CM: Clinical Reasoning II

Question 1

What are the five steps of designing a study?

Answer 1

define the purpose of the study
formulate the null hypothesis
choose a study design
avoid sampling error: calculate sample size
avoid bias

Question 2

What are the three general purposes of a study?

Answer 2

descriptive: observational study
hypothesis-generating: observational study
hypothesis testing: experiment or observational

Question 3

What is the null hypothesis?

Answer 3

non-directional: will be no different than… - use 2 tailed test
directional: will be no higher/lower than… - one tailed statistical test, more likely to yield significant result, preferred if possible

Question 4

What is a simple experiment?

Answer 4

subjects randomly assigned to treatment or control and followed to see who becomes ill
ideally is double blind

Question 5

What is a repeated measures experiment?

Answer 5

same as simple, except two or more measurements in all groups, one before intervention and one after
each pt has his/her own control

Question 6

What is a repeated measures experiment with crossover?

Answer 6

after completing repeated measures study, experiment continued in same subjects except treatment and placebo groups are switched after time delay for residual treatment effects to be washed out
useful when extreme variation in outcome event

Question 7

What is a cohort study?

Answer 7

subjects selected based on whether they were exposed to risk factor and then followed to see who gets dz = follow-up study

Question 8

What is a case-control study?

Answer 8

subjects selected based on if they have dz and then studied to determine which had been exposed to risk factor or treatment

Question 9

What are type I and type II sampling errors?

Answer 9

type I = finding an association when variables not actually associated
type II = finding no association when variables are actually associated

Question 10

What is a selection bias?

Answer 10

distortion in estimate of association resulting from manner in which subjects were selected (ex: those exposed and those not exposed selected differently, longer survival rate or higher dropout rate in one group than other)

Question 11

What is an information bias?

Answer 11

distortion in estimate of association due to measurement error or misclassification of subjects

Question 12

What is confounding?

Answer 12

bias when causal effect of risk factor or treatment is inflated or deflated in set of data by presence of one or more extraneous variables colinked w risk factor and outcome variable

Question 13

What are incidence and incidence rate?

Answer 13

incidence = number of new events occurring in a defined population in a defined time period
rate = incidence divided by number of people at risk

Question 14

What is the prevalence and prevalence rate?

Answer 14

prevalence = number of active cases present in defined population in a defined time period
rate = prevalence divided by number of people at risk