Biostats, Ethics, Patient Safety, and QI

Question 1

How do you calculate the negative predictive value of a test (probability of not having disease in patient with negative test)?

Answer 1

True negatives/ All negative tests

remember to set up square with test on left and disease on top

Question 2

How do you calculate the sensitivity of a test (how accurate it identifies those who have disease)?

Answer 2

True positives/ (true positives + false negatives)

Question 3

How do you calculate the positive predictive value of a test (probability of disease in a patient with positive screen)?

Answer 3

true positive/ all positive tests

all positive= true positives + false positives

Question 4

A …. study is when similar subjects are divided into groups based on presence or absence of the outcome of interest and then comparisons are made regarding the frequency of risk factors in each group.

Answer 4

Case-control

Question 5

A …. study is when subjects are divided into groups based on the presence or absence of a presumed risk factor and are followed for development of outcome.

Answer 5

Cohort study

Question 6

A … study is when the presence of a risk factor and the presence of an outcome are reviewed simultaneously in a population.

Answer 6

Cross sectional

Question 7

How do you calculate specificity (how well the test excludes those that do not have the disease)?

Answer 7

True negatives/ (true negatives + false positives)

Question 8

A type …. error is defined as failing to reject the null hypothesis when it is really false.

Answer 8

2

concluding there is no difference when there actually is; exposed by investigating the power of the study

Question 9

A type … error is defined as rejecting the null hypothesis when it is actually true

Answer 9

1

concluding there is a difference in outcomes when there is not; exposed by investigating the p-value of results

Question 10

How do you calculate the number needed to treat (NNT, number of patients who need to be treated in order to prevent or cure a disease)?

Answer 10

NNT = 1/ ARR (absolute risk reduction)

NNT = 1/ (Proportion of bad outcomes for placebo - proportion of bad outcomes in treatment)

Question 11

…. bias is an error that is often introduced when data is subjectively gathered for a substantial time after the studied event occurred

Answer 11

Recall

individuals tend to remember details with skewed degree of accuracy

Question 12

… bias is when an intervention of the study uncovers a process at an earlier stage than is commonly experienced, which appears to artificially extend the time course of the event

Answer 12

Lead time

Question 13

… bias arises when the choice of participants for the study differentially includes/ excludes specific characteristics that directly influence the results of the study protocols

Answer 13

Selection

Question 14

… bias is present when the research instrument of the topic under study is not valid or reliable

Answer 14

Measurement

for example, the diagnostic instrument is not properly calibrated

Question 15

An odds ratio of > 1 means there is a …. chance the outcome has occurred due to the exposure, and an odds ratio of < 1 means the outcome is … with the exposure

Answer 15

higher; less likely