Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
2 views

Sensitivity, Specificity

The document outlines a lecture on sensitivity and specificity in the context of epidemiology, detailing the differences between diagnostic and screening tests. It emphasizes the importance of assessing the performance of these tests through validity and reliability, and introduces key metrics such as sensitivity, specificity, positive predictive value, and negative predictive value. Additionally, it discusses the implications of test performance in relation to disease prevalence and the potential need for combined testing strategies.

Uploaded by

Deborah
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

Sensitivity, Specificity

The document outlines a lecture on sensitivity and specificity in the context of epidemiology, detailing the differences between diagnostic and screening tests. It emphasizes the importance of assessing the performance of these tests through validity and reliability, and introduces key metrics such as sensitivity, specificity, positive predictive value, and negative predictive value. Additionally, it discusses the implications of test performance in relation to disease prevalence and the potential need for combined testing strategies.

Uploaded by

Deborah
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

Tropical Health and

Epidemiology
MODULE: THEP 310, 220

LECTURE 7: SENSITIVITY AND SPECIFICITY

LECTURER: DR MWAYIWAWO MADANITSA. (MBBS, PHD)


Module Outline
▪ Epidemiology
- Definitions and concepts in epidemiology
- Degrees (levels) of endemicity (disease occurrence)
- Sampling
- Measures of disease (outcome) occurrence and frequency
- Measures of effect
- Bias, Confounding and Stratification
- Causation and causal associations
- Specificity and sensitivity
- Measure of vaccine efficacy
- Measures of disease transmission (R0 and Positivity rate)
- Screening and surveillance
Assessment of Diagnostic and Screening Tests
▪ Two (2) kinds of tests used for assessing people’s health
1. Diagnostic tests and
2. Screening tests

▪ Screening tests - primarily used in asymptomatic populations to determine the likelihood


of having or developing a particular disease
▪ primary function is to identify individuals that are at a higher risk of developing a
particular disease.

▪ Diagnostic tests - used to detect the presence or absence of disease in a symptomatic


individual
▪ warranted if results will change management of the patient or provide meaningful
clinical information about the patient’s expected outcomes or prognosis/survival
Assessment of Diagnostic and Screening Tests
▪ Differences between screening and Diagnostic Tests
Assessment of Diagnostic and Screening Tests
▪ It is important to assess the performance of diagnostic and screening tests (measurement
error) due to the implications for application

▪ Two principles of measurement (error) that are considered

1. Validity – how close the test comes to the true measure of the variable of interest
- Comparing results of test with true values from a “gold standard” test
- Gold standard test: the test or benchmark that is the best available under
reasonable conditions - the most accurate test possible without restrictions
- Gold standard test can be displaced with advances in technology for new test
methods

2. Reliability – how consistent the test is when used by different observers, time
periods or places.
Assessment of Diagnostic and Screening Tests
▪ Validity measurements can be done for
1. Continuous variables
2. Categorical variables

▪ Validity of continuous variables

- How well the test values correlate with the true values from gold standard test

- Measured by the Pearson correlation coefficient (r)

- r measures the strength and direction of a linear relationship between two variables
on a scatterplot, ranges from +1 to -1
Assessment of Diagnostic and Screening Tests
Assessment of Diagnostic and Screening Tests
▪ Validity of categorical variables

- Particularly concerned with inaccuracies of measurement leading to misclassification


of a result

- Assessed using 2 x 2 tables

- Measured using 4 metrics:


1. Sensitivity
2. Specificity
3. Positive predictive value
4. Negative predictive values
- expressed as proportions (%)
Assessment of Diagnostic and Screening Tests
▪ Sensitivity (Se)
- Proportion of true positive results correctly identified by the test, “True Positive (TP)
rate”
- Measure of ability of the test to correctly identify those who have the disease

▪ Specificity (Sp)
- Proportion of true negative results correctly identified by the test, “True Negative
(TN) rate”
- Measure of ability of the test to correctly identify those who do not have the disease

▪ Se = a/(a+c)
▪ Sp = d/(d+b)
Assessment of Diagnostic and Screening Tests
▪ Positive Predictive Value (PPV)
- Proportion of individuals with a positive result who actually have the disease
- Probability (patient having disease when the test is positive)

▪ Negative Predictive Value (NPV)


- Proportion of individuals with a negative result who actually do not have the
disease”
- Probability (patient not having disease when the test is negative)

▪ PPV = a/(a+b)
▪
▪ NPV = c/(c+d)
Assessment of Diagnostic and Screening Tests
▪ Sensitivity and Specificity are not related to prevalence of a disease… BUT

▪ PPV and NPV are related to the prevalence of a disease


Assessment of Diagnostic and Screening Tests
▪ Examples

Prevalence = =
Sensitivity (Se) = =
Specificity (Sp) = =
PPV = =
NPV = =
Assessment of Diagnostic and Screening Tests
▪ Under what circumstance would you really want to minimize the false positives?

When the costs or risks of follow-up therapy are high and the disease itself is not
life-threatening

▪ When would you want to minimize the false negatives?

If the disease is often asymptomatic and


1. is serious, progresses quickly and can be treated more effectively at early stages
OR
2. easily spreads from one person to another
Assessment of Diagnostic and Screening Tests
▪ Ideal test would have 100% sensitivity and 100% specificity…but this is not possible in a
single test
▪ Common practice to combine tests….
▪ initial “screening” test then
▪ a “confirmatory” test
▪ Sensitivity of combined test
= 1 - (1 - sensitivity of test 1) × (1 – sensitivity of test 2)
▪ Specificity of combined test
= 1 - (1 - specificity of test 1) × (1 – specificity of test 2)
▪ HIV testing algorithm
▪ Determine, Se = 99.4%, Sp = 99.6%
▪ Unigold: Se = 99.76% , Sp = 99.76%

Calculate the Se and Sp of the algorithm

You might also like