Biostadistics Flashcards
Intention to treat analysis
subject results are analyzed according to the group that they were initially assigned to (not according to adherence) is one technique used to analyze outcome data and preserve randomization
When the cut-off point of a test is decreased, what happen with specificity and sensitivity?
Decrease in specificity (TN/(TN+FP)) and an increase in sensitivity (TP/(TP+FN)).
To identify more patients (usually in the form of decreasing the cutoff value) but with more false positives occurring.
p-value of 0.052 meaning in a study that compares two medications?
There is a 5.2% chance that A is more effective than B is due to chance.
If the cutoff value of a diagnostic test is increased (meaning it takes more of a finding to suggest a diagnosis), what should we say about specificity and sensitivity?
The sensitivity decreases while the specificity increases.
If the confidence interval of a relative risk contains the value 1
The result is likely not epidemiology significant.
Berkson bias
occurs when hospitalized study subjects are more likely to have a greater burden of illness than other possible subjects.
Attrition bias
Occurs because patients who are lost to follow-up may be different from those who remain in the study.
Selection bias
A sampled population is not representative of the population researchers are trying to study.
Due to inappropriate recruitment or attrition of study participants. E
- non-response bias (participants who answer a survey may be less sick than participants who don’t)
- the healthy worker effect (employed subjects may be healthier than others)
- volunteer bias (volunteers may be different from those who do not volunteer)
- late-look bias (patients with severe disease may be less likely to be studied due to death or disability)
-Attrition bias (patients who are lost to follow-up may be different from those who remain in the study).
Response bias
Occurs when the outcome metric is a subjective patient-reported measure because patients may change their responses in a non-random manner.
The factors that increase prevalence
Increase in new cases (increased incidence)
An improvement in the quality of care (prolonged duration of disease)
Improved diagnostic ability (early detection thus more cases).
Normal distribution, mean, median and mode.
In a normal distribution, the mean, median, and mode are identical
Hawthorne effect.
participants change their behavior when they are aware that they are being studied.
The Pygmalion effect
an investigator inadvertently conveys his desired result to the participant, who then alters his behavior accordingly.
External validity
is the ability to use results from a study to draw conclusions about populations different from those used in the study.
Internal validity
refers to the degree to which a study’s results are accurate and can be used to establish a cause-and-effect relationship. A randomized and controlled trial has the highest level of internal validity.
A latency period
The negative effects of a disease or the positive benefits of a treatment take a long time to become clinically apparent.
Lead time bias
occurs when the early detection of a disease falsely elevates the survival time of a disease.
T-test
Compare 2 groups in a clinical trial. This test is used to determine whether a significant difference exists between 2 means. Therefore, it can only be used for continuous variables
Chi-squared
analyze categorical (not continuous) variables
Effect modification
occurs when an external variable positively or negatively modifies the effect of a risk factor on a certain disease, effect modification changes the magnitude or direction of an effect.
Confounding
occurs when a third factor is associated with both the exposure and the outcome of interest. For example, if smoking is associated with chewing gum, it could seem that chewing gum is associated with lung cancer even though smoking is the confounding variable.
The hazard rate
The probability of an event occurring in the next time interval divided by the length of that interval.
If the ratio is less than 1, then the treatment reduces risk, and if it is greater than 1, then it represents an increased risk.
Measure in Case control studies
Odds ratio
Measure in Cohort studies
Relative risk
