NOTES FOR POPHLTH MOD 1 Flashcards
Epidemiology
Frequency of health and disease in different populations- causes of poor health and how to improve
Epidemiology equation
E= (N/D) / T
GATE standards for what?
Graphic approach to epidemiology
Cohort study “observational follow up”
CG and EG have measurements/observations
Outcomes over a period of time- can measure incidence and prevalence
Incidence (OUTcome measurement)
Events that occur are counted forward from starting point OVER PERIOD OF TIME
Prevalence (OUTcome measurement)
Disease are counted at ONE POINT IN TIME. time is not included
Describe incidence overview
easy to measure
categorical measure eg” death’ gender”
NOT NUMERICAL value
depends only on the number of events
Prevalence overview
hard to measure
numerical and categorical measures
depends on how much event occur and how death and cure occur
Describe cross sectional studies
Exposure and outcome are measured at the same point in time- eg checking pulse rate
Good to measure the prevalence within a group
Describe the overview of Cohort study
have measurements
over period of time
can measure incidence and prevalence
Strengths of Incidence
Clean measure as its only determined by risk-
includes N/D/T
Weaknesses of Incidence
Difficult to measure as its over time
Weakness of prevalence
Less informative only includes N/D
determined by incidence cure rate and death rate
Strengths of Prevalence
easy to measure as you stop time and count
Describe RCTS
Are like cohort studies but allocated randomly to EG and CG
Strength of RCTS
equal chance of being allocated to EG and CG
- good when highlighting differences in a drug trial
- reduce chances of confounding
What is a double blind RCT
Participants/ investigators are both unaware
What are the requirements of RCTS
they can only be completes if they are ethical and practical-
Large studies are expensive
Drugs must not harm
what is RD
EGO- CGO
also known as the absolute risk difference
RR= 0 meaning?
EGO= CGO and there is no difference
RISK ratio
EGO/ CGO
Describe RR and how to compare it
RR has no units
if RR is less than 1 it = RRR
if RR is more than 1 = RRI
this is seen as a percentage
Acronym for non random error
R A M BOM AN- used in non random error
Describe R
REcruitment
who they are
do they represent the population
what is the response rate like
Describe A
ALLocation
how well were participants allocated
how were they allocated
was it subjective- questionaire or objective- measurement
Describe ADjustment in Analyses
separating the factors that influence bias
Describe confounding
exposure is mixed with another factor which will affect the outcome
Describe M
maintenance
did participants remain in there groups= EG and CG
were any groups lost in follow up
Describe Cross sectional studies relationship with M
Maintenance is not an issue- follow up does not affect outcome
Describe blind or Objective measurement
Objective is measure- not affected by personnel interpretation
questionnaire is subjective- people have different opinions
Double blind studies are good to reduce error
What is Confounding’s relationship with studies
will always influence all ecological studies
Describe RAndom error causes
everyone can not be studied
identical studies will have different results
Occurs when there is a smaller sample= greater chance results will not represent the population
Describe the 95% confidence interval
95% chance the true value in a population lies within the 95% confidence interval
-range of results likely to include the true results in the whole population
Describe the 95% CI interval on a chart
If EGO and CGO do not overlap and 95% of CI does not overlapRD= 0 or RR= 1= Statistically significant
If EGO and CGO overlap and CI of RR=1 overlap RD= 0
this means there is NO significant difference
What decreases the chances of Random error
A bigger study= less random error
WHat is a method that can be used instead of a bigger study to decrease Random error
Meta analysis
describe Meta analysis
take multiple studies and put them together
ECOlogical studes and Error
most common error Maintenance
prone to confounding
difficult to adjust
Cross sectional studies and error Strength
Strength:
recruitment
Measurement- if self reported
little to none
maintenance error
quick and cheap
Weaknesses:
Reverse causality/ confounding
Cohort study and error/ strengths/ weaknesses
easier to adjust and manage cohort in analyses
Strength: provides relationship between exposure and outcome
W: prone to confounding
Maintenance error
RCTS and error
Minimal confounding- meta analysis
maintenance error common if long-term
Very expensive= difficult to do so
RD or absolute risk is known as what?
EGO- CGO
What is the no equivalent effect value
RD= 0
or
RR= 1
What are the 4 types of random error that can occurr
R sampling error
R measurement error
Inherent biological gradient
R allocation error
Describe R sampling error and how to reduce it
a study population can never be identical. Results will produce an estimate of the truth in the population
To reduce it: averaging
Describe R measurement error and how to reduce it
it is difficult to measure biological gradients the same each time
It is hard for a human to measure each time
Reduction via averaging
Describe R allocation error and how to reduce it
Only exists with RCTS. Groups may differ by baseline by chance
Reduction: Bigger sample
Describe Inherient Biological randomness and how to reduce it
Biological phenomena are inherently random. Therefore a individuals measurement can change from time to time
How to reduce it: averaging
What is reverse causality
occurs when a relationship between EG and CG cannot be established.
What occurs when the risk difference is lower in the exposure group
results in an absolute risk reduction
What occurs when the risk difference is higher within the exposure group
results in a absolute risk increase
Describe stratified analysis
confounding can be reduced if participants are placed in different age groups.
if age groups= no difference= confounding has not occurred.
If age group= difference in numbers= confounding has occured
Describe AN within Ramboman
are differences between EG and CG adjusted for within the analysis
