Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
31 Cards in this Set
- Front
- Back
|
What kind of study measures incidence?
|
Cohort
|
|
|
What kind of study measures prevalence?
|
Cross-sectional
|
|
|
What is the problem with a cross-sectional study?
|
The cannot determine causality because information is only obtained at a SINGLE POINT in time!
|
|
|
What is the false negative ratio?
|
1-sensitivity
|
|
|
What is the false positive ratio?
|
1-specificity
|
|
|
How does prevalence affect NPV and PPV?
|
The higher the prevalence, the higher the PPV.
The lower the prevalence, the lower the NPV |
|
|
What is the positive likelihood ratio?
|
sensitivity/1-specificity
|
|
|
What is the negative likelihood ratio?
|
specificity/1-sensitivity
|
|
|
What do likelihood ratios measure in general?
|
How much more or less likely a given test result is in diseased as opposed to nondiseased people. They show how much the odds (or probability) are increased/decreased if the test result is positive/negative
|
|
|
What is Posttest odds?
|
Pretest odds x LR
|
|
|
Cohort studies are also known as?
|
Longitudinal or incidence studies
|
|
|
What is the advantage of a case-control study?
|
They can be used to study rare diseases
|
|
|
What is the disadvantage of a case-control study?
|
The prevalence, incidence, and RR cannot be calculated because the numbers of subjects with and without the disease are determined by the investigator.
|
|
|
What are odds
|
Odds = probability of event/(1-probability of event)
Probability = odds/1+odds |
|
|
What is the attributable risk?
|
Incidence of disease in exposed - incidence in unexposed
|
|
|
What is the relative risk?
|
Incidence in exposed/incidence in unexposed
|
|
|
What is the odds ratio?
|
How much more likely it is that a person with a disease has been exposed to a risk factor than someone without the disease. It is used to estimate RR in case-control studies. The lower the disease incidence, the more closely it approximates the RR.
Odds ratio = odds that a diseased person is exposed/odds that a nondiseased person is exposed |
|
|
Selection bias
|
Samples/participants are selected that differ from other groups in additional determinants of outcome (patients with a fhx of breast cancer may self-select to enter mammography program so that prevalence seems higher than it is)
|
|
|
Measurement bias
|
Measurement or data gathering methods differ between groups (one group assessed by CT and another by MRI)
|
|
|
Confounding bias
|
A third variable is either positively or negatively associated with both the exposure and outcome variables
|
|
|
Recall bias
|
Difference between 2 groups in the retrospective recall of past factors or outcomes (cancer patients are more likely to remember exposures to chemical than healthy subjects)
|
|
|
Lead time bias
|
Earlier detection of disease gives an appearance of prolonged survival when the natural course of the disease is not actually altered
|
|
|
Length bias
|
Screening tests detect a disproportionate number of slowly progressive diseases but miss rapidly progressive ones, leading to an overestimation of the benefit of the screen.
|
|
|
What is a type 1 error (alpha)
|
Concluding that there is a difference in treatment effects between groups when there is actually not. FALSELY REJECTING the null hypothesis. (false positive conclusion)
|
|
|
What is a type 2 error (beta)
|
Concluding that there is no difference in treatment effect when there actually is. FALSELY ACCEPTING the null hypothesis (false negative conclusion)
|
|
|
When are results statistically significant?
|
p < 0.05
|
|
|
What is power?
|
The probability that the study will find a statistically significant difference when one is truly there. More subjects = more power
Power = 1 - type II error |
|
|
What does it mean if the confidence interval contains the null value (RR of 1.0 or odds ratio of 0%)?
|
The results are not statistically significant
|
|
|
What is reliability?
|
Equivalent to precision. How reproducible are the results? Random error reduces reliability and precision
|
|
|
What is validity?
|
Equivalent to accuracy. Measure the trueness of a measurement. Systemic error reduces validity and accuracy (miscalibration of equipment)
|
|
|
When do you use chi-squared, T-test, and ANOVA?
|
Chi-squared: Common percentages or proportions, nominal (non-numerical) or ordinal info
T-test: Compare 2 means (continuous info) ANOVA: 3 or more means |
