EVIDENCE BASED MEDICINE

Question 1

What are the steps in Evidence Based Psychiatry?

Answer 1

5 A’s

Ask - a relavent question about the patient

Access - database of available research

Appraise - look through research and appraise articles

Apply - apply what you’ve learned to your patient

Assess - evaluate your performance

Question 2

How do you formulate a clinical question?

Answer 2

PICO

Patient/problem

Intervention

Comparison

Outcome

Question 3

When would you use a case report/case series?

Answer 3

For some rare disease, that you want to do cheaply and quickly. But you cannot draw meaningful conclusions about cause and effect.

Question 4

When would you use a Cross Sectional Study?

Answer 4

To determine prevelance. You can demonstrate cause and effect also.

Unfortunately you can get something called Prevelance-Incidence Bias - a risk factor appears to have caused a disease when it could just be something that alters duration or prognosis.

Question 5

When do you use Cohort Studies?

Answer 5

Take 2 groups (defined by whether or not they were exposed to a risk factor). Compare who develops a disease with who doesn’t in both groups to give you an idea of whether exposure results in development of the disease.

PROSPECTIVE

Question 6

What is Central Tendency and how is it measured?

Answer 6

• It is a description of a data set using a single value that represents the centre of a data set.*
• Mean, Median, Mode*

Question 7

Mean?

Answer 7

Sum of all measurements divided by the number of elements.

Best for normally distributed data.

Painfully skewed by Outliers.

Question 8

Median?

Answer 8

Middle value in a data set. Not affected by outliers (therefore can be used with NON NORMAL or skewed distributions). Okay when dealing with ordinal data.

Question 9

Mode?

Answer 9

Most often occuring data.

Question 10

What is Dispersion and how is it measured?

Answer 10

Dispersion is the spread about the mean. Measured by :

• SD (which is the square root of variance)
• Standard Error of the Mean
• Quartiles
• Percentile
• Range

Question 11

Skewness and Kurtosis

Answer 11

Skewness - a measure of assymetry

Kurtosis - refers to the degree of the presence of outliers (spread wide or thin tailed).

Question 12

Variance

Answer 12

Statistical measure of variations

Question 13

Standard Deviation

Answer 13

Measure of spread about the mean.

Question 15

Parametric vs Non Parametric

Answer 15

Parametric - makes assumptions about the distribution of the population from which the data was taken. The assumption is usually that the data is NORMALLY DISTRIBUTED.

Non parametric - does not make assumptions about the distribution. Better used to non normally distributed data or when making the assumption of normallity is worrisome (small sample size).

Question 16

And yet….

Answer 16

Non Parametric tests are LESS POWERFUL (so you need a bigger study size).

But Parametric tests can make the WRONG CONCLUSIONS about associations.

Question 17

How many phases of DRUG TRIALS?

Answer 17

4

Question 19

Phase 1 Drug Trials

Answer 19

safety and pharamacokinetics

small group, assess tolerability

Question 20

Phase 2 drug trials

Answer 20

test effective dose, dosing schedule, often by COMPARING IT TO ANOTHER drug

Question 21

Phase 3

Answer 21

Double Blind RCT to further test efficacy and safety

Question 22

Phase 4

Answer 22

post marketing

helps you find out less common SE and test the drug in specific populations