Tölfræði (EBM)

Question 1

Hvernig lítur histogram út?

Answer 1

Súlurit!

Question 2

Hvað er likelihood ratio?

Answer 2

Þetta er tengt grafinu sem Hjalti er alltaf að sýna.
Likelihood ratios are defined as the ratio of the probability of a test result among patients with a target disorder to the probability of that same test result among patients who do not have the target disorder.

Question 3

Hvað er Fagan nomogram (likelihood ratio nomogram)?

Answer 3

The likelihood ratio nomogram (or ‘Fagan’ nomogram) enables a post-test probability to be graphically calculated if the pre-test probability and likelihood ratio are known. A line can be drawn connecting the pre-test probability and likelihood ratio, and this will intersect at the post-test probability on the right hand side:

Question 4

UK national census er dæmi um hvernig rannsókn?

Answer 4

Cross-sectional study

Question 5

Er hægt að nota cross-sectional rannsóknir til að meta relative risk á að fá sjúkdóm? En cause vs effect?

Answer 5

Neibb, hvorugt.

Question 6

Hver er munurinn á parametric vs non-parametric statistical significant tests?

Answer 6

Statistical significance tests can be either parametric or non-parametric. Parametric tests usually assume that data is normally distributed, whereas non-parametric tests are usually based on rank order.

Question 7

3 dæmi um parametric significance tests:

Answer 7

• Unpaired student’s t
• Paired student’s t
• Pearson’s test

Question 8

5 dæmi um non-parametric signifance tests:

Answer 8

• Mann-Whitney U
• Wilcoxon paired
• Sign
• Spearman’s rank
• Kendall’s test

Question 9

Aðeins er hægt að nota chi-square test með hvernig data?

Answer 9

Nominal data (líka kallað categorical data)

Question 10

Hvað er Gaussian distribution?

Answer 10

Normalkúrfa sem fylgir týpískri “bjöllulögun”.

In a Gaussian distribution:

• The median is the middle point of the observations
• The mode is the most commonly observed measurement
• The mean is the arithmetic average

In a Gaussian distribution 95% of the observations fall within +/- 2 standard deviations of the mean.

Question 11

Hvað er relative risk? Hvað mælir það?

Answer 11

Relative risk, or risk ratio, (RR) is used to compare the risk in the two different groups. It is the ratio of the absolute risks of the disease in the treatment group (ART) to the absolute risk of the disease in the control group (ARC). It measures the strength of association between a factor and an outcome.

Question 12

Hvernig er number needed to treat reiknað út?

Answer 12

The NNT is defined as the inverse of the absolute risk reduction (ARR) and can therefore be calculated by:

NNT = 1 / ARR

Question 13

Hvað er specificity?

Answer 13

Sértæki.

The specificity is the proportion of true negatives correctly identified by the test.

Question 14

Hvað er kaplan meier graf og hvernig lítur það út?

Answer 14

Notað til að bera saman survival hjá hópum sjúklinga sem fá mismunandi meðferð.

Question 15

Hvað er Hospital Episode Statistics í UK?

Answer 15

Hospital Episode Statistics (HES) is a data collection process that involves the collection details of all hospital admissions, outpatient appointments and A&E attendances at NHS hospitals in England.

Question 16

Hvernig reiknar maður likelihood ratio fyrir positive vs. negativt test?

Answer 16

The likelihood ratio for a positive test = sensitivity / (1-specificity)

The likelihood ratio for a negative test = (1-sensitivity) / specificity

Question 17

Hvernig er sensitivity reiknað?

Answer 17

Test positives / true positives

Question 18

Hvernig er specificity reiknað?

Answer 18

Test negatives / true negatives

Question 19

Ef próf er með minna en 100% specificity, þá fáum við falskt…

Answer 19

…jákvæð próf

Question 20

Ef próf er með minna en 100% sensitivity, þá munum við fá…

Answer 20

…falskt neikvæð próf

Question 21

Hvað er positive predictive value?

Answer 21

Sá hluti test positives sem eru í raun með sjúkdóminn.

Question 22

Ef próf er með minna en 100% specificity, þá fáum við falskt…

Answer 22

…jákvæð próf.

Question 23

Hvað er negative predictive value?

Answer 23

Sá hluti test negatives sem er í raun ekki með sjúkdóminn.

Question 24

Hvað er Type I error?

Answer 24

Möguleikinn á að hafna null hypothesis þegar hún er í raun sönn (þeas að trúa að effect sé til staðar þegar það er í raun ekki). Gerist í 1 af 20 tilfellum þegar við notum p gildi undir 0,05.

Question 25

Hvað er Type II error?

Answer 25

Að hafna EKKI null hypothesis þegar hún er EKKI sönn (þeas það er raunverulegur effect til staðar en við fundum hann ekki í þessari tilraun og verðum því að segja að það sé ekki munur). Gerist t.d. þegar rannsóknin er ekki með nægt power.

Question 26

Hvað er receiver operating characteristic curve? Hvað er það notað til að meta?

Answer 26

A receiver operating characteristic (ROC) curve is a graph used to assess diagnostic tests. The cut off point for a positive and negative test is varied and then the sensitivity and specificity for each of these cut off is calculated. A perfect test has a ROC curve that passes through the upper left corner (100% sensitivity, 100% specificity). Therefore the closer the ROC curve is to the upper left corner, the higher the overall accuracy of the test.

The area under the curve is a measure of how good the test is:
1 = perfect test
0.5 = test no better than chance

The y-axis of the ROC curve is the true positive rate, which = sensitivity

The x-axis of the ROC curve is the false positive rate, which = 1 – specificity

Question 27

Hvernig eru case control studies gerðar og fyrir hvernig sjúkdóma eru þær sérlega gagnlegar?

Answer 27

A case-control study is a type observational study in which two groups of patients, one with the disease and one without, are compared on the basis of a proposed causative factor that occurred in the past. They are therefore retrospective in nature and are useful in hypothesis generation.

They are suitable to be used when investigating a rare disease or as a preliminary study in cases where little is known about the disease and the proposed aetiological factor. They can look at multiple risk-factors (exposures) but can only look at a single outcome.

Question 28

Hver er munurinn á case control stúdíum vs. RCT og prosepctive cohort studies?

Answer 28

Compared with randomised controlled trials, case-control studies are generally relatively inexpensive to run but provide less evidence for causal inference.

Compared with prospective cohort studies, case-control studies are usually less expensive and also shorter in duration.

Question 29

Positive predictive value breytist m.a. eftir því…

Answer 29

Hversu algengur viðkomandi sjúkdómur er í viðkomandi þýði.

The positive predictive value is not intrinsic to the test and also depends upon the prevalence of the disease within the test population.

Question 30

8 mismunandi rannsóknaform í röð frá lægsta hierarchy of evidence upp í hæsta:

Answer 30

• Expert experience
• In-vitro research
• Animal research
• Case reports
• Case series
• Case control studies
• Cohort studies
• RCT

Question 31

Hver er munurinn á sensitivity og positive predictive value?

Answer 31

Sensitivity is the proportion of true positives correctly identified by the test.

The positive predictive value describes those patients with a positive test who also have the disease.

Question 32

Hver er munurinn á specificity og negative predictive value?

Answer 32

Specificity is the proportion of true negatives correctly identified by the test.

The negative predictive value describes those whose test results are negative who are disease free.

Question 33

Hvers konar error er selection bias?

Answer 33

Selection bias is a form of systematic error.

Question 34

Hvað er confounding factor?

Answer 34

Randomisation will reduce confounding.

A confounding factor is a background variable that is not of direct interest to the study.

Question 35

Hvenær eru random errors til staðar í mælingum og hvað veldur þeim?

Answer 35

Random error is always present in a measurement. They are caused by unpredictable fluctuations in the readings of a measurement apparatus or in the experimenters’ interpretation of the instrumental reading.

Question 36

Hvernig er experimental event rate reiknað?

Answer 36

T.d. í samanburðarhópi þar sem einn hópur fær samanburðarlyf og hinn nýtt:

(fjöldi þeirra sem FÁ sjúkdóm/einkenni þrátt fyrir að vera á nýja lyfinu) / (allir sem eru á nýja lyfinu)

Question 37

Hvernig er control event rate reiknað?

Answer 37

T.d. í samanburðarhópi þar sem einn hópur fær samanburðarlyf og hinn nýtt:

(fjöldi þeirra sem FÁ sjúkdóm/einkenni þrátt fyrir að vera á gamla lyfinu) / (allir sem eru á gamla lyfinu)

Question 38

Hvernig er relative risk reduction reiknað út?

Answer 38

(Control event rate - experimental revent rate) / Control event rate

Question 39

Hvernig er absolute risk reduction reiknað út?

Answer 39

Control event rate - experimental event rate

Question 40

Hvernig er number needed to treat reiknað út?

Answer 40

1 / absolute risk reduction

Question 41

Hvernig lítur Pareto diagram út og hvað sýnir það? Hver er tilgangur þess?

Answer 41

A Pareto diagram (also called a Pareto chart) is a type of chart that contains both a bar graph and a line graph. Each bar represents a category or set of qualitative data. The bars are arranged in order of frequency, so that more important categories are emphasized. The cumulative total is represented by the line graph.

The main purpose of a Pareto diagram is to highlight the most important data set among a set of different factors.

Question 42

Hvað verður um confidence interval þegar sample size verður stærra?

Answer 42

Confidence interval verður þá þrengra.

Question 43

Hvað er mode?

Answer 43

Tíðasta gildi

Question 44

Hvað er median?

Answer 44

Miðgildi

Question 45

Hvað er mean?

Answer 45

Meðaltal

Question 46

Studnet´s t test er statistical test fyrir parametric data. Hvað eru margir hópar í því?

Answer 46

2

Question 47

ANOVA er statistical test fyrir parametric data. Hvað eru margir hópar í því?

Answer 47

Fleiri en 2

Question 48

Fyrir hvernig data er chi squared test?

Answer 48

Fyrir qualitative data.

Question 49

Hvernig test er Kendall´s test?

Answer 49

Non-parametric significance test

Question 50

Hvernig lítur Funnel plot út og hver er tilgangur þess?

Answer 50

Funnel plots are a graphical means of checking for publication bias in meta-analyses and systematic reviews. The basic assumption is that trials showing no effect are less likely to be published if they are small than if they are large. Researchers conducting a large trial (which requires a lot of effort and funding) usually have an interest to publish the trial results even if negative. Furthermore, the running of a large trial is usually widely known in the field and difficult to hide. By contrast, it has been shown that small trials with a negative result more often remain unpublished, possibly because of a lack of interest or because the researcher (who may have links with the drug company funding the trial) may want to hide the negative result from the public. Evidence for publication bias can be assumed if smaller trials show more positive effects than larger trials (because small studies with a negative effects are more likely to remain unpublished).

The vertical axis of a funnel plot is a measurement of the precision of the estimated treatment effect (which is also an approximate measure of the size of the trial). The horizontal axis measures the treatment effect. The point estimate from each study is then plotted onto the graph and a vertical line inserted where the pooled estimate from the meta-analysis will lie. If the dots denoting the individual studies roughly form a symmetrical triangle, then publication bias is probably absent. If there is asymmetry, then publication bias can be suspected.