Miller's Stats Review Flashcards

Ortho Stats Review

1
Q

Observational study designs

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Experimental study designs

A

Prospective RCT

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Only study type you can use to calculate incidence of disease

A

Cohort study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Difference between systematic review and meta-analysis

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Phases of FDA research

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

5 levels of evidence

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What makes a randomized controlled trial a level 2 study

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What determines the level of evidence in a systematic review

A

The lowest quality study used in the review

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Bias vs confounding

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Types of bias before the study

A

Selection, channeling, chronology

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Types of bias during the study

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Selection bias

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Detection bias

A

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

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Recall bias

A

Knowledge of disease alters the patients recall of specific prior events

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Assessor bias

A

Person doing the assessment is not blinded

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Performance bias

A

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

17
Q

Hawthorne effect

A

Subjects alter behavior when they know they’re being observed

18
Q

What study is used to determine bias in meta-analysis

A

Funnel plot

19
Q

Incidence vs. prevalence

A

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

20
Q

Relative risk vs. odds ratio

A

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

21
Q

How to calculate prevalence

A

Cross-sectional study

22
Q

Categorical data

A

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

23
Q

Continuous data

A

Intervals and ratios

24
Q

Abnormally distributed continuous data analysis

A

Mann-Whitney U-test

25
Q

Normally distributed data analysis

A

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

26
Q

Sensitivity

A

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

27
Q

Specificity

A

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

28
Q

PPV

A

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

29
Q

NPV

A

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

30
Q

Type I error

A

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

31
Q

Type II error

A

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

32
Q

Confidence interval distributions

A

95% = 1.96 distribution

33
Q

Variance

A

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