EBM Flashcards Preview

► Med Misc 09 > EBM > Flashcards

Flashcards in EBM Deck (45):
1

Give an example of a patient oriented outcome.

Quality of life stuff

2

External validity means

How you can apply the study in the real world. For example, a surgery study has no external validity for a psychiatry practice

3

What is internal validity?

Was the study you conducted valid?

4

Odds ratio limits

2x2 table
Odds ratios are the cross products
OR>1 means increased risk of said outcome, good or bad

5

Can you calculate an ARR from an odds ratio?

No

6

Can likelihood ratios be applied to different populations of patients?

YES because a LR takes into account all the data in the 2x2 table.

7

Construct a 2x2 table

Sick Not sick
Test + a b
Test - c d

8

An easy way to calculate a negative LR:

1/LR = -LR

9

sPIN means:

SpPIn - in a test with a high Specificity, a Positive test rules In the diagnosis

10

sNOUT

SnNOut - in a test with a high Sensitivity, a Negative test rules Out the diagnosis

11

+LR =

Sensitivity/(1-Specificity)

12

-LR =

(1-sensitivity)/Specificity

13

The equation for sensitivity based on the 2x2 table

Sick Not sick
Test + a b
Test - c d

Sensitivity = a/(a+c) = true positives/all disease positives

14

The equation for specificity based on the 2x2 table

Sick Not sick
Test + a b
Test - c d

Specificity = d/(b+d) = true negatives/all disease negatives

15

What are LR cutoffs to remember?

+LR > 5 is good
+LR .2 not good
-LR

16

List and identify the important validity concepts for articles about therapy:

1. randomization
2. allocation concealment
3. blinding
4. intervention
5. outcome assessment
6. withdrawals and followup
7. similarity of comparison groups and statistical adjustment as needed
8. power analysis
9. intention to treat analysis

17

T/F: A trial comparing a new anti-hypertensive medication to an older anti-hypertensive medication is NOT a valid therapy study because it does not include a placebo.

False

18

Examine the following abstract, paying particular attention to the outcomes studied.

Glucose Control and Vascular Complications in Veterans with Type 2 Diabetes
Duckworth W, et al.
ABSTRACT
Background The effects of intensive glucose control on cardiovascular events in patients with long-standing type 2 diabetes mellitus remain uncertain.
Methods We randomly assigned 1791 military veterans (mean age, 60.4 years) who had a suboptimal response to therapy for type 2 diabetes to receive either intensive or standard glucose control. Other cardiovascular risk factors were treated uniformly. The mean number of years since the diagnosis of diabetes was 11.5, and 40% of the patients had already had a cardiovascular event. The goal in the intensive-therapy group was an absolute reduction of 1.5 percentage points in the glycated hemoglobin level, as compared with the standard-therapy group. The primary outcome was the time from randomization to the first occurrence of a major cardiovascular event, a composite of myocardial infarction, stroke, death from cardiovascular causes, congestive heart failure, surgery for vascular disease, inoperable coronary disease, and amputation for ischemic gangrene.
Results The median follow-up was 5.6 years. Median glycated hemoglobin levels were 8.4% in the standard-therapy group and 6.9% in the intensive-therapy group. The primary outcome occurred in 264 patients in the standard-therapy group and 235 patients in the intensive-therapy group (hazard ratio in the intensive-therapy group, 0.88; 95% confidence interval [CI], 0.74 to 1.05; P=0.14). There was no significant difference between the two groups in any component of the primary outcome or in the rate of death from any cause (hazard ratio, 1.07; 95% CI, 0.81 to 1.42; P=0.62). No differences between the two groups were observed for microvascular complications. The rates of adverse events, predominantly hypoglycemia, were 17.6% in the standard-therapy group and 24.1% in the intensive-therapy group.
Conclusions Intensive glucose control in patients with poorly controlled type 2 diabetes had no significant effect on the rates of major cardiovascular events, death, or microvascular complications. (ClinicalTrials.gov number, NCT00032487 [ClinicalTrials.gov] .)
Volume 360:129-139 January 8, 2009 Number 2

The outcome "median glycated hemoglobin level" is:
A. DOE - A disease-oriented outcome
B. POE - A patient-oriented outcome

DOE - A disease-oriented outcome

19

Randomization is:

the best way to "allocate" subjects in your study to the comparison groups. There are right and wrong ways to conduct randomization, but in general a random number table or computer-generated randomization are the best ways.

20

What is allocation concealment?

Allocation concealment is not intervention "blinding". Instead, point is to ensure that the person recruiting for the study (the person who invites people into the study, applies the inclusion and exclusion criteria and consents them for the study) does not know the group into which the subject will be placed.

21

Examples of continuous data in a study:

For continuous data, you'll see means with their standard deviations, and you'll see medians with interquartile ranges (similar to standard deviations)

22

Examples of categorical data in a study:

For categorical data, you'll may see confidence intervals around the percentages

23

Number Needed to Treat Means:

Number Needed to Treat is a way to think about the effectiveness of a therapy as a clinician with a panel of patients.

An example:
CER = 20%, EER - 10% (the outcome is a bad outcome)
then ARD = 10% or 0.10 - there is a 10% reduction in the bad outcome for the intervention group.
therefore, the NNT = 1/0.10 = 10. We'd need to treat 10 people before we prevent an additional bad outcome.

Do not use means to calculate NNT. Use ratios.

24

NNT =

1/ARR or 1/ARD (ARR and ARD are the same thing)

the absolute risk reduction, risk difference or excess risk is the change in risk of a given activity or treatment in relation to a control activity or treatment

25

Control Event Rate (CER) =

comparative statistics

# events in control group / # subjects in control group

26

Experimental Event Rate (EER) =

comparative statistics

# events in intervention group / # subjects in intervention group

27

Relative Risk (RR) =

comparative statistics

EER/CER

# events in intervention group / # subjects in intervention group DIVIDED BY # events in control group / # subjects in control group

28

Relative Risk Reduction (RRR) =

comparative statistics

1 - RR = 1 - (EER/CER) = 1 - (# events in intervention group / # subjects in intervention group DIVIDED BY # events in control group / # subjects in control group)

29

Absolute Risk Difference (ARD, same as ARR) =

comparative statistics

CER - EER = (# events in control group / # subjects in control group) - (# events in intervention group / # subjects in intervention group)

the absolute risk reduction, risk difference or excess risk is the change in risk of a given activity or treatment in relation to a control activity or treatment

30

Your question:
In patients with a lower extremity deep vein thrombosis, does low molecular weight heparin (LMWH) vs. adjusted dose unfractionated heparin (UH) lead to less disease progression (especially pulmonary embolism) and fewer complications (especially bleeding)?

The study:
A well designed and executed randomized controlled trial of LMWH vs. UH reveals that the cost of LMWH is much lower than using UH primarily because of decreased hospital and laboratory monitoring costs.

When you are summarizing the critical appraisal of this study, your major criticism will be:

A. Biased - The validity of this study is too open to bias.
B. No answer - This study is valid, but does not answer the question I asked.

No answer - This study is valid, but does not answer the question I asked.

It has the wrong outcomes.

31

List and identify the important validity concepts for systematic reviews:

1. comprehensive search for evidence
2. publication bias
3. explicit inclusion/exclusion criteria
4. explicit assessment of the quality of included studies
5. assessment of heterogeneity
--qualitative
--quantitative

32

Systematic reviews are meant to:

locate and summarize ALL existing valid evidence on a topic.

33

The validity of a systematic review has to do with how well the REVIEW was done first, not how well the individual trials were done. While the validity of the included studies is important, the first step is to appraise the REVIEW. This includes:

Was the SEARCH for evidence comprehensive?
Did it include at least 2 different databases? Was there a search for unpublished literature? Is is likely that relevant studies were missed?

Were there explicit INCLUSION and EXCLUSION criteria stated and followed? Was there a table of excluded studies so that you can tell which studies didn't make the cut?

Were the included studies appraised for their validity (here's where the individual trials are graded)? Did the authors analyze the results with and without the low-quality studies (did they perform a sensitivity analysis)?

Was there an assessment for heterogeneity?
heterogeneity refers to differences between the studies that make it unwise to combine their results - these differences can be from patient population, methods, etc.
there are statistical measures of heterogeneity (p-values, I-squared statistics), etc. The most important issue, however, is that the authors didn't just look for it, but EXPLAINED the heterogeneity if they found it.

Did the authors assess for publication bias? Was a funnel plot provided?

if studies don't show a difference between treatment and control, then they frequently don't get published, so they cannot be found in a regular search. A funnel plot can help show this graphically, but the authors should discuss whether publication bias was present or not.

34

Which of the following sets are important validity criteria for systematic reviews?

A - Randomization, blinding, allocation concealment and follow-up rates.
B - Comprehensive search, explicit inclusion criteria, validity assessment of included studies and assessment for heterogeneity.

B - Comprehensive search, explicit inclusion criteria, validity assessment of included studies and assessment for heterogeneity.

A. has to do with therapy studies and RCTs

35

T/F: Meta-analysis should be reported with every systematic review, because it can fix the problem of too much heterogeneity between studies.

False
meta-analysis can FIND heterogeneity, but not fix it.

36

T/F: meta-analysis can be performed on both continuous and dichotomous data,

True

37

Systematic reviews have the potential to be the highest level of evidence about a given clinical topic, but only if:

they are performed well (the validity issues mentioned above)

they include high quality evidence themselves.

38

List and identify important validity concepts for studies about diagnostic tests:

1. appropriate reference standard for comparison
2. independent comparison of test and reference standard
3. blind comparison of test and reference standard
4. consecutive or random enrollment of subjects

39

Sensitivity and specificity can be determined from which of the following study designs:
Coh/CC - Cohort or case-control studies
RCT - Randomized controlled trial

Cohort studies are the best type of studies to use for diagnostic tests, case-control studies work well also.

40

What is a reference standard?

recognized reference standard - this test should be a non-controversial, definitive test used to diagnose the condition - e.g., a pathologic diagnosis, a confirmatory x-ray, etc.

41

In a study of the physical diagnosis of splenomegaly by percussion over Traube's space, the researcher performed the percussion, then did the ultrasound himself to see if he was correct. This violates which validity criterion?

A. Recognized reference standard - ultrasound does not diagnose splenomegaly
B. Blinding - the person doing the ultrasound knew the results of the percussion test

B. Blinding. The ultrasonongrapher should not know the results of the percussion test and vice-versa.

42

Positive Predictive Value (PPV) =

Sick Not sick
Test + a b
Test - c d

PPV = a/(a+b) = true positives / all test positives

43

Negative Predictive Value (NPV) =

Sick Not sick
Test + a b
Test - c d

NPV = d/(c+d) = true negatives / all test negatives

44

pre-test and post-test probabilities are:

the probabilities of disease before and after the test you are considering. Obviously, for a test to be useful, it will change the probability of disease - making it more or less likely, depending on the result. Frequently, we can use the overall prevalence of a disease in a population as the baseline pre-test probability.

45

Prevalence =

Sick Not sick
Test + a b
Test - c d

Prevalence = (a+c) / (a+b+c+d) = all disease positives / all