Prevalence And Incidence Flashcards
Define Prevalence, how is it calculated and expressed?
A measure of how common a disease is
Number of people with condition / number of people in total
As a % or number/n people
What are the 3 types of prevalence?
Point prevalence - % in specific moment in time
Period - % over a period eg month
Lifetime - % that will suffer with that disease during lifetime
Define Incidence, how is it calculated and expressed?
Rate at which new events occur in a population, over a defined period of time
Number of new cases / number of people in total x years observed
As per n people per time period (eg 100 cases per 1000 people per year)
Or per n person-years (100 cases per 1000 person-years)
What factors affect prevalence?
Incidence rate
Recovery rate
Death rate
Transfer rate
Give examples of conditions with:
High prevalence + High incidence
High prevalence + Low incidence
Low prevalence + High incidence
Low prevalence + Low incidence
High prevalence + High incidence: common not brief condition (common cold)
High prevalence + Low incidence: uncommon, long term condition (Type 2 Diabetes)
Low prevalence + High incidence: common, brief condition (nose bleeds)
Low prevalence + Low incidence: uncommon, short term condition (pancreatic cancer)
What’s statistical inference?
Best guess based on data of the true value and describe the level of uncertainty
What’s point estimation?
Best guess based on sample data
What is sampling error and how can it be reduced?
Difference between sample point estimates and the true value
Reduce by testing a larger population
What’s standard error? How can it be reduced?
Numerical value representing the sampling error
Small SE = best guess closest to the truth and will get small value from a larger sample size
What are confidence intervals? How are they usually expressed?
Should present estimates from a sample with a range of plausible values to represent the level of uncertainty
The study finds 80% of mothers hold their baby on the left. A 95% CI is 75-85%
“You can be 95% confident that the true value lies between 75-85% of new mums holding their baby on the left”
What effect does a larger sample size have on:
Sample error
Standard error
Confidence intervals
Reduces both sample and standard error and provides narrower confidence intervals = estimate closer to the true value
How can confidence intervals be compared?
To see if there is a statistically significant (real) difference between two groups
If CI overlap = statistically significant
