Term
|
Definition
| The likely hood that the null hypothesis is true. |
|
|
Term
|
Definition
| A false positive. Or you deny the null hypothesis when it is true. Or you find a significant difference when there isn't one. |
|
|
Term
|
Definition
| A false negative. Accepting the null hypothesis when it is false. Or not finding a significant difference when there really is one. |
|
|
Term
|
Definition
| The probibility the test will reject the the null hypothesis when it is false. So probability you will not make Type II error. The power is usually .8. Power increases with increased participants or events. Themore participants you have the more likely you are to see a difference if there is one is how I look at it. |
|
|
Term
|
Definition
Probability of an event occurring in the exposed group vs. a non-exposed (or control) group. A relative risk of 1 indicates no difference between comparison groups. A
RR < 1 means that the intervention reduced the risk of the outcome. |
|
|
Term
Relative Risk calculation
if disease progression occurred in 28% of placebo-treated
patients and in 16% of metoprolol-treated patients what is the relative risk |
|
Definition
| Relative risk is 16/28 or 0.57. So patients treated with metoprolol were 57% as likely to have disease progression compared to placebo. |
|
|
Term
Relative Risk Reducation calculation
if disease progression occurred in 28% of placebo-treated
patients and in 16% of metoprolol-treated patients, what is the relative risk reduction |
|
Definition
Relative risk reduction = 1- relative risk
Relative risk = 16/28 = 0.58
1-0.58 = 0.42
meaning there is a 42% risk reduction in disease progression in patients being treated with metoprolol
Not the best for measuring clinical significance. For example if the disease progession happened to 4% of patients in placebo and 2% in metoprolol group the relative risk reduction would be 50% but is this just 2% difference clinically significant. For clinically significant absolute risk reduction is best |
|
|
Term
| Absolute Risk Reduction definition |
|
Definition
| absolute difference in outcome rates between the treatment and control groups |
|
|
Term
Absolute Risk Reduction calculation
if disease progression occurred in 28% of placebo-treated
patients and in 16% of metoprolol-treated patients what is the absolute risk reduction? |
|
Definition
| 28-16% = 12% absolute risk reduction |
|
|
Term
| Number needed to treat definition |
|
Definition
represents the number of people who would need
to be treated with the intervention in order to prevent 1 adverse event |
|
|
Term
number needed to treat (NNT) calculation
if disease progression occurred in 28% of placebo-treated
patients and in 16% of metoprolol-treated patients. What is the number needed to treat? |
|
Definition
NNT = 1/absolute risk reduction
Absolute risk reduction = 28-16% = 12%
1/0.12 = 8.3
NNT = 9 patients (round up) |
|
|
Term
|
Definition
| ratio of the odds of an event occurring in the treatment group to the odds of an event occurring in the control group |
|
|
Term
Odds ratio calculation
In a group of 100 smokers, 40 people developed
lung CA while 60 people do not. In a similar group of 100 non-smokers, lung CA developed in 10 people. What is the odds ratio? |
|
Definition
The odds of a smoker developing lung CA would be 40/60 = 0.66 (number of people that had the event over number of people that did not), whereas the odds of a non-smoker developing lung CA would be 10/90 = 0.11
The odds ratio would then be 0.66/0.11 = 6
If the odds ratio is close to 1 then there is not much different between the 2 groups
|
|
|
Term
|
Definition
| Discrete data has a limited number of values usually within categories or classifications. Descrete data can be ordinal or nominal data. |
|
|
Term
|
Definition
Subtype of discrete data, assigned numbers have meaningful order but not necessarily an equal distance between them.
Ex. No pain, mild pain, moderate pain, severe pain
|
|
|
Term
|
Definition
Subtype of discrete data. Categories have no rank or order.
Ex. male/female, Yes/no
|
|
|
Term
| Continuous data definition |
|
Definition
| Values are within a defined range. Equal distance between values. Broken down into ratio and interval data |
|
|
Term
|
Definition
Subtype of continuous data. Has an absolute zero.
Ex. Height, age, blood pressure |
|
|
Term
|
Definition
It has no absolute zero.
Ex. Temperature in Fahrenheit or celcius (Kelvin has absolute zero) |
|
|
