Miller's Stats Review

Question 1

Observational study designs

Answer 1

Retrospective (case control or case series), single (cross-sectional survey or case report) or prospective (cohort)

Question 2

Experimental study designs

Answer 2

Prospective RCT

Question 3

Only study type you can use to calculate incidence of disease

Answer 3

Cohort study

Question 4

Difference between systematic review and meta-analysis

Answer 4

Meta-analysis is a systematic review with addition of combining data and forest plots

Question 5

Phases of FDA research

Answer 5

1: pharmacology, 2: initial efficacy, 3: confirm efficacy, 4: post-market analysis

Question 6

5 levels of evidence

Answer 6

1: RCT, 2: Cohort, 3: Case-control, 4: Case series, 5: Expert opinion

Question 7

What makes a randomized controlled trial a level 2 study

Answer 7

Heterogeneous results, follow-up less than 80%, poor randomization

Question 8

What determines the level of evidence in a systematic review

Answer 8

The lowest quality study used in the review

Question 9

Bias vs confounding

Answer 9

Bias = systematic error in methodology, confounding = something that wasn’t accounted for that may influence results

Question 10

Types of bias before the study

Answer 10

Selection, channeling, chronology

Question 11

Types of bias during the study

Answer 11

Detection, recall, interviewer/assessor, performance and Hawthorne affect

Question 12

Selection bias

Answer 12

Patient factors are not equal between groups, outpatient total knees have lower infection than inpatient because they are healthier to begin

Question 13

Detection bias

Answer 13

Inconsistency in outcome assessment, tendency to look more closely for an outcome in one group vs the other

Question 14

Recall bias

Answer 14

Knowledge of disease alters the patients recall of specific prior events

Question 15

Assessor bias

Answer 15

Person doing the assessment is not blinded

Question 16

Performance bias

Answer 16

Unintended difference in quality of care provided to two different groups in the study

Question 17

Hawthorne effect

Answer 17

Subjects alter behavior when they know they’re being observed

Question 18

What study is used to determine bias in meta-analysis

Answer 18

Funnel plot

Question 19

Incidence vs. prevalence

Answer 19

Incidence = new cases over period of time, Prevalence = total number of current cases

Question 20

Relative risk vs. odds ratio

Answer 20

RR can only be calculated in cohort study = incidence in exposed/incidence in unexposed.

Question 21

How to calculate prevalence

Answer 21

Cross-sectional study

Question 22

Categorical data

Answer 22

Nominal and ordinal, chi square for larger numbers, Fisher exact if smaller numbers

Question 23

Continuous data

Answer 23

Intervals and ratios

Question 24

Abnormally distributed continuous data analysis

Answer 24

Mann-Whitney U-test

Question 25

Normally distributed data analysis

Answer 25

T-test if there are only two groups, ANOVA if 3+ groups, Pearson correlation coefficient and regression analysis

Question 26

Sensitivity

Answer 26

Ability to detect disease (True Positives). TP/TP+FN

Question 27

Specificity

Answer 27

Ability to detect healthy (True Negatives). TN/TN+FP

Question 28

PPV

Answer 28

Probability they have disease given a positive test = TP/TP+FP

Question 29

NPV

Answer 29

Probability they are healthy given negative test = TN/TN+FN

Question 30

Type I error

Answer 30

Alpha error = false positive result, minimize risk so this happens < 5% of the time, hence p < 0.05

Question 31

Type II error

Answer 31

Beta error = false negative; minimize risk so this happens <20% of the time, hence 1 – B = 1-20% = 80% power

Question 32

Confidence interval distributions

Answer 32

95% = 1.96 distribution

Question 33

Variance

Answer 33

Distribution of your bell curve, large sample size with narrow your variance