Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Epidemiology > Study design in etiologic research > Flashcards

Study design in etiologic research Flashcards

1
Q

What are the three points in Design of data collection

A
  • Experimental or observation
  • Census or sampling
  • Time
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are the four types of epidemiological research questions? Are they descriptive of causal?

A
  • Diagnostic, predicts presence if disease (descriptive)
  • Etiologic, explains occurance of disease (causal)
  • Prognostic, predicts the course of a disease (descrptive)
  • Therapeutic/intervention, defines medical action that should be taken (causal)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is the design of data collection in Cohort studies?

A
  • Non experimental
  • Time
  • Census (most of the time) or sampling
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Definition of a cohort study

A

A group of subjects from whom data are collected and followed over time. A cohort can change in size, characeristics and membership over time. There is a difference in a fixed cohort and a dynamic population.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Describe a cohort with a fixed population.

A

Once a person experiences the defining event they remain part of the cohort as long as they are alive. A fixed cohort is changing over time in size (people die or are lost to follow up), age and gender distribution, but not in membership. E.g. birth cohort of babies born in 2017.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Describe a cohort with a dynamic population.

A

Persons can join and leave the cohort at every time. In a stable situation a dynamic cohort is not changing in determinants (age, sex) or size, but is changing in membership (new people replace old people). E.g. residents in a city, members of a health insurance plan

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the meaning of a retrospective of prospective data collection?

A

In causal research, the study design is always prospective (determinant comes before outcome). In diagnostic research outcome and determinant occur at the same time. Data collection is prospective when data is collected after the research question and retrospective/historical if data is collected before the research question.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Strengths of a cohort study (5)

A
  • able to calculate incidence in exposed/unexposed and true relative risk
  • Best design for rare exposure
  • Less subject to biases because outcome is not known. Since subjects are enroled before outcome, you can collect information on multiple determinants without recall bias
  • multiple effects of exposure can be investigated
  • temporal relationship exposure-outcome is clear
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Limitations of a cohort study (7)

A
  • Not suitable for rare outcomes
  • Generally time-consuming (except retrospective cohort)
  • Expensive (except retrospective cohort)
  • Long latency period
  • Change in exposure
  • Lost to follow up (slective)
  • Ethical considerations (because it takes a long time te perform)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Definition of domain?

A

The theoretical population to which the study findings apply (target population)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Definition of source population? And which two things are important for a source population in cohort studies?

A

The population from which the study population is sampled.
- The source population does NOT have to be representitive of any particular broader population
- Important is that the exposed and the unexposed groups come from the same source population in order to be comparable

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Definition of study population?

A

Subjects that are selected from the source population and are observed by the investigator. Should be part of the domain and source population and is the result of selection criteria and logistic circumstances.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Characteristics exposure distribution (4) in cohort studies

A
  • Exposure distribution in study population does not have to be representative for the exposure distribution in the source population or domain
  • you can select al exposed and sample unexposed persons
  • Power is highest when exposed and unexposed are of equal size (e.g. select)
  • Because the exposure is measured before outcome, there is slection but not slection bias
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is the main difference between a follow-up (cohort) study and a case-control study?

A

In follow-up study the slection is based on exposure status, in case-control it is on outcome

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Characteristics of follow-up in a cohort study (2)

A
  • Presence or absence of risk factors are determined before outcome occurs (t>0)
  • Follow-up is until the outcome, death, lost to follow up or end of the study
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Reasons for follow-up in a cohort study (3)

A
  • Determine if and when study endpoints happen
  • Determine which member are currently under observation
  • Obtain more information (change in determinant or confounders) –> completeness is essentioal because of loss of statistical pwer and could lead to bias when there is a relation between loss to follow-up and the outcome
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Strengths of a cohort study (5)

A
  • Able to calculate incidence in exposed/unexposed and true relative risk
  • Best design for rare exposure
  • Less subject to biases because outcome is not known. Since subjects are enroled before outcome, you can collect information on multiple determinants without recall bias
  • multiple effects of exposure can be investigated
  • temporal relationship exposure-outcome is clear
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Prospective vs retrospective cohort

A

Retrospective is not neccessarily a negative qualification, but sometimes prospective data collection may provide data that are qualitativly or quantitavily better because it is collected after the research question. They are not synonymous to case-control and cohort!

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What are the four typical summaries of exposure

A
  • Current level (yes, no; ever, never; current dose)
  • Average exposure (in a specific period)
  • Duration of use (of a minimum dose)
  • Cumulative exposure (sum of each exposure multiplied by the time; pack years)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What are the four typical summaries of exposure

A
  • Current level (yes, no; ever, never; current dose)
  • Average exposure (in a specific period)
  • Duration of use (of a minimum dose)
  • Cumulative exposure (sum of each exposure multiplied by the time; pack years)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Definition of induction time:

A

The time required for the effect of a specific exposure to become manifest, measured as the interval between first exposure - or achievement of cumulative exposure - and the occurence of overt disease.
- Study may be uninformative when follow-up is too short (or too long)
–> stratify for time since exposure to prevent dilution effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Latency period

A

The time interval during which a disease is present but not detected.
- Important in early detection/screening –> lead-time
- If survival is compared between screen-detected cases and clinically detectes cases –> lead time bias
–> i.e. screen-detected cases seem to live longer just because they were detected earlier, even if earlier treatment does not improve survival.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What type of frequency and association measures are used in a fixed cohort and a dynamic population?

A

Fixed cohort
- Cumulative incidence (denominator is fixed number of people contribute to the cohort) (assumption = all subjects are followed for the same amount of time)
- risk ratio
- risk difference
- incidence density and rate ratio when you have to deal with competing risks, lost to follow-up (long follow-up time)
Dynamic population
- Incidence density (denominator is total time of follow-up)
- Rate ratio
- Rate difference

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Which three assumptions does a rate have?

A
  • incidence rate is assumed to be constant over time-window in which it is calculated
  • If not: follow-up time should be defined in finer strata
  • It is reasonable to be constant for a particular Age category
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

How do you calculate incidence density when follow-up is long and incidence vary with age

A

Split time period into categories for when the incidence rate can assumed to be constant.
- Assumption: Homogeneity of person time: experience of individuals is interchangeable. Ten months of 1 individuals is equal to 1 month of 10 individual’s

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Thee aspects of vailidity

A
  • Selection bias: in prognosis between exposed and unexposed
  • Confounding bias: in determinants of the outcome
  • Information bias: in complete, valid and precise collection of data
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Thee aspects of vailidity

A
  • Selection bias: in prognosis between exposed and unexposed
  • Confounding bias: in determinants of the outcome
  • Information bias: in complete, valid and precise collection of data
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

Selection bias in cohort studies (4)

A

External validity if
- the non-response is related to exposure status and to the disease status (will affect disease rates and rate ratio in an over- or underestimation), but unlikly in cohort studies because of unknown outcome at the start.
- if there is a relationship between lost to follow-up and determinant or outcome
- more a problem in case-control than cohort
- you can not adjust, you have to prevent. You can perform a sensitivity analysis - worst/best case scenario (all follow up do have the outcome or not)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

Confounding bias in cohort studies

A

Effect estimate is distorted because mixed with effect of extraneous variable
- variable is associated with the determinant and the outcome, but not in a causal pathway

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

Dealing with confounding (3), (2)

A

Design: randomization, restriction/selection, matching
Data-analysis: stratification (rule of thumb 10%), multivariable analysis (many confounders)
- Confounders should be completely measured at baseline
- Keep in mind residual confounding (not measured confounders)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

Effect modification deffinition (3)

A

= statistical interaction
- e.g. the effect of smoking depends on gender
- if present, results should be presented stratified
- Interaction should be considered before start of study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

Information bias in cohort studies

A

Present when there are errors in obtaining information on exposure or outcome status, also called miclassification.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

Differential misclassification

A

Misclassification of outcome differs between expsoure groups
- in a cohort study: differential misclasification of disease causes a distortion of the relation between exposure and outcome that results from exposure dependent misclassification

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

Non-differential misclassification

A

Under ascertaiment only (decreased sensitivity, but perfect specificity of disease mearurement method)
- cases are only percentage of actual cases and precentage is similar for expsoed and unexposed: underestimation of both CI’s but no difference in relative risk
–> relative risk unaffected
–> risk difference underestimated
Over ascertainment only (decreased specificity, but perfect sensitivity of disease mearurement method)
- dilution with subjects without disease
–> bias in relative risk towards null (underestimation
–> risk difference also underestimated

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

Dealing with information bias in cohort studies (2)

A
  • use of standard techniques
  • use blinding of researchers
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
33
Q

Properties of causal diagrams (2)

A
  • Directed: the arrwos can be interpreted as direction of the causal effects
  • Acyclic: a variable cannot cause itself, either directly or through other variables (if you start in V1, you never come back to V1 only to V2, V3, V4…)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
34
Q

Definition cross-sectional study

A

Determinant and outcome are collected at the same point in time for each patient. Regarding to the theoretical design: a cross-sectional study van be etiologic or diagnostic. All individuals are observed only once.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
35
Q

Theoretical design of a cross-sectional study

A
  • Observational
  • t=0
  • census
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
36
Q

Purposes of cross-sectional studies (6)

A
  • Prevalence estimation/studies (descriptive)
  • Etiologic, causal inferences
  • Ecologic (etiologic): unit of observation is larger than individual
  • Diagnostic research/study
  • Reference range/reference values/normal range
  • Measurement of change (time, age)
36
Q

Purposes of cross-sectional studies (6)

A
  • Prevalence estimation/studies (descriptive)
  • Etiologic, causal inferences
  • Ecologic (etiologic): unit of observation is larger than individual
  • Diagnostic research/study
  • Reference range/reference values/normal range
  • Measurement of change (time, age)
37
Q

Purposes of cross-sectional studies (6)

A
  • Prevalence estimation/studies (descriptive) – population burder
  • Etiologic – causal inferences
  • Ecologic (etiologic): unit of observation is larger than individual
  • Diagnostic research/study
  • Reference range/reference values/normal range
  • Measurement of change (time, age)
38
Q

What is the most suitable design for diagnostic research?

A

Cross-sectional study

39
Q

In etiologic studies looking for causal inferences, what is prerequisite? (2)

A
  • In etiologic inference studies a stable determinant is needed (do not change over time).
  • Sample of domain, not necessarily random, selection can be efficient (but prevalence cannot be estimated).
40
Q

In which study purpose/design of cross-sectional studies is a representative sample of domain needed?

A

Prevalence estimation, reference values/range, diagnostic value. In diagnostic research, patients suspected for having the outcome, not the general population!

41
Q

What are strengths of a cross-sectional study? (6)

A
  • Saves time (no follow-up)
  • Saves cost (no follow-up)
  • No lost to follow-up
  • Recall of current exposure more precise, less recall bias
  • No variability over time (equipment, technicians, measurements)
  • No learning effect (because information is collected once)
42
Q

What are limitations of a cross-sectional study? (4)

A
  • t=0, what is causal effect?
  • induction time, current exposure to determinant be less relevant
  • Etiologic interference: cases with long duration are overrepresented
    –> determinant related to duration
    –> determinant related to fatality (survivor bias), severe cases die before sample
  • selection bias –> selective participation
42
Q

What are the cohort (3) and period effects (3) when studying change over time in a single cross-sectional study?

A

Cohort effect:
- each age is a cohort, but people in the past are more likely to have smoked when they were a teenager
- cohorts change: change in life style
- cohort studies: no problem
Period effect:
- change in measurement techniques
- Learning effects
- change in disease definition or surveillance

43
Q

What are the period effects (4) and chance in selection bias (2) when studying change over time in a cohort study?

A

Period effect:
- change in measurement techniques
- Learning effects
- change in disease definition
- change in surveillance
Selection bias
- selective mortality
- selective loss to follow-up

44
Q

What are the period effects (4) and chance in selection bias (1) when studying change over time in multiple cross-sectional studies?

A

Period effect:
- change in measurement techniques
- Learning effects
- change in disease definition or surveillance
- change in surveillance
Bias:
- In and exclusion criteria must be the same at baseline and follow-up

45
Q

Strengths of realworld data (4)

A
  • Rich data sources with a lot of information
  • No laborious & expansive data collection necessary
  • Direct reflection of patient population
  • Reflect actual behavior better
46
Q

Limitations of realworld data (6)

A
  • Data collection is not for research purpose
  • Missing data (because of the participant, or free text)
  • Context is absent
  • No quality control
  • Selection-bias
  • Privacy issues
47
Q

Definition of one health

A

Human epidemiology + veterinary epidemiology

48
Q

Explain the healthy worker effect in veterinary epidemiology.

A

If you want to check if heavy lifting causes backpain and you are sampling nurses, you do not find an association because nurses with backpain do not work. Your population is healthy and therefore selected (selection bias).
Farmers do active selection / culling in production level, many diseases and fertility.The farmer is sleceting the best cows to get the most out of it.

49
Q

Explain the occurance relation in veterinary epidemiology.

A

In veterinary epidemiology the disease is exposure and the productivity the outcome.

50
Q

How is the study population of a case control study composed?

A

Selection of two samples by the presence (cases) of absence (controls) of the outcome. The objective is prospective.

51
Q

In which forms of epidemiologic research is the case-control design used?

A

Mostly in etiologic, but also in diagnostic and prognostic. Not in intervention/therapeutic.

52
Q

What are the four steps in case-control studies?

A

Sampling is essential
1. identify cases
2. identify and sample controls from the same study base as the cases
3. Assess exposure in cases and controls
4. Calculate measure of association (odds ratio)

53
Q

When to perform a case-control study? (4)

A
  • Outcome is rare
  • Assessment exposure is expensive
  • many exposure categories of interest
  • long/unknown latency period
54
Q

What are the principles of a case control study?

A

Study base = imaginative pool where cases emerge from
Every time a cases occurs you fish them out of the pool and look for a control in the pool
Controls are sampled to obtain estimates of exposure/non-exposure in study base, and are therefore referents drom the study base

54
Q

What are the principles of a case control study?

A

Study base = imaginative pool where cases emerge from
Every time a cases occurs you fish them out of the pool and look for a control in the pool
Controls are sampled to obtain estimates of exposure/non-exposure in study base, and are therefore referents drom the study base
CASES ARE COMPARED WITH THE STUDY BASE, NOT THE NON-CASES!!!

55
Q

Describe the classic (but wrong) approach

A

Collect cases, collect non-cases, and compare them. Estimates may be correct, but non-cases do not represent the study base.

56
Q

Describe the modern (correct) approach

A

Collect cases. identify study bases from which cases emerge, sample controls from THAT study bases, compare them.

57
Q

Explain the terms primary study base and secondary study base.

A

Primary study base: clearly defined and numerated population (e.g. city, region, country, GP registry.). “easy sampling controls”
Secondary study base: defined by cases. Sampling controls is more difficult (e.g. catchment area for child cancer is bigger than for breastcancer). More difficult to conceptualize.

58
Q

Explain data collection for cases in case control studies (3)

A
  • Define appropriate criteria: what makes a case a case?
  • Index date: when are criteria fulfilled? At what day is my case becoming a case?
  • Diagnostic procedures or questions should not be influenced by the determinant. e.g. if is known that OAC give a higher chance in DVT, a GP would recognice a DVT more early when knows that some one is using AOC –> selection bias!
58
Q

Explain the terms incident cases and prevalent cases

A

Incident cases: wait for cases emerge and sample suitable controls (hard in rare disease) –> you study the determinant of getting the disease

Prevalent cases: are already present in study base. –> study the determinant of incidence and determinant of duration of disease (survival?). CAUTIOUS!: because you study the determinants of having the disease it is possible that these determinants cause survival/better prognosis/less severe cases (al others died and are not prevalent in study base but still can come but are underrepresented). Use when incidence is low.

59
Q

What do you have to think about when you sample cases (1)? Why would you sample cases and not take them all (4).

A

That you sample your controls in the same way!

  • When the outcome incidence is high.
  • When certrain determinant is rare
  • When data collection is expensive
  • in a stratified sample from (or restricted) those with particular categories of a confounder/modifier
60
Q

Explain the two situations in case control studies: dynamic population and closed cohort

A

Dynamic population: primary care center, district
Closed cohort: register, cohort

60
Q

Explain person moment sampling = incidence density sampling = risk set sampling (6)

A

= nested case control
- is ideally
- every time a case occurs you sample a control in the study base
- at that point of time the study base does only consist of non-cases
- non-cases/controls can become cases later on
- does not violate the study base principle
- OR = ID-ratio

61
Q

Explain sampling study base on regular time intervals or random date each case occurs (3)

A

= nested case control
- in extreme cases: if exposure does not change, you can sample the study base at 1 point in time
- assumption: exposure is constant over the sampling interval
- if the exposure incidence is 1/week you control sampling must be 1/week

62
Q

Explain sampling at the end

A

= nested case control
At t=1: identification of cases
At t=1: identification of non-cases, sample non-cases = controls

63
Q

In which five ways can you sample a fixed cohort?

A
  • risk set sampling
  • regular time interval sampling
  • sample at random date each time a case occurs
  • sampling at the end (= wrong, sample of non-cases)
  • at the beginnen –> case cohort
64
Q

Explain sampling in a case-cohort study

A

= nested case-cohort
At t=1: identify all cases
At t=0: sample of cohort = controls,
NB Some controls will become cases

65
Q

What are the advantages (1) and disadvantages (2) of sampling at the end?

A

Advantage: efficiency (exposure measured in sample only)
Disadvantage: “classical” approach (sampling non-cases)
–> in concept wrong
–> “rare disease assumption”

66
Q

What are the advantages (3) and disadvantages (2) of sampling in the beginning?

A

Advantage:
- efficiency (as all case-control studies) + you can use another subgroup for another research question
- valid sample from study base
- provides absolute rates (ratios, differences)!
Disadvantage:
- analysis less straightforward, but no problem
- less suited when many are lost to follow-up

67
Q

What are the criteria for being a control in a case control study (2, 3 notes)

A
  1. Controls should provide for a distribution of the determinant representative of the population from which cases emerge
  2. Controls should be identified as cases, if they develop the outcome under study during the study experience
    Note: Controls should not be representative of non-diseased (NOT of “general population”)
    Note: Sampling of controls from same (existing) registry as cases is efficient
    Note: Restriction or stratified sampling of controls should be analogous to cases
68
Q

What are the advantages (1) and disadvantages (4) of sampling of a registered population?

A

Advantages: sampling frame readily available (cases?)
Disadvantages: low response rate?, selection bias?, information bias? (recall bias), ethical issues: accessibility

69
Q

What are the advantages (4) and disadvantages (3) of sampling hospital controls

A

Advantage: efficiency, relatively high response, from same catchment population? (depending on the disease), limited recall bias?

Disadvantage:
- diseased by definition.
- Distribution of important factors (notably determinant) could differ from study base.
- I.e. referent disease associated with determinant and therefore do not give a good reflection of the exposure rate.
Example: Smoking or alcohol as determinants –> hospital controls smoke/drink more
Note: Avoid “cocktail” reference. Every reference disease should be verified (same catchment area, representative for study base, confounding). Every disease must meet above criteria

70
Q

What are the advantages (1) and disadvantages (4) of sampling population controls

A

Advantage
- Feasible when cases from same population eg. sudden death in Rotterdam
Disadvantage
- “expensive” (vs hospital controls): many calls
- Response relatively low
- Selection bias (exposured people are more likely to participate)
- Information bias? (recall bias?)

71
Q

What are the advantages (2) and disadvantages (1) of sampling special controls, e.g. spouses, friends, neighbors

A

Advantage:
- Relatively high response rate
- Efficiency
Disadvantage: if determinant made (more) comparable through selection referents –> dilution of effect
Examples: Smoking –> childhood asthma, radiation – leukaemia. Controls: neighborhood/sibling controls?

72
Q

Is sampling outside the study base possible?

A

Sampling control groups from outside study base is only possible when “external” sampling frame is comparable with study base concerning distribution of determinant (and confounders).

73
Q

How many controls are needed per case?

A

Number of controls per case refers to precision, NOT to validity
One per case often optimal, notably when:
- Data collection is expensive
- Case series is large
>4/5 per case of little additional value (increasing precision only marginally)

74
Q

Give the odds ratio. How do you interpret the odds ratio?

A

Ratio of odds of exposure among cases and odds of exposure among controls = OR = [a/b] / [c/d] = ad/bc. Interpretation of the odds ratio depends on moment of sampling controls.
- incidence density sampling/person moment sampling
OR = IDR
- Sampling from non cases (at end)
OR = +/- RR if outcome is rare
- Sampling at baseling
OR = RR

75
Q

Can absolute and relative measure of disease risk be calculated in a case control study?

A

In those instances where cases and controls are sampled from a population of known size, absolute and relative measures of disease risk can be calculated as in regular full cohort studies.

76
Q

Definition confounder

A
  • (Causal) risk factor for the outcome AND
  • Unequally distributed among comparison groups AND
  • No intermediate in the causal chain
77
Q

Confounding vs effect modification

A
  • Confounding is sort of bias, effect modification is part of nature
  • Confounding can be prevented in a design (RCT), effect modification can/should not be prevented and should be reported.
  • Confounding be adjusted by stratification and regression, effect modification should be reported (stratification).
    –> If you find an effect modification after stratifying, then you report the separate OR’s! You never pull two different OR’s together.
78
Q

What is a big misconception about case control studies in research?

A

When disease is exposure, it is a cohort study, not a case control study!
Study population
- Women and men referred for coronary angiography but no CAD
- Asymptomatic women and men matched on age, sex, GP
Outcome: Health care consumption

79
Q

Selection bias in case-control studies.

A

Comparability in prognosis between exposed and non-exposed: Not an issue in case-control studies as you start with the outcome (because you select on outcome)

Selective lost to follow-up (non-response): Not an issue in case-control studies as there is no follow-up

In case control studies: distortion that arise from procedures used to select study participants and from factors that influence participation. (when cases only become known to the researcher when they have certain determinants). The relation between exposure and outcome differs in those who participate compared to those who were eligible but do not participate.
Examples: Neighborhood controls, Friends/family as controls, Hospital-based controls

80
Q

Information bias in case-control studies.

A

Present when there are errors in obtaining information on exposure or outcome status: Also called misclassification (Comparibility in data-collection).
- Differential: misclassification differs between cases and controls or between exposed and unexposed.
- Non-differential: Misclassification is random
So, as you can see, not much different between cohort and case-control

81
Q

What is the objective of matching in case control studies?

A

not to limit confounding bias because there is no validity issue! Matching is to efficiently obtain information from study base in study of confounders/modifiers  efficiency issue!
One artificially makes diseased and “non-diseased” (sample study-base) similar; yet diseased and “non-diseased” differ according to many (prognostic) factors.

82
Q

Matching in cohort studies vs case-control studies

A

Matching exposed and non-exposed in follow-up studies “natural” and underappreciated (should probably done more)
Matching cases and controls in case-control studies “unnatural” and overrated (too many case control studies are matching their cases and controls)

83
Q

When do you match in case control studies?

A

Matching according to age, sex, blood pressure, body mass index, family history, exercise, smoking, etc. –> Cholesterol of cases and controls may then be similar as a result of matching procedure! –> Cases and controls should differ in many aspects !!!!!

Especially efficient when number of controls in confounder/modifier strata of base sample would otherwise be disproportional to number of cases in strata. E.g. when you want to look to sex differences and your cases or controls include low numbers of men you can match on sex to have more men.
ALTERNATIVE: Large unmatched sample (but is unefficiënt) –> you get oversampling of one part of the category (you don’t need more than 5x controls/case) and under sampling of the other part of the sampling (you have less than 1 control per case)

84
Q

What do you have to do if you match in case control studies for efficient reasons?

A

Matching often posed logistic problems, may be very inefficient and is often expensive.
In case you do match (efficiency reasons): perform CONDITIONAL data analysis
i.e. conditional on matching factors
i.e. adjusts for making cases and controls more similar

In OR’s you can only use disconcordant pairs.
(Exposed-Unexposed-pair)/(Unexposed-Exposed-pair)
Exposed-exposed and unexposed-unexposed do not count

85
Q

Disadvantages of a case-cohort approach (5)

A
  • less suited when many cohort members are lost to follow-up
  • analysis less straightforward, but no problem
  • reviewer’s ignorance of case-cohort approach
  • time of assessment of exposure in cases may be different for subsample cases:
    • bias?
    • e.g. blood samples
  • Controls are identified from the beginning: bias?
86
Q

Advantages case-cohort (5)

A
  • as in all case-control studies: efficiency
  • absolute risks and incidence rates can be calculated in exposed / nonexposed
  • several measures of association (notably risk differences)
  • multiple case-diseases (same baseline sample for several diseases)
  • rapid control selection
87
Q

Explain case-crossover study. When is it feasible (4)?

A

Mimics cross-over trial wherein all individuals exposed and unexposed. Each person acts as own control (no controls?). Differences with cross-over trial in 1) order not random: non-experimental study, and 2) person time sampling, not persons.

only feasible when:
- Exposure changes over time in individuals
- Outcome occur acutely
- Exposure effect transient (thus no “carry-over effect”) (you need a fast wash-out period, smoking is a bad example because when you stop smoking you still can have effects for years).
- No systematic change in exposure / confounders over time (or known and documented)
In a 2x2 table you match the cases and controls the same as in ‘matched case control studies’.

88
Q

Explain of case-only studies

A

In gene-environment interactions. 2x2 table of exposure cases to 2 determinants. Case-only odds ratio (ad/bc): multiplicative interaction?
Example: (= patient with neurodevelopmental disorder)

89
Q

Explain of case-only studies

A

In gene-environment interactions. 2x2 table of exposure cases to 2 determinants. Case-only odds ratio (ad/bc): multiplicative interaction?
Example: (= patient with neurodevelopmental disorder)

Epidemiology (2 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions