Quiz
PUBH Important Concepts
Dieses Quiz enthält Fragetypen, die sich im gewohnten Player öffnen.
Fragen in diesem Quiz (35)
1You can protect randomisation in an RCT by 1) concealing allocation and 2) using intention to treat analysis
- ARichtig
- BFalsch
2If you do not follow intention to treat analysis, what do you follow?
- APer protocol analysis
- BRandom analysis
- CDefective route analysis
3Potential sources of bias:
[blank_start]Lack of blinding[blank_end] - if participants/researchers know which treatment
[blank_start]Loss to follow up[blank_end] - no data/someone leaves study, could underestimate harms of treatment
[blank_start]Non-adherence[blank_end] - if people do not follow protocol, may not learn of efficacy- ALack of blinding
- BLoss to follow up
- CNon-adherence
4_______ is the genuine uncertainty about a benefit or harm of intervention.
- AClinical equipose
- BClinical equity
5[blank_start]Internal validity[blank_end] comprises of if the findings are real or if they could be explained by chance, bias, or confounding. [blank_start]External validity[blank_end] is considered afterwards and evaluates the generalisability of a study.
- AInternal validity
- BExternal validity
6As sample size increases, likelihood of getting a representative sample [blank_start]increases[blank_end]. The accuracy of the parameter estimates also [blank_start]increases[blank_end] i.e. 95% confidence interval has a [blank_start]smaller[blank_end] range.
- Aincreases
- Bdecreases
- Cincreases
- Ddecreases
- Esmaller
- Flarger
7Which statement best explains confidence intervals?
- AA CI represents the range of values the true parameter will lie in 95% of the time (if we continue to repeat with the same samples)
- BA CI represents the range of values the true parameter will exist in (if we continue to repeat with the same samples)
- CA CI is used to estimate a true average
8When interpreting 95% CI, if the null value [blank_start]lies within[blank_end] the interval, then the findings are consistent with chance as an explanation and are [blank_start]not statistically significant[blank_end].
If the null value [blank_start]does not lie within[blank_end] the interval, then it is unlikely the value is due to chance and so it is [blank_start]statistically significant[blank_end].- Alies within
- Bdoes not lie within
- Cnot statistically significant
- Dstatistically significant
9The null hypothesis (Ho) states there is [blank_start]no association[blank_end] between exposure and outcome. The alternate hypothesis (Ha) states there is [blank_start]an association[blank_end].
P values test the null hypothesis. P values are the probability that differences are due to chance if the [blank_start]null hypothesis[blank_end] is true.- Ano association
- Ban association
- Cnull hypothesis
10The p value is the probability that null hypothesis or alternate hypothesis is true.
- ARichtig
- BFalsch
11A Type I error is finding a statistically significant result when there is actually [blank_start]no association (Ho is true)[blank_end]. There occur [blank_start]5%[blank_end] of the time if we use 0.05 as a p value.
A Type II error is finding a statistically significant result where there actually is [blank_start]an association (Ha is true)[blank_end].- Ano association (Ho is true)
- Ban association (Ha is true)
- Can association (Ha is true)
- Dno association (Ho is true)
- E5%
- F105
12Which concept is described by "any systematic error which results in an incorrect estimate between the exposure and risk of disease"?
- ABias
- BConfounding
- CChance
13Of the three main types of bias, which can only be controlled for in the design phase?
- ASelection and information bias
- BInformation and publication bias
- CSelection and recall bias
14Selection bias is a [blank_start]systematic[blank_end] difference between 1) people included in the study vs. not or 1) people included in the [blank_start]exposure vs. control[blank_end] groups.
Loss to follow up can create a selection bias. To minimise loss to follow up, you can collect [blank_start]alternate[blank_end] details when recruiting participants, maintain [blank_start]regular[blank_end] contact with participants, and make [blank_start]several[blank_end] contact attempts.- Asystematic
- Bexposure vs. control
- Calternate
- Dregular
- Eseveral
15Case-control studies are susceptible to [blank_start]recall bias and selection bias[blank_end]. Since the exposure and outcome have already occurred, and cases and controls are selected [blank_start]separately[blank_end], it is important that selection is [blank_start]independent[blank_end] of exposure status.
When selecting cases, you must ensure high [blank_start]participation[blank_end], have a clearly defined [blank_start]population of interest[blank_end], and have a reliable method of ascertaining cases.
When selecting controls, you must ensure the controls are from the [blank_start]same[blank_end] defined population, and have the same inclusion/exclusion [blank_start]criteria[blank_end] as the cases, and once again ensure high participation.- Arecall bias and selection bias
- Bseparately
- Cindependent
- Dpopulation of interest
- Eparticipation
- Fsame
- Gcriteria
16Measurement error creates [blank_start]information[blank_end] bias. A participant, interview technique, or data collection instrument could cause measurement error.
Measurement error can be systematic or random. Accurate data is collected when there is no [blank_start]systematic[blank_end] error. Precise data is collected when there is no [blank_start]random[blank_end] error.- Ainformation
- Brecall
- Csystematic
- Drandom
17An accurate but unprecise study had little to no systematic error, but had random errors.
- ARichtig
- BFalsch
18Non-differential misclassification is systematic i.e. misclassification occurs [blank_start]equally between groups[blank_end]. Differential misclassification means that groups had [blank_start]different levels of misclassification[blank_end].
- Aequally between groups
- Bdifferent levels of misclassification
19Which of these is not a method to minimise recall bias?
- AUsing objective measures
- BValidating self reported measures with other information
- CUsing memory aids
- DStructured questionnaire and standard prompts
20Which of these is not a method of minimising interviewer bias?
- AClearly defined study protocol
- BInterviewer training
- CBlinding
- DObjective measures
21Confounders are: associated with the exposure, associated with the outcome, but not on the causal pathway.
- ARichtig
- BFalsch
22The Simpson's paradox is when a confounder reverses the association between an exposure and outcome.
- ARichtig
- BFalsch
23Ways to reduce confounding at the design phase:
1. [blank_start]Restricting[blank_end] who can enter the study
2. [blank_start]Randomisation[blank_end] of groups (eg RCT)
3. [blank_start]Matching[blank_end] (1:1 or frequency)- ARestricting
- BRandomisation
- CMatching
24Ways to control for confounding at the analysis stage:
1. [blank_start]Stratification[blank_end] (grouping)
2. [blank_start]Standardisation[blank_end]
3. Regression models/[blank_start]Multivariable analysis[blank_end]- AStratification
- BStandardisation
- CMultivariable analysis
25It's possible to over-match when matching.
- ARichtig
- BFalsch
26Steps of a systematic review:
1. [blank_start]Formulation of a clear question[blank_end]
2. Develop [blank_start]search[blank_end] strategy
3. [blank_start]Diagnostic test[blank_end] strategy
4. Search for and [blank_start]identify relevant studies[blank_end]
5. [blank_start]Data extraction[blank_end]
6. [blank_start]Individual and group crit. appraisal[blank_end]
7. [blank_start]Synthesis of findings[blank_end]
8. [blank_start]Interpretation of results[blank_end]- AFormulation of a clear question
- Bsearch
- CDiagnostic test
- Didentify relevant studies
- EData extraction
- FIndividual and group crit. appraisal
- GSynthesis of findings
- HInterpretation of results
27Choose the incorrect statement about levels of prevention.
- APrimordial prevention is establishing and maintaining conditions that minimise harm to health.
- BPrimary prevention aims to limit the incidence of disease by controlling risk factors.
- CSecondary prevention aims to reduce the prevalence of disease, by shortening duration, severity, and occurrence.
- DTertiary prevention aims to reduce the time spent with disability.
28Examples of types of prevention:
[blank_start]Secondary[blank_end] = Weight loss programmes
[blank_start]Primary[blank_end] = reduce HIV cases by promoting condoms
[blank_start]Primordial[blank_end] = reduce car crash injuries by providing good infrastructure for cycling
[blank_start]Tertiary[blank_end] = Stroke rehabilitation- ASecondary
- BPrimary
- CPrimordial
- DTertiary
29An exposure with a large population attributable risk is a good target for intervention.
- ARichtig
- BFalsch
30Choose the incorrect statement about prevention.
- AIndividual prevention has a favourable benefit-to-risk ratio.
- BIndividual prevention is behaviourally inappropriate.
- CPopulation prevention is radical.
- DPopulation prevention is radical is behaviourally appropriate.
- EPopulation prevention is has a high benefit-to-risk ratio.
31Surveillance can be [blank_start]universal[blank_end] (the whole population) or [blank_start]sentinel[blank_end] (in certain sites eg labs). It can also be [blank_start]active[blank_end] (an established system) or [blank_start]passive[blank_end] (based on health care worker reports).
- Auniversal
- Bsentinel
- Cactive
- Dpassive
32When deciding if screening is appropriate we must consider:
1) [blank_start]The seriousness of the disease[blank_end] (ie don't screen for colds)
2) The ability to [blank_start]alter the course of disease[blank_end]
3) [blank_start]Lead time[blank_end] (in what period can we detect the disease)
4) [blank_start]Prevalence of preclinical disease[blank_end]- AThe seriousness of the disease
- BPrevalence of preclinical disease
- CLead time
- Dalter the course of disease
33A screen test has good [blank_start]sensitivity[blank_end] if it has a high ratio of people diagnosed with the disease, over all people with the disease - diagnosed and undiagnosed. (a / a+c or, true positives / false negatives + true positives)
A screen test has good [blank_start]specificity[blank_end] if it has a high ratio of people without the disease diagnosed as disease free, over all people with the disease. (d / b+d or, true negatives / false positives + true negatives)- Asensitivity
- Bspecificity
34Sensitivity & specificity are [blank_start]intrinsic[blank_end] properties of the test. Predictive values measure test [blank_start]performance[blank_end] in a
particular population. The positive predictive value is the true positives over the true positives and false negatives ([blank_start]a / a+b[blank_end]). The negative predictive value is the true negatives over the false positives and true negatives ([blank_start]d / c+d[blank_end]).- Aintrinsic
- Bperformance
- Ca / a+b
- Dd / c+d
35Predictive values are influenced by disease prevalence.
- ARichtig
- BFalsch
Informationen
- Beschreibung
- Quiz am PUBH Important Concepts, erstellt von Mer Scott am 25/10/2017.
- Autor
- Mer Scott
- Sprache
- English
- Aufrufe
- 6
- Schlagwörter
- Bewertung
- Noch keine Bewertungen Bewerte dieses Quiz