparent symptoms
What is lead time bias?
Those with disease picked up through screening have a longer lead time so have a longer survival from diagnosis
How do we get around lead time bias?
Measuring deaths prevented rather than survival
Test for colorectal cancer =
Faecal occult blood
Flexi-sig
Test for prostate cancer =
Prostate specific antigen (PSA)
PPV of FOB
2%
PPV for PSA
30%
Who benefits from screening
- Not false -ve
- Not false +ve
- True -ve somewhat
- Some true +ves but not all
What is the problem with prostate cancer screening?
Catch too many slow growers or non-progressors.
What logic is at the heart of ethics?
Utilitarian
Autonomy issues with screening:
Are people actually chosing, especially children and infants
What is good about good screening?
Early detection can reduce risk of death or illness
What is bad about good screening?
False +ves
Over investigation and treatment
Some people get true +ves but this doesn’t prevent deaths
