SCREENING Flashcards
Types of Screening
Mass
High risk/Selective/Targeted
Opportunistic/Disease finding
Multipurpose
Multiphasic
Uses of Screening
Case detective; Prescriptive
Control of disease: Prospective
Research
Education/Prevention
Criteria for choosing a screening test in relation to the disease
Significant burden in community
Detectable but has long pre-clinical phase
Support facilities and effective treatment are available
Evidence that early detection has outcome benefit
Test is appropriate for early detection
Natural history of disease is inadequately understood
Policy screening is in place
Criteria for choosing a screening test in relation to test itself
Inexpensive
Acceptable (invasive or non-invasive)
Yield valid and reliable results
Contingency Table
Columns represent disease presence/absence
Rows represent test results (negative/positive)
Sensitivity
Ability of the test to correctly identify who has the disease
TP / (TP+FN) * 100
Specificity
Correctly identifies who does not have disease
TN / (TN + FP) * 100
Problem of FN
False assurance that all is well
Disease signs and symptoms ignored, leading to detrimental outcome
Problem of FP
Undue and unnecessary tests, discomfort, anxiety, emotional trauma.
Thus the need to confirm screening results and suspicions with diagnostic test (gold standard test).
Sequential screening test
Two stage tests; screening test done first, then all positives brought back for second test to reduce false positives; increases specificity, but reduced sensitivity of screening test
Simultaneous Screening test
2 or more tests done in parallel; maximizes probability that true positives will be identified, increased sensitivity, but reduced specificity because false positives will also be identified.
Positive Predictive Value
Number of positives that actually have disease (TP)
TP / (TP+FP)
Probability that disease is present with positive test result
Negative Predictive Value
Number of negatives that actually do not have disease
TN / (TN+FN)
Probability that disease is absent with a negative test result
Likelihood Ratio
Percentage of sick with a result / Percentage of well with same result
Positive LR
relative probability of a positive test in a diseased person compared to a disease-free person
TP/FP
Sensitivity / 1 - Specificity
Negative LR
Relative probability of a negative test in a diseased compared to a disease-free
FN/TN
1 - Sensitivity / Specificity
Ordinal Variables/Levels
Describe patient’s characteristics, attitude, behaviour, status
eg. education, pain level, satisfaction level, agreement level, stage of cancer
Reproducibility/Repeatability/Reliability
How results of tests measured are identical or closely similar each time conducted
Can be affected by variations in lab procedures, observers, and test subjects
Intra-subject Variation
variation in result of testing same individual over a short period of time, due to environmental/physiological changes happening to individual during that period
Intra-observer Variation
variation in test results due to same observer examining results at different times
Inter-observer variation
variation in results due to multiple observers examining results
Yield
amount of unrecognized disease detected and brought to treatment as a result of screening
TP + FP / (TP + FP + TN + FN)
Lead Time Bias
Systematic error of apparent increased survival due to early detection (which could be due to improved screening) but without effective treatment
5-Year Survival
Proportion of patients alive 5 years after diagnosis
Generally accepted to mean improved treatments
Influenced by improved screening tests capable of detecting subclinical cases
