Lecture 31 & 32: Measures of Disease Frequency Flashcards
Why do we measure disease frequency?
To describe:
-Estimate burden of disease for prioritization
-Determine current status - establish “normal” to interpret changes over time
To compare:
-Between populations
-Overtime
-Between risk factor groups
-Evaluate preventative measures
How do we measure frequency?
-Counts (total # of what your measuring but by itself not sufficient) ex 132 not very helpful
-Proportions 132/232= 57% more useful
-Rates (important include time variable) 0.80 dogs per student-month
*note they are not synonymous (despite what you may read)
What are proportions?
-Number with disease divided by the population size ex A/(A+B) so numerator also included in denominator
-Often expressed as a percentage
-Values range from 0-1 (or 0-100)
What are rates?
-An expression of an event occurrence in a defined population in a specified period of time (measures speed)
-Not dimensionless (“PER UNIT SOMETHING”)
-ex 60km per hour, the unit changes meaning giving it value
-How fast the new cases of disease occur in population in a particular time frame
What is the difference between prevalence vs incidence?
Prevalence: the number of existing cases of disease in the population (At a certain point or period)
-SNAP SHOT in time can use proportion, no units, and ranges from 0-1 or 0-100
Incidence: the number of new cases of disease in the population (during a given time period)
How do you calculate prevalence?
2 different ways to calculate prevalence:
Point prevalence:
-Proportion of the population that has the disease of interest at a particular time (ex today)
= # with disease at single point in time/ # in pop that could have disease at that point in time
Period prevalence:
-Proportion of the pop that has the disease of interest during certain period (ex all of 2020)
-Period of time can be anything depending on what looking for
= # with disease during a specific time period/ # in pop that could have the disease during that period of time
What is censoring?
-Observation is stopped because:
-End of study
-Individual left the cohort
-Was lost to follow-up (reach out no reply)
-Died from a cause OTHER THAN THE ONE OF INTREST
What are important things to consider when talking about prevalence?
-Does NOT take into account WHEN the disease occurred therefore not a measure of risk
-Numerator includes individuals with disease that is of varying durations
-Cases may have developed yesterday, last week, last year etc
-But can still be an important and useful measure that tells us how many individuals in a population are affected (can be used to determine health care needs)
What is an example of a prevalent interpretation?
Ex prevalence of diabetes in 2015
national prevalence= 7.9%
-7.9% of Canadians had diabetes during the year 2015
What is incidence?
-A dynamic measure of disease occurrence
-New disease occurrences over time (needs a time component)
-Requires at least 2 observations or tests
1. to establish which individuals are disease-free
2. To identify which ones developed disease in the observation period
*have to be disease free from the start in order to measure when they get disease
What is an example of a incidence interpretation?
Diabetes in 2015 national incidence = 613.6 cases per 100,00 person-years
-If you followed 100,000 people for 1 year you would expect to see ~ 614 new cases of diabetes
What is the difference between, incidence risk vs rates?
2 ways to express incidence
-RISK
-RATE
-Numerator for both is the same (# new cases)
-Denominator (and interpretation) differs
What is RISK?
risk aka cumulative incidence
-Proportion of unaffected individuals who, on average will develop the disease of interest over a specified time period
-or the likelihood that an individual will contract the disease of interest during the given time period
-Report as proportions with no units, that ranges from 0-1 or 0-100
What is the equation for calculating risk?
risk= # new cases in pop during specified time period/ initial NAR - 1/2 withdrawals during the time period
NAR= number at risk at time 0
withdrawals= individuals that leave the study for other reasons, but did not develop the disease of interest
Subtract by 1/2 the number of withdrawals bc we assume that individuals leaving the pop did so on average 1/2 way through the time period
What is rate?
Rate aka incidence density
-Numerator is still new cases, but its a true rate
-Measures the rapidity / average speed with which newly diagnosed cases of the disease develop
-Denominator is the sum of units of time each individual was risk and observed
2 ways to calculate rates; denominator
1. Exact denominator (preferred method)
-use only if you know exact details for each individual
2. Approximate denominator
-Use if you only have summary data (ie dont know exact details of each individual)
