Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

Year 1 Epidemiology > 1 > Flashcards

1 Flashcards

(63 cards)

Start Studying
1
Q

Null Hypothesis

A

Two sets of data/ results are unrelated (always true in statistics until other test done)

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

P Value

A

The probability that if the null hypothesis is true, the sample would be the same or higher as the statistical results just due to chance.

Tested to check if Probability that null hypothesis is true –> Cut off for statistical significance = 0.05

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

Confidence Intervals

A

Intervals in which the true result of test is likely to be present

When repeating the test, chances are that in 95% of the repetitions, results would be in the 95% confidence intervals

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

Case control and cohort studies: compare and explain the features of case-control and cohort studies, list the individual strengths and weaknesses and evaluate their appropriateness for answering research questions

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

Incidence

A

Number of new cases in a defined population within a defined period of time

(Number of new cases / total number of population)

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

Prevalence

A

Number of cases of disease within a defined population at a specific point of time (include new and pre-existing), often expressed In percentage of the total population

–> Proportion !

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

Mortality

A

The incidence of deaths in a certain time frame

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

Case-control studies

(way it is done + pro + con)

A

One Group of people with disease and one group of people without disease and they are being compared for exposure

–> Retrospective

Pro: cheap, quick, good for rare diseases, investigation for many exposures at once

Con: selection bias in choosing control, recall bias (retrospective), uncertainty of exposure-diseases-time relationship, poor for rare exposures, no direct calculation of incidence rates

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Standardised mortality ratios definition + explenation

A

Observed deaths / Expected deaths

–> higher than 1 = higher than expected

Comparison of death rates in observed group in comparison to standard population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Cohort Studies (design, pro, con)

A

Select two groups –> expose one to exposure, other don’t –> see how many with/ without in each group

Prospective

Pro: Multiple outcomes, rare exposures, incidence calculation, minimization of bias in estimating exposure (prospective), natural history of disease

Con: Expensive, time-consuming, Healthy worker/volunteer bias, not for rare disease, bias when loosing follow up

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

Clinical trial design

A

planned experiment in humans to measure the efficiency of effectiveness of an intervention

Requires:

  • Control group
  • prospective
  • randomization
  • both groups (intervention + control) followed for the same time
  • ideally (double) blinded
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Case-Control Studies (design, pro-con)

A

Compare people with disease and people without disease for exposure

Looking for cases with and without disease and then checking for exposure afterward

Retrospective

Pro:

Cheap+ quick, Suitable for rare diseases, Investigation for many exposures at once

Con:

choosing control –> selection bias, Recall bias (retrospective, the uncertainty of exposure - diseases -time relationship, Poor for rare exposures (if not many people are exposed to one thing), Can’t calculate incidence rates directly

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

Relative risk (calculation + interpretation)

A

calculation + data can be collected from a cohort study:

Relative risk = incidence rate in the exposed/

the incidence rate in unexposed

RR=1 –> no difference

RR>1 –> exposure causes a higher risk for disease

RR<1 –> exposure causes lower risk for disease

–> DIVISION (vs subtraction in attributable risk)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Odds ratio (calculation + interpretation)

A

Can be calculated with data from Case-Control Studies

—> anhalts Punkt für relatives risikso, jedoch night so genau–> bei seltenen Erkrankungen schätzwert für relatives Risiko

odds Krank/odds gesund

Odds krank: Exposed / Nonexposed cases

Odds gesund: Exposed /Nonexposed cases

OR= no difference

OR>1 = exposure is a potential risk factor for disease

OR<1= exposure potentially protective for disease

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

Difference between Systematic review and meta-analysis

A

Systematic review: systematic collection of data

Method of providing a summary of existing data + evidence–> should be reproducible and limit bias

Metha analysis Combines statistical + quantitative analysis of data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Hierarchy of scientific evidence

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Attributable risk

A

Attributable risk= how much higher the frequency in exposed vs unexposed

Attributiutbale risk = incidence exposed - incidence unexposed

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

List possible causes for association

A

Chance

  • chance might lead to a false association between exposure and disease
  • Chances for testing: e.g. p values(–> chance?) and CI –> where is real value?

Bias

  • systematic error in the system (design, the conduct of study)

Confounding

  • a third factor that gives false association between exposure and disease
  • AT trial design this can be influenced by Randomization and matching (same age, sex etc in case-controls)

Causation

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

List and explain the different forms of Biases

A

1. Selection bias

false selection of participants (e.g. different exposures in case and control group in a case-control study) 2. Recall bias

2. Recall bias

Recall of exposure is difficult, biased via false, disturbed memory of participants

3. Observer bias

more relevant in RCT, the researcher is aware of treatment given (not blinded) –> biases interpretation of results, symptoms

4. Information /measurement bias

differences in the collection of data so that there is a different quality to the data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Bradford Hill criteria

A

Criteria used to consider whether an association is due to causation

  1. The strength of association –> the stronger the more reliable
  2. The consistency of association –> same finding in different trial designs and different population etc.
  3. Specificity of association –> one to one relationship between cause and outcome
  4. Temporal relationship to association –> risk factor must occur during or before the disease
  5. Dose-response relationship –> higher dose= greater risk
  6. Biological plausibility–> logical explanation?
  7. The coherence of association–> absence of conflict with other knowledge
  8. Experimental evidence (reversibility) –> remove of risk factor decrease in risk?
  9. Analogy – other similar associations? –> absence only lack fantasy, no real association
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Methods to Attempt to identify confounders

A

Association is due to third factor –> confounder

It can be dealt with at

  1. Study design
  • Randomization in RCT –> confounding factor is the same in both groups
  • Matching in Case-Control –> Try to match basic characteristics like sex age–> fewer confounders
  1. Data analysis
  • Stratification –> risk is calculated individually for each variable
  • Standardization –> e.g. adjusting for age structure in country etc
  • Regression analysis
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What are the principles of screening?

A

Try to detect early stages of the disease in healthy individuals

It is:

  • detect people at risk
  • cheap, noninvasive, simple
  • healthy, asymptomatic individuals

Ideally: High sensitivity and high specificity – BUT High Sensitivity = Low Specificity

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

Sensitivity in screening

A

Ability to pick up true positives (miss as fewer people as possible)

Sensitivity = True positives / (True positives + false negatives –> all people who have disease)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Specify in screening

A

The ability of a test to correctly identify those without disease

Specitiy = True negatives / (True negatives + False positive)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Criteria for the screening programme
Disease: * \* An important health problem. * \* Has a well recognised pre-clinical stage. * \* The natural history is well understood. * \* The disease has a long latency period (asymptomatic period). Screening test: * \* Valid (sensitive & specific). * \* Simple and cheap. * \* Safe and acceptable. * \* Reliable. Diagnosis and Treatment: * \* Facilities are adequate. * \* Effective, acceptable and safe treatment. * \* Available. * \* Cost effective and sustainable.
26
The validity of a screening test
An indication of the extent to which a measure is a true indicator of what it purports to measure (the disease) Depends on * Specificity * sensitivity * Predictive value --\> How likely is patient to have the disease given the test is positive or negative Depends on Specificity, Sensitivity **and Prevalence in population)**
27
Different approaches in screening --\> who is offered screening test?
* Mass screening – applied to the whole population (defined by age or gender). * Targeted screening – selected sub-groups thought to be at increased risk. * Systematic screening – the population is called for screening with a register --\> Can be mass or targeted in nature * Opportunistic screening – population approached when they make contact for another reason.
28
Different Levels of Prevention
1. Primordial – prevention of factors that promote the emergence of risk factors. e.g. banning alcohol consumption. 2. Primary – altering behaviors associated to exposure to risk factors by use of campaigns or health promotion. e.g. campaigns to reduce alcohol consumption. 3. Secondary – procedures that detect and treat pre-clinical pathological changes and thereby limit disease progression. e.g. prescription of statins after a heart attack. 4. Tertiary – methods of easing symptoms of a disease. e.g. stroke rehabilitation.
29
Why does evidence-based medicine matter?
Why EBM matters to clinicians * Patient * Medical Knowledge * Practice-Based Learning and Improvement * Interpersonal and Communication skills * Professionalism Also: evidence-based decision making in politics
30
Phases of a clinical trial
1. Test safeness of treatment ( only a small number of people, often healthy volunteers) 2. Test for effectiveness (short-term) and safety (often few hundred people with disease) 3. Compare new drug with current treatment or placebo, continue monitoring side-effects (several thousand people from different backgrounds, different locations) 4. After marketing: gather more information about more populations and monitor long-term side effects
31
Effectiveness
Results in a day-to-day basis/environment
32
Efficacy
Results under ideal conditions (outcome in trials)
33
Experimental Event Rate
Data collected in clinical trials EER= incidence in the intervention arm
34
Control event rate
In clinical trial: CER= Incidence in the control group
35
The relative risk in clinical trials (how to calculate)
EER/CER = Incidence in the intervention arm/ Incidence in the control arm
36
Relative reduction of risk (in clinical trials)
(CER-EER)/CER --\> Incidence in control-incidence in intervention / incidence in control
37
Absolute risk reduction (in clinical trials)
CER-EER control group incidence - intervention group incidence
38
Number needed to treat in a clinical trial
1/ARR (Absolute risk reduction) Number of patients you need to treat to prevent one bad outcome Example: NNT=5 --\> 5 patients need to be treated to prevent one stroke to happen
39
Summarise information of a paper in a scientific journal
A sentence for each of the following will often suffice: * Why did they do it? * What did they do? * What did they find? * What did they conclude? * In your opinion, was the study conducted well?
40
Evaluate the quality if research paper: What to look for, which criteria to ask?
General Appraisal Checklist (full in notes) 1. Question – Relevant? Hypothesis? 2. Design – appropriate? 3. Population – sample size? 4. Methods – measure? Appropriate? 5. Analysis – appropriate statistical tests? 6. Confounders 7. Bias – measurement, selection 8. Ethics – ethical? Consent? 9. Interpretation – casual inference (Bradford Hill)
41
Which protocols are used for evaluation of which kind of study?
RCT: Consort Observational study: STROBE Systematic review/meta-analysis: PRISMA/MOOSE
42
Pooled risk estimate
Can be generated from meta-analysis when there is a lot of data out there --\> taking into account several studies to get to Pooled risk
43
Phases in a systematic review
1. Planning * defines the research question to be addressed 2. Finding Research * Identification of research --\> define search criteria to include all published data * Selection of studies --\> inclusion/exclusion criteria * Study Quality assessment (based on recognized or user-defined criteria) 3. * Reporting and dissemination – Study details need to be abstracted and details tabulated to show a summary of the findings * --\> We can estimate an overall effect by combining the data in a Meta-analysis
44
Meta-analysis
Meta-analyses combine the published estimates of effect from each study to generate a **pooled risk** estimate. * may not be possible/ misleeding if studies to topic a too heterogenous Advantages: * more reliable and precise estimate of effect * Differences (heterogeneity) between published studies can be identified and explored In a meta-analysis, the effect of each individual study is pooled to produce a weighted average effect across all studies. A Forest plot is the most common way of presenting the results from a meta-analysis. \* Each study is represented by a box and line – the size of the box corresponds to the weight given to that individual study; the horizontal lines correspond to the 95% confidence interval. \* The overall estimate from the meta-analysis is usually shown as a diamond at the bottom of the plot. The centre of the diamond and dashed line corresponds to the summary effect estimate; the width of the diamond represents the confidence interval around this estimate.
45
Biases and limitations of meta-analysis
_Publication bias_ * studies are less likely to be published if no significant findings Also: * lack of heterogeneity * low study quality * inconsistency of results --\> May distort meta-analysis
46
Mass screening programmes in the UK
* **Cervical Cancer** - every 3 years for women aged 25-49, every 5 years to women 50-64 * **Breast Cancer** - every 3 years for women aged 50-70, women aged 70 and over can self-refer * **Bowel Cancer** - every 2 years for men and women aged 60-74 * **Abdominal Aortic Aneurysm** - offered to men in their 65th year * **Diabetic Eye Screening** - offered to people with type 1 or type 2 diabetes over the age of 12
47
Why do you screen?
* Early diagnosis of disease where early intervention improves prognosis * Identification of high risk individuals where intervention improves prognosis * Identification of those posing a risk to others e.gscreening for infectious disease e.ghepatitis B
48
Predictive value in screening test
Proportion of tests that are correct
49
The positive predictive value in screening
* Positive predictive value =likelihood that a patient with a positive test result will actually have the disease * PPV= a/(a+b) --\> True positives / All positives (TRUE+ False positives)
50
The negative predictive value in screening
* Negative predictive value =likelihood that a patient with a negative test result will not have the disease * NPV= d/(c+d) True negatives / all negatives (true + false)
51
Attributional risk
Number of cases which could be prevented without the exposure --\> Incidence in the exposed- incidence in unexposed SUBTRACTION (VS division in Relative risk)
52
Which plot shows results in meta-analysis
Forest plot
53
How is heterogeneity shown? (in which graph)
Galbraith plot
54
How is publication bias shown?
Funnel plot
55
List four main approaches to intervention
Clinical intervention Health Education Healthy public policy community development
56
What is Clinical intervention?
Biomedical (classically thought of under the category Prevention-but others can be prevention too!)
57
What is health education?
The traditional type of health promotion (knowledge- attitudes-behavior-practice).
58
What is community development?
Is an approach to intervention ## Footnote Individuals setting up their own initiatives e.g. Sports for children
59
What is Healthy public policy?
Approach to intervention Legal, fiscal and regulatory (HIA, European directive). e.g. making health care more/easier accessible for everyone
60
Top 6 infectious diseases
Leon had Diarrhea that made mad. 1. Lower respiratory infections- 3.9M 2. HIV/AIDS-2.8M 3. Diarrheal diseases -1.8M 4. Tuberculosis- 1.6M 5. Malaria- 1.2M 6. Measles-0.6M
61
Glod standard
the term is given to a definitive test which exposes the true disease status of an individual
62
Hosts and treatment of schistosomiasis 8+ problem)
Hosts: Humans and snails Treatment: Praziquantel Problem: Larve in bladder: might cause bladder cancer
63

Year 1 Epidemiology (1 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions