Population screening and interventions

Question 1

Which level of prevention is screening?

Answer 1

Secondary - prevents disease > illness

Question 2

Definition of screening

Answer 2

Identifying apparently healthy people who have or are at an increased risk of developing disease

Test is not 100% accurate - aim for high Sn, Sp less important

Question 3

Diabetic eye screening UK

Answer 3

Annually for diabetics >12

Question 4

Cervical screening UK

Answer 4

every 3 years 25-49

Every 5 years 50-64

Question 5

Breast screening UK

Answer 5

Every 3 years from 50-71

Question 6

Bowel cancer screening UK

Answer 6

55 - one-off scope

60-74 - every 2years

Question 7

AAA screening UK

Answer 7

Men on 65th birthday

Question 8

NSC criteria for condition

Answer 8

1. Important health problem
2. Known natural Hx (incl. carriers)
3. Primary prevention in place

Question 9

NSC criteria for test

Answer 9

1. Valid
2. Acceptable
3. Agreed cutoff
4. Agreed diagnostic pathway

Question 10

NSC criteria for treatment

Answer 10

1. Early intervention -> better outcomes
2. Agreed health pathway (who is offered treatment and how)

Question 11

NSC criteria for programme

Answer 11

1. Adequate facilities
2. All other measures to improve outcome taken
3. Cost-benefit analysis - outweighing opportunity cost
4. Benefit > physical and psycho harm from testing
5. Acceptable to population and professionals
6. Evidence-based for improved outcomes
7. Information available to the public re: accuracy and implications
8. Quality assurance plan

Question 12

Potential biases in screening programme evaluation

Answer 12

Self-selection: Healthier people more likely to participate in screening

Lead time bias - increased survival length due to earlier dx rather than improved outcome, look at mortality rate instead

Length bias - improved survival from detecting longer conditions (e.g. slow-growing, less aggressive cancers)

Question 13

Population attributable risk

Answer 13

Risk in whole population (Exposed + unexposed) - risk in unexposed

c.f. attributable risk E-U

Question 14

Population attributable fraction

Answer 14

PAR/risk in whole population

Question 15

Nuffield ladder of population-level health interventions

Answer 15

1. Do nothing
2. Inform and educate
3. Enable choice
4. Make healthier option the default option
5. Incentivise healthier options (financial or other)
6. Disincentivise unhealthy options
7. Restrict choice
8. Eliminate choice

Question 16

Drawbacks of natural experiments for assessing population-level health interventions

Answer 16

Use before-intervention as control

Increasing noise to get from exposure > health outcome as loss of control > insufficient power for distal relationships

Cannot establish causality

Question 17

Preventable fraction

Answer 17

Mathemeatially identical to RRR

Absolute risk reduction / event rate

Question 18

Attributable risk

Answer 18

Absolute risk: Exposed - unexposed

Question 19

Attributable fraction

Answer 19

attributable risk / risk in exposed population

Question 20

relative risk reduction

Answer 20

absolute risk reduction / total risk in exposed population