EBM 1 - HEALTH AND DISEASE in the population

Question 1

types of data

Answer 1

categorical;

• ordered; examination grades
• non ordered; fav colours

numerical;

• discrete ; no of goals scored
• continous ; height

Question 2

how to display data i.e. what charts

Answer 2

continous ; histograms

discrete; bar chart, pie charts etc

Question 3

binary variable

aka?

Answer 3

basically a variable where there are only two options
eg do you have the disease? YES or NO

DICHOTOMOUS

Question 4

FORMULAS;

• proportion
• risk
• prevalence
• incidenCE RATE
• HOW can death rate be calculated
• use incidence formula to come up with another formula for prevalence & when only can we use this formula

Answer 4

• prop; no. of people with disease over;
  total people altogether
• risk; no. of NEW cases of disease over;
  total people initially disease free
• prevalence; no. of people with the disease ATM over;
  no. of total ppl atm

risk; NEW cases
prevalence ; EXISTING cases

• incidence rate ; how fast disease is occurring
  no of NEW cases of disease over;
  no. of people initially disease free X TIME INTERVAL
• death rate can be calculated in this way (i.e. incidence rate) also but the event is death

prevalence; incidence X average duration of disease
check if that includes time interval etc ?
this is only true when incidence is constant for the duration of a disease

Question 5

measures of central tendency

Answer 5

mode ; mot common
median ; middle
mean ; average

Question 6

measures of variability or spread

Answer 6

range
interquartile range
standard deviation

Question 7

normal distribution

Answer 7

uses reference range
95% of values will lie between;
mean ± 1.96 SD

Question 8

POSSIBLE SHAPES of distribution

Answer 8

symetrical
positively skewed- skewed to the right - down the stairs
negatively skewed - skewed to the left- up the stairs

(opposite to slope: that’s how to remember it)

Question 9

2) types of studies

1) 2 main CLASSIFICATION of studies

Answer 9

types ; RCT, cohort, case control, cross sectional ecological

note ; these are in order of strongest to weakest evidence of causality

classified ; observational; no intervening e.g. observing teenagers wifi usage
OR
experimental/ INTERVENTIONAL
YOU intervene change something about treatment or exposure to the disease
eg RCT

Question 10

cross sectional study

• what is it?
• adv & disadv?
• how would you set one up

Answer 10

what? measures PREVALENCE of disease

adv; good way of measuring prevalence of health conditions
can generate hypothesis

disadv; reverse causality ie as disease & outcome are measured at the same time it can be hard to know which caused which i.e. x cause y or y cause x

Question 11

selection bias

generalisability

Answer 11

SB; bias in selection of participants meaning the sample doesn’t represent the target population

G; to what extent can we relate this to other populations

Question 12

accuracy vs precision

Answer 12

accuracy ; how representative the study is of the population (whole or target?)
- will be affected by selection bias

precision; how much variability there is between the sample statistics
measure using CI

THUS studies can be inaccurate but precise & visa versa

Question 13

• ecological study
• what?
• good and bad?

Answer 13

studies where the unit of analysis is populations not individuals

adv ; good for generating hypothesis to use in individual - studies

• quick and cheap
• good for measuring group exposure

disadv ; can not always relate to individuals
eg smokers have a higher mortality rate but that doesn’t mean each and every individual that smokes does also etc
- often heavily CONFOUNDED (CH 4 RETURN TO)

Question 14

if mean is bigger than the median, what does this to the graph

Answer 14

positively skewed ie skewed to the right , DOWN the stairs

visa versa

Question 15

what has the highest prevalence ?

Answer 15

the one that has the highest incidence rate &; the highest survival rate which is equivalent to the average duration of disease