Skip to content
Scintria
ToolsGrant writerFunding
HomeLiteratureClinical trialsEvidence LabProtocol packData analysisLearnProjectsLibraryReferencesApplications
Start free
ScintriaStart free

Learn / Epidemiology / Screening and diagnostic tests

Evaluating screening and diagnostic programmes in practice

12 minute read · 6-question quiz

Download this lesson as a colour Word or PDF study guide with the figures and answers. Included in paid plans; reading and quizzes stay free.Sign in

What you will learn

  • Map a diagnostic or screening pathway as a cascade and find where people are lost

  • Interpret programme indicators such as coverage, yield and positivity

  • Apply these ideas to HIV, TB and cervical cancer programmes

A test does not help anyone unless the person is reached, tested, receives the result, is linked to treatment and stays in care. Programme evaluation therefore follows the cascade from eligible people to the final outcome.

10,000Eligible women aged 30 to 494,000 (40%)Screened400 (10% positivity)Screen positive340 (85%)Received results255 (75%)Treated or referred
An illustrative cervical screen-and-treat cascade. Only 2.6% of eligible women completed the pathway, even though the test itself works.

Reading the cascade

Of 10,000 eligible women, 4,000 are screened (coverage 40%). The test is positive in 400 (positivity 10%). 340 get results and 255 are treated. Overall, 255/10,000 = 2.6% of eligible women reached the end; the biggest loss is at the first step (60% never screened). Improving the sensitivity of the test would help little: the programme needs outreach, not a better test. A cascade shows which stage to fix first.

Standard indicators

IndicatorDefinitionWhat it shows
CoverageScreened ÷ eligibleReach
PositivityPositives ÷ testedWorkload and test performance; compare across sites
YieldNew cases found ÷ testedEfficiency
Linkage to careLinked ÷ diagnosed (within a set time)Success of referral
RetentionIn care at 12 months ÷ startedContinuity
Turnaround timeFrom sample to result receivedDelays that cause loss
Interval cancersCases arising between screensMissed or fast-growing disease (sensitivity of the programme)

The HIV 95-95-95 cascade

The UNAIDS targets are that, by 2030, 95% of all people living with HIV know their status, 95% of those diagnosed receive sustained antiretroviral therapy and 95% of those on treatment have viral suppression. Multiplying the steps: 0.95 × 0.95 × 0.95 = 0.857, so 86% of all people living with HIV would be virally suppressed. Several East and Southern African countries have reported results close to these targets in national population surveys. A shortfall at any step lowers the final figure: 90-90-90 gives only 73%.

The TB cascade

The TB care cascade from estimated incidence: people with TB who are notified (the 'missing millions' who are never diagnosed); notified who start treatment; those who complete and are cured. Tools to close the gap: community screening, symptom screening for all, Xpert and other rapid molecular tests, contact investigation and TB preventive treatment.

Think in cascades

Any programme is the product of its steps. A 90% result at five steps is only 59% at the end (0.9⁵). Small losses accumulate. Measure every step, find the biggest absolute loss, and address it.

Facility data can mislead

Denominators in routine data are often unreliable: 'eligible' populations come from census projections, and double counting or transfers inflate numbers. Triangulate facility data with surveys (such as the Uganda Population-based HIV Impact Assessment) and spot-check records.

Evaluation vocabulary

Process indicators (coverage, positivity, linkage) tell you whether the programme was delivered; outcome indicators (mortality, incidence, suppression) tell you whether it worked; impact needs a comparison group.

Summary

  • A screening or testing programme works through a cascade; evaluate every step.

  • Coverage, positivity, yield, linkage, retention and turnaround time are core indicators.

  • The UNAIDS 95-95-95 targets multiply to 86% viral suppression among all people living with HIV.

  • Losses accumulate: five steps at 90% leave 59%.

  • Fix the largest absolute loss first; triangulate routine data with surveys.

Check your understanding

Question 1
1

The product of 95%, 95% and 95% is about:

Question 2
2

Screening coverage is:

Question 3
3

In the example cascade, the largest loss occurs at:

Question 4
4

A cascade with five steps each at 90% retains about:

Question 5
5

Interval cancers measure:

Question 6
6

The best way to check the accuracy of routine facility denominators is to:

Draft study guide. Spotted an error or something unclear? Tell us at hello@scintria.com. This is educational material, not clinical or regulatory advice.

Validity and reliability of testsThe randomised controlled trial