1. Introduction to Screening

Screening is defined as the active search for unrecognized disease or defects in apparently healthy people using rapid tests, examinations, or other procedures.

It is important to distinguish screening from diagnosis: * Screening: Applied to asymptomatic populations to identify those likely to have the disease. * Diagnosis: Applied to symptomatic individuals (or those with positive screen results) to confirm the disease.

The Iceberg Phenomenon: Screening aims to detect the “submerged” portion of the iceberg—the pre-symptomatic or undiagnosed cases—to intervene early and reduce morbidity and mortality.

2. Criteria for Screening (Wilson & Jungner)

Before a screening program is initiated, it must meet specific criteria. The classic criteria proposed by Wilson and Jungner (WHO, 1968) remain the gold standard:

  1. The Condition: Must be an important health problem.
  2. Natural History: The natural history of the condition should be well understood (from latent to symptomatic stages).
  3. Latent Stage: There should be a recognizable latent or early asymptomatic stage.
  4. Treatment: There should be an accepted treatment for patients with recognized disease.
  5. The Test: There should be a suitable test or examination.
  6. Acceptability: The test should be acceptable to the population (non-invasive, distinct).
  7. Policy: There should be an agreed policy on whom to treat as patients.
  8. Resources: Facilities for diagnosis and treatment should be available.
  9. Cost: The cost of case-finding should be economically balanced in relation to possible expenditure on medical care as a whole.
  10. Continuity: Case-finding should be a continuing process and not a “once and for all” project.

3. Types of Screening

  • Mass Screening: Screening the whole population (e.g., General health checkups).
  • High-Risk (Selective) Screening: Screening specific groups with known risk factors (e.g., Mammography for women over 50, LCCT for heavy smokers).
  • Multiphasic Screening: Applying two or more screening tests to a large population at one time.

4. Evaluation of a Screening Test

To evaluate a test, we compare it against a Gold Standard (the definitive diagnostic test). We use a 2x2 contingency table.

4.1 The 2x2 Table

Standard 2x2 Table for Screening Evaluation
Status Test_Positive Test_Negative
Diseased True Positive (TP) False Negative (FN)
Healthy False Positive (FP) True Negative (TN)

4.2 Key Metrics

Let’s simulate a real-world scenario. Suppose we are screening for Diabetes Mellitus in a population of 1,000 people using Fasting Plasma Glucose. * Prevalence: 10% * Sensitivity of Test: 80% * Specificity of Test: 90%

# Define population parameters
total_pop <- 1000
prevalence <- 0.10
sensitivity <- 0.80
specificity <- 0.90

# Calculate the 2x2 values
diseased <- total_pop * prevalence
healthy <- total_pop * (1 - prevalence)

TP <- diseased * sensitivity
FN <- diseased - TP

TN <- healthy * specificity
FP <- healthy - TN

# Create the matrix
matrix_data <- matrix(c(TP, FP, FN, TN), nrow = 2, byrow = TRUE)
colnames(matrix_data) <- c("Disease Present", "Disease Absent")
rownames(matrix_data) <- c("Test Positive", "Test Negative")

kable(matrix_data, caption = "Simulated Data: Diabetes Screening")
Simulated Data: Diabetes Screening
Disease Present Disease Absent
Test Positive 80 90
Test Negative 20 810

Definitions and Formulas

  1. Sensitivity (True Positive Rate): The ability of the test to identify correctly those who have the disease. \[ Sensitivity = \frac{TP}{TP + FN} \times 100 \]

  2. Specificity (True Negative Rate): The ability of the test to identify correctly those who do not have the disease. \[ Specificity = \frac{TN}{TN + FP} \times 100 \]

  3. Positive Predictive Value (PPV): The probability that a patient with a positive test result actually has the disease. \[ PPV = \frac{TP}{TP + FP} \times 100 \]

  4. Negative Predictive Value (NPV): The probability that a patient with a negative test result is truly free of disease. \[ NPV = \frac{TN}{TN + FN} \times 100 \]

# Calculating PPV and NPV based on the data above
ppv_calc <- (TP / (TP + FP)) * 100
npv_calc <- (TN / (TN + FN)) * 100

print(paste("PPV:", round(ppv_calc, 2), "%"))
## [1] "PPV: 47.06 %"
print(paste("NPV:", round(npv_calc, 2), "%"))
## [1] "NPV: 97.59 %"

5. The Effect of Prevalence on Predictive Values

This is a critical concept for MBBS students. Sensitivity and Specificity are intrinsic properties of the test. However, PPV relies heavily on the prevalence of the disease.

If you screen for a rare disease (low prevalence) in the general population, your False Positives will overwhelm your True Positives, drastically lowering your PPV.

Clinical Implication: This is why we do not screen the general population for rare cancers (like Pheochromocytoma) without specific indications. A positive result is more likely to be a False Positive than a True Positive.

6. Biases in Screening Evaluation

When evaluating if a screening program actually saves lives, we must watch out for biases.

A. Lead Time Bias

Screening detects disease earlier than clinical presentation. Survival time is measured from diagnosis to death. If screening diagnoses a cancer 2 years earlier, but the patient dies at the same age they would have otherwise, “survival time” appears increased by 2 years, but life was not actually prolonged.

B. Length Time Bias

Screening tends to detect slowly progressing diseases (longer pre-clinical phase) more often than rapidly progressive diseases (short pre-clinical phase). This makes screening cases appear to have a better prognosis, simply because we “caught” the slower/milder cases.

7. Real-Life Clinical Examples

Example 1: Cervical Cancer (Pap Smear)

  • Type: Cytological screening.
  • Success: One of the most successful screening programs globally.
  • Natural History: HPV infection \(\rightarrow\) CIN 1/2/3 \(\rightarrow\) Invasive Carcinoma (takes years).
  • Intervention: Excision of pre-cancerous lesions prevents cancer.

Example 2: Newborn Screening (Phenylketonuria - PKU)

  • Type: Metabolic screening (Heel prick test).
  • Criteria: Meets Wilson & Jungner perfectly.
    • Condition is severe (mental retardation).
    • Latent stage exists (birth to first few weeks).
    • Treatment exists (dietary restriction of Phenylalanine).
    • Result: Preventable intellectual disability.

Example 3: Prostate Cancer (PSA) - A Controversy

  • The Test: Prostate Specific Antigen.
  • Issue: PSA has low specificity (elevated in BPH, prostatitis). It also detects slow-growing cancers that might never have harmed the patient (Overdiagnosis).
  • Guideline: Most guidelines now recommend “Shared Decision Making” rather than mass screening, due to the risk of overtreatment (incontinence, impotence) outweighing the survival benefit in many men.

8. Summary

  1. Screening is for asymptomatic individuals.
  2. A good screening test must be sensitive (don’t miss cases) and specific (don’t scare healthy people).
  3. PPV depends on disease prevalence.
  4. Screening programs must be evaluated for Lead Time and Length Time biases to ensure they truly reduce mortality.

End of Chapter ```

Instructions for the User:

  1. Install R and RStudio if you haven’t already.
  2. Install the required packages by running this command in your R console: R install.packages(c("rmarkdown", "knitr", "ggplot2", "tidyverse"))
  3. Copy the code block above into a new file in RStudio.
  4. Save the file with the extension .Rmd (e.g., Screening_Chapter.Rmd).
  5. Click the Knit button (icon looks like a ball of yarn) in RStudio to generate the formatted chapter.