EPI820 EvidenceBased Medicine

1
EPI-820 Evidence-Based Medicine

• LECTURE 5 SCREENING
• Mat Reeves BVSc, PhD

2
Objectives

• Two components of screening (Dx Tx).
• Determinants of pre-clinical phase
• Concept of lead time
• Characteristics of screening tests (Se, Sp,
  yield)
• Biases - lead time, selection, length-biased
• The only valid measures of screening efficacy
• Evaluations and criteria for screening

3
I. Introduction

• Objective to reduce mortality and/or morbidity
  by early detection and treatment.
• Secondary prevention.
• Asymptomatic individuals are classified as either
  unlikely or possibly having disease.

4
Introduction

• Screening involves both diagnostic and treatment
  components
• Screening differs from diagnostic testing

5
Screening - Introduction

• There are two forms of screening which involve
  fundamentally different formats, organization and
  intent
• Mass or population-based screening
• Case finding
• This lecture covers mass screening for
  non-infectious (chronic) diseases only.

6
II. Characteristics of Disease

• The Pre-Clinical Phase (PCP)
• the period between when early detection by
  screening is possible and when the clinical
  diagnosis would usually be made.

7
Prevalence of Pre-clinical disease

• A critical determinant of the potential utility
  of screening.
• Prevalence is affected by
• disease incidence
• average duration of the PCP
• previous screening
• sensitivity of the test
• Example
• long PCP Colorectal cancer
• short PCP childhood diabetes

8
Lead Time
Lead time amount of time by which diagnosis is
advanced or made earlier
9
Relationship between screening, pre-clinical
phase, clinical phase and lead time
10
Lead Time

• Equals the amount of time by which treatment is
  advanced or made early
• Not a theory or statistical artifact but what is
  expected and must occur with early detection
• Does not imply improved outcome
• Necessary but not sufficient condition for
  effective screening.

11
Lead Time

• Impossible to determine lead time for any
  individual, can only compare distribution between
  screened and non-screened populations (RCT)
• Knowledge of expected lead time useful to
• Indicates amount of time diagnosis and treatment
  must be advanced
• Determine frequency of screening
• Example lead times
• Mam screening women 40-49 1-2 years
• Mam screening women 50-69 3-4 years
• Invasive Colo-rect cancer 7-10 years

12
Example of Estimation of Lead Time Distributions
within an Screening RCT
Cumulative Number of Cases
Years after Screening
Total lead time 5 4 5 2.5 2 18.5
years Average lead time 18.5/ 9 2.05 years
13
IV. Characteristics of screening tests

• a) Sensitivity (Se)
• Definition the proportion of cases with a
  positive screening test among all individuals
  with pre-clinical disease
• Influences
• the prevalence of pre-clinical disease
• the distribution of lead times
• Concept of the Sensitivity Function
• Average probability of detection for cases a
  certain time away from clinical diagnosis

14
Issues Related to Determining Sensitivity

• Determining the denominator
• who are the false negatives? (FNs)
• Cannot justify full work-up of negative test
  results
• Verification bias
• FNs estimated by counting number of interval
  cases
• Spectrum bias
• Se varies with spectrum of disease
• Screening intensity
• No previous screening more advanced PCP
  higher Se
• Se decreases with repeat screening

15
IV. Characteristics of screening tests

• b) Specificity (Sp)
• Definition the proportion of individuals with a
  negative screening test result among all
  individuals with no pre-clinical disease
• Imperfect Sp affects many (the healthy),
  imperfect Se affects few (the sick)
• Screening PVP usually low because prevalence of
  PCP is low
• Influences
• the number of false-positive test results
• the PVP and thereby the feasibility and
  efficiency of the screening program

16
IV. Characteristics of screening tests

• c) Yield
• Definition the amount of previously
  unrecognized disease that is diagnosed and
  treated as a result of screening.
• Another measure of screening efficiency .
• Influenced by
• Se
• Prevalence of PCP

17
V. Effects of screening on disease incidence
18
V. Effects of screening on disease mortality
19
VI. Evaluation of Screening Outcomes

• Study Designs
• RCT
• Compare disease-specific cumulative mortality
  rate between those randomized (or not) to
  screening
• Eliminates confounding and lead time bias
• But, problems of
• Expense, time consuming, logistically difficult,
  ethical concerns, changing technology.

20
VI. Evaluation of Screening Outcomes

• Study Designs
• Observational Studies
• Cohort
• Compare disease-specific cumulative mortality
  rate between those who choose (or not) to be
  screened
• Case-control
• Compare screening history between those with
  advanced disease (or death) and healthy.
• Ecological
• Compare screening patterns and disease experience
  (both incidence and mortality) between
  populations

21
Problems with Observational Studies

• Confounding due to health awareness - screenees
  are more healthy (selection bias)
• More susceptible to effects of lead-time bias and
  length-biased sampling
• Poor quality, often retrospective data
• CCS Difficult to determine appropriate control
  group
• Difficult to distinguish screening from
  diagnostic examinations

22
b) Measures of Effect of Screening

• Disease-specific Mortality Rate (MR)
• the number of deaths due to disease
• Total person-years experience
• The only gold-standard outcome measure for
  screening
• NOT affected by lead time,
• when calculated from a RCT - not affected by
  selection bias or length-biased sampling.

23
Effects of screening on disease survival - I
24
Effects of screening on disease survival - II
25
Effects of screening on disease survival - III
26
Biases that effect survival duration

• The efficacy of screening cannot be assessed by
  comparing survival rates (or CFR) because
• Selection bias
• Volunteers are more healthy
• Lead-time bias
• Introduced into survival experience of screen
  detected cases
• Length-biased sampling
• Screening preferentially identified slower
  growing or less progressive cases with a better
  prognosis

27
Length-biased Sampling
DX
DX
Worse Prognosis Cases
DX
DX
X
X
Better Prognosis Cases
X
X
SCREENING
28
VII. Pseudo-disease and Over-diagnosis

• Over-diagnosis
• Limited malignant potential
• Extreme form of length-biased sampling
• Examp Ductal-carcinoma in-situ
• Competing risks
• Cases detected that would have been interrupted
  by an unrelated death
• Examp Prostate CA and CVD death
• Serendipity
• Chance detection due to diagnostic testing for
  another reason
• Examp PSA and prostate CA, FOBT and CR CA

29
Over-diagnosis Effect of Mass Pap Screening in
Connecticut (Laskey 1976)
30
VIII. Assessing the feasibility of screening

• Acceptability
• convenience, comfort, efficiency, economical
• Efficiency
• Low PVP
• Potential to reduce mortality
• Effectiveness of treatment without screening
• Effect of competing mortality
• Cost-effectiveness
• Mam screening 30 50K /YLS

31
IX. Summary

• Three questions to ask before advocating
  screening

• Efficacy
• Effectiveness
• Cost-effectiveness

• Should we screen? (scientific)
• Can we screen? (practical)
• Is it worth it? (scientific, practical, policy,
  political)

32
Need and Feasibility of Screening?

• All cases of disease can be classified into three
  groups relevant to screening
• 1. Cure is necessary but not possible (Group I)
• 2. Cure is possible but not necessary (Group II)
• 3. Cure is necessary and maybe possible
  (Group III only group that could benefit!)

33
Lung Cancer?
34
Colorectal Cancer?
35
Prostate Cancer?