EBM Quiz 1 Flashcards
Evidence Based Medicine
The application of population based information to decision making about individual patients
(previously clinical epidemiology)
Case series study
Certain characteristics of a group (or series) of patients (or cases) are described in a published report; interesting or intriguing observations that occurred for a small number of patients, usually leads to a hypothesis and case control, cross sectional, or cohort study
Cross sectional study
Do an analysis on data collected on a group of subjects at one point in time rather than over a period of time
Could be a short period of time moving to the future
Cohort study
Also known as prospective studies
Moving forward in time, subjects are selected at the onset of the study and then determine whether they have the risk factor or have been exposed
(i.e. Framingham study of CV disease)
Historical cohort study
A cohort study by using information collected in the past and kept in records or files (retrospective) Direction of inquiry is still forward in time, from a possible cause or risk factor to an outcome.
Cross over study
A randomized trial except at a certain point in time, the subjects stop the treatment and placebo and then switch with the control group.
Unaware of which group they are in
Double blind study
Subjects are randomly put into study
They do not know which group they are in
Researchers do not know which group they are in
The gold standard (double blind randomized study)
Blind study
Subjects are randomly put into study
They do not know which group they are in
Randomized clinical trial
The gold standard
greatest justification for concluding causality
subject to the least number of problems or biases
Patients are randomly assigned different treatments
one is treatment, the other is control
Control may be placebo or commonly used treatment
they are expensive and long
Controlled vs uncontrolled studies
Can use another investigators research as control
Can use historical controls (subjects the investigator has previously treated)
Uncontrolled studies are more often used with procedures
Primary sources
Published in journals
Most popular, critically acclaimed by peer reviewers before being published
Dissertations
for an academic degree, critically reviewed before a degree is given
Technical reports
Conference proceedings
Surveys
Patents
Secondary sources
Review journals
Text books
Summary of research
Most newspaper articles
Tertiary
Summarizes and condenses information from primary and secondary sources
Dictionaries
Encyclopedias
Almanacs
Impact factor
The measure of frequency an article has been cited in the past year
Quantitative
Consists of numbers that represent counts or measurements
Qualitative
(Categorical)
can be separated into non numeric categories
Nominal
Categories only
Data cannot be arranged in an ordering scheme
i.e. gender or occupation
Ordinal
Categories are ordered, differences can’t be found or are meaningless
(i.e. health categories - poor, reasonable, good, excellent)
Discrete random variable
Take on countable values
i.e. number of offspring or respirators. Or shoe size
Continues random variables
The amount of time it takes to accomplish a task
(i.e. has infinite number of times to accomplish this task)
(hours-mins-secs-microsec-nanosec-etc)
Mean
Average
5.4
1.1
0.42
.73
.48
1.1
add all together and divide by 6
= 1.538
Problem is one outlying variable can affect mean
Median
Middle number
arrange in descending order
take middle number
if two middle numbers, add together divide by 2
Mode
mode is the data set that occurs most frequently
if two values occur with the same frequency it is bimodal
if greater than two values occur with the same frequency it is multi modal
If no value is repeated there is no mode
5.4, 1.1, .42, .73, .48, 1.1 = mode is 1.1
27, 27, 27, 55, 55, 55, 88, 88, 99 = bimodal 27 & 55
1,2,3,4,5,6,7,8,9 = no mode
1,2,3,1,2,3,4,5,6,7, = multi modal 1, 2, & 3
Weighted mean
values may vary in degree of importance so we use a weighted mean to adjust this
Find the mean of 3 test scores
Scores are 85, 90, 75
Weights are 20%, 30%, 50%
20+30+50
=
8150
———-
100
= 81.5 %
x̅ = ∑(w*x) / ∑w
Standard deviation
the measurement of of variation of values around the mean
(how much variability from mean)
S= Square root of
/ ̅ ̅ ̅ ̅ ̅ ̅ ̅ ̅ ̅ ̅ ̅ ̅ / ∑ (x-x̅)² s = / ----------- √ n-1
Find the deviation of each observation from the mean, square the deviations, add them and then divide the sum by n - 1.
The square root is used to express the standard deviation on the original scale of measurement
Variance
Is a measure of variation
This is equal to the square of the standard deviation
(either sample or population)
Variance examines how a set of data values vary among themselves.
Large variance means great variation within the data
Significant and insignificant value
Significantly low (µ - 2σ) or Lower Significantly high (µ + 2σ) or Higher
Values not significant: between (µ - σ) and (µ + σ)
Significance is not used unless it has been determined to be statistically significant
Significance example
Mean (+) or (-) 2 times the standard deviation
Significant
If standard deviation is 1.0 and mean is 6.
2 x 1 = 2
2 (+) or (-) 6 = 4 - 8
this means
4 or lower is significant
or
8 or higher is significant
Not significant values are between 4 and 8
σ =
Standard deviation in a population
µ =
Mean of all values in a population
s =
standard deviation in a sample
σ² =
Variance
standard deviation squared
N =
number of values in a population
n =
number of values in a sample
∑ =
The addition of a set of values
Greek letters are used for
Population
English letters are used for
Sample
x =
Used to represent on value in a data set.
x̅ =
The mean of a set of sample values