Screening 1

Screening

What does screening mean?
Screening: The presumptive identification of unrecognized disease or defect by the application of test, examination, or other procedures which can be applied rapidly to a symptomatic population to sort out those who probably have the disease from those who probably not. ( A test that will help to identify a condition in an asymptomatic person)

A screening test is not intended to be diagnostic; and those with positive test sent on for further evaluation.
Diagnostic and screening tests have major differences due to the reasons for performing them. Diagnostic tests usually done on a patient who has a clinical problem, more likely to have the disease. While screening tests done on apparently healthy people, before they have any symptoms.

The goal of screening is to make people live longer or feel better

The basic purpose of screening for disease is to separate from a large group of apparently well persons who have the probability of having the disease under the study. The concept of screening is that early detection, before the development of symptoms → more favorable prognosis because the treatment began before the disease become clinically manifested → More effected → Reduction in morbidity and mortality. (2ndry prevention)

What do we screen for?

1) Risk factors for disease or conditions
2) Presence of disease or conditions

What makes a disease or risk factor an appropriate target for screening?

Disease must be:
Seriousneed to affect morbidity or mortality. e.g. no screening test for male pattern baldness because it will not affect morbidity or mortality.
Treatable. e.g. no screening tests for ALS (amyotrophic lateral sclerosis) or Huntingtons Disease because we cant do anything for the diseases.
And also treatment giving before symptoms develop should be more beneficial. (Early treatment is better than latewant to be able to intervene and improve morbidity and mortality): E.g. HT, TB, Ca breast, PKU..etc.
Pre-clinical detectable periodtest must be available to detect the disease in pre-clinical stage. e.g. HIV can be present for decades before the symptoms present themselves in the patient vs. pneumonia where disease and symptoms occur simultaneously
Prevalent


What makes a test a good screening test? Screening Tests must have the following characteristics
Validity (sensitivity/specificity), use the best test we can. e.g. PAP smear is not very sensitive, but conveniently cervical cancer is not a rapidly progressing disease. So if we dont pick up the disease the first year, we can the next year. But if we had a more sensitive test, we would use itdont toss a test because it doesnt have perfect sensitivity.
low-riskneed to include the risk of follow-up test on a false positive
Non invasive tolerable,
low-expenseefficiency consideration

Does more screening=better health care?

Effectiveness of Early Detection Even if a test accurately detects early-stage disease, one must ask whether this will benefit the patient. This highlights the importance of treatment efficacy. However, even with effective therapy, for screening to be worthwhile, early detection must offer added benefit over conventional diagnosis and treatment. A screening strategy has questionable effectiveness if patients seeking medical attention because of symptomatic disease have similar outcomes to those undergoing screening and subsequent treatment for early disease.
Screening tests are standardized examinations or tests used to identify patients who would benefit from the early detection of disease.

Accuracy of Screening Tests

Tests are valued because they discriminate one group of patients from another.
We start by identifying a "gold standard", usually a well defined pathologic or clinical finding that defines a disease, a condition, or a syndrome. Then we assess conditional probabilities, namely, the frequency with which various clinical findings and test results occur in these defined entities, (eg, the probability of post-operative bleeding in patients with prolonged partial thromboplastin times).

Although some tests can have several results and can be applied to patients who might belong to one of several different groups, it is often useful to consider a simplified situation in which the test is either positive or negative and in which a patient either has one particular disease or does not. In such simple cases, we use 2x 2 table to simplify calculations terms:
Results of screening test

Screening test Gold standard Test

Disease Disease Free
Total
Positive

Negative

(a) True Positive

(b) False Positive

Total Test positive
(a+b)

(c) False Negative



(d) True NegativeTotal Test Negative

(a+b)
TotalTotal Disease positive
(a+c)Total Disease Negative
(b+d)Grand Total
(a+b+c+d)
1- Validity
Sensitivity is the likelihood of a positive result in patients known to have the disease (TP+|dis)X 100


Specificity is the likelihood of a negative result in patients known to be free of the disease (TN-|no dis) X 100

2- Predictive values

Positive Predictive Value (PV+): Is the probability that person actually has the disease giving that he or she tests positive.
PV+ = a / a+b X 100%, is also called "Yield" of the test

Negative Predictive Value (PV-): Is the probability that person actually not has the disease giving that he or she tests negative.
PV- = d / c+d X 100%

3- Also we have:

The false-positive rate is the likelihood of a positive result in patients known to be free of the disease (pT+|no dis) X 100 and equals (1-specificity);
The false-negative rate is the likelihood of a negative result in patients known to have the disease (pT-|dis) X 100 , and equals (1-sensitivity).

Accuracy of test. (TP+TN) / Total

Thus, sensitivity and the false-negative rate describe how the test performs in patients with disease, whereas specificity and the false-positive rate describe how the test performs in patients without disease. The use of screening tests with poor sensitivity or poor specificity may result in potentially serious consequences due to false negative or false positive results. These in turn may lead to a false sense of security and inappropriate delay in treatment, or the patients unnecessary exposure to the risks of treatment. (When we should increase sensitivity?)

Example: Supposing one is interested in validating the use of CXR for the diagnosis of pulmonary TB. The gold standard for diagnosing TB is the culture of AFB from the sputum. To validate the use of CXR, we would have to select 200 TB suspect's people and perform both CXR and sputum cultures for them. The results given in the table:

Screening test
CXR Gold standard Test (Culture)
Disease Disease Free
TotalPositive


Negative(a) 80(b) 70(a+b) 150(c) 20(d) 30(c+d) 50 Total(a+c) 100(b+d) 100(a+b+c+d) 200
Sensitivity = a / a+c X 100%
= 80/100 X100 = 80%
Specifity = d / b+d X 100%
= 30/100 X100 = 30%
PV+ = a / a+b X 100%
= 80/150 X 100 = 53%
PV- = d / c+d X 100%
= 30/50 X 100% = 60%

False positive rate

False negative rate
Accuracy of the test

Relation of PV+ to disease prevalence

The PV+ of the screening test (yield of the test) is increased as the prevalence of the disease is increased as shown in the below table:

Disease prevalence

Test Result Gold standard Test
Disease Disease Free
Total
PV+


1%
Positive

Negative99 595594

17%194059406 Total100990010000
5%Positive

Negative495475970

51%590259030 Total500950010000

Prevalence affects predictive value (the chance of having the disease in the presence of a positive test)

The more prevalent a condition, the fewer false positive test there will be

The less prevalent a condition, the fewer false negative test there will be

EX: High Risk Population: Prevalence=40%
PPV=0.985 NPV=0.993

Low Risk Population: Prevalence=0.01%
PPV=0.0098 NPV=0.999


It is recommended to screen high risk population









Al Kindy College of Medicine

Community Medicine Department Lab. 6
Practical Lab. Of General Epidemiology

PAGE 

PAGE 5