Types of Studies Flashcards
2 types of edpidemiologic studies
observational
experimental
Experimental Studies
an active attempt to change a disease determinant or the progress of a disease through treatment
what do major experimental study designs include?
- randomized controlled trails,
- field trials, (healthy participants)
- community trials (ppl in the community)
randomized controlled studies (clinical trials)
- designed to stude the effects of a particular intervention, usually a treatment for a specific disease
- partients are alloved into groups randomly
- ## should be comparable at the start of the investigation
Field Trials
- involve healthy people who are presumed at risk
- aimed at reducing exposure
- often ligistically complicated and expensive
FIeld Trial groups
preventiatice intervention and no intervention (control group)
community trials
treatment groups are communities rather than individuals
- diseases that are influenced by social conditions
- prevention includes changing behaviours
Limitations of a community trials
- small number of communitites can be involved
- differences can be attributed to the study and not other factors
- random isn’t always possible
- difficult to isolate the general social change
Sources of Error
- Random
- Systematic
Random Errors
when the value of a sample measurement diverges from the true population value due to chance alone.
3 major sources of random errors
- individual biological variation
- sampling error
- measurement error
individual biological variation
always occurs no measurement is perfectly accurate
Sampling Error
caused by the fact that a small sample is not representative of all the population’s variables
- increasing size can help, not necessarily
Measurement error
can be reduced by strict protocols and making individual measurements as precise as possible
What do you need to perform sample size calculations?
- statistical significance to be able to tell the difference
- acceptable error
- magnitude of the effect
- amount of disease
- relative sizes
how many controls should you have for each case
more than four controls for each case
Systematic Error (or bias)
occurs when results differ in a systematic mannor from true values
high accuracy
a study with small systematic error, not affected by sample size
2 principle biases
selection bias
- measurement or classification bias
Selection Bias
a systematic difference between the characteristics of the people selected for a study and the characteristics of those who are not
Healthy Worker Effect
when the disease or factor under investigation makes people unavailable for the study
Measurement Bias
occurs when the individual measurements or classification of disease or exposure are inaccurate
- dif labs cna get dif results
recall bias
occurs when there is a differential recall of information by cases and controls
participants don’t remember past things accurately.
blind study
investigators do not know how the participants are classified
double blind study
investigators and participantsand researchers do not know the treatment or intervention
Confounding
can occur when another exposure exists in the study population and is associated both with the disease and the exposure being studied.
methods of controlling confounding
- randomization
- restriction
- matching
analysis stage: - stratification
- statistical modeling
Randomization to control confounding
the method for ensuring that potential confounding variables are equally distributed among the groups being compared.
Restriction to control confounding
limiting the study to people who have particular characteristics, limit such that confounding characterisitcs are not present in the sample
Matching to control confounding
selecting study participants to ensure that potential confounding variables are evenly distributed in the two groups being compared, matching a person with a confounding variable to another.
Stratification /Statistical Modelling to control confounding
used in large studies because it preferable than doing things during design
Stratification
involves the measurement of the strength of associations in well defined and homogenous categories, if age is a confounding variable measure in age groups
Statiscal modeling
range of statistical techniques
validity
an expression of the degree to which a test is capable of measuring what it is intended to measure
valid = responds to the truth
2 types of validity
internal validity
external validity
Interval validity
the degree to which the results of an observation are correct for the partical group being studied
necessary but does not guarantee external validity
External Validity
the exntent to which the results of a study apply to people or laborities not involved
- external quality control
- judgements
- supported if similar results are found in studies in other populations
Informed Consent
free and voluntary informed consent must be obtained, must retain the right to withdraw at any time
an observational study
investigate measures, but it does not intervene
2 types of observational studies
- descriptive study
- analytical study
descriptive study
limited to a description of the occurrence of a disease in a population often the first step in an epidemiogic investigation
analytical study
goes further by analysing relationships between bealth staus and other variables
ecological studies
or correlational studies and useful for generating hypotheses, units of analysis are groups
- comparing different populations through time or place
ecological fallacy
inaproriate conclusions fromt he study are drawn, the association observed at the group level does not reflect that at the individual level.
Cross sectional studies or prevalence studies
measure of the prevalence of disease
measurements of exposure and effect are made at the same time
effect line follows exposure line there is a correlation
Case-control studies
involves cases and suitable control group of people unaffect by the disease or outcome variable
cases
people with a disease or other oucome variable of interest
control group
the comparison of reference of the case group - healthy ppl
Cohort studies
aka follow-up or incidence studies
begin with a group of people who are free of disease and who are classified into subgroups according to exposure to a potential cause of disease or outcome
Nested Case-Control Studies
cases and controls are both chosen from a difined cohort
risk =
attack rate
risk among people who are exposed =
people with disease / total number of people exposed
Risk among people not exposed
number of people with disease / total number of people not exposed
Relative risk =
risk among people exposed / risk among people not exposed
If relative risk >1
exposure and disease are positively associated (more exposed people have the disease)
Relative risk = 1
exposure and disease are not associated
relative risk <1 exposure
and disease are engatively associated (more unexposed people have the disease), those exposed are protected from the disease
attributable risk =
risk of people exposed - risk of people not exposed
the proprtion of people who showed syptoms that can be credited to the exposure
null hypothesis
the exposure is not related to the disease, is 1
alternatie hypothesis
if the null hypothesis prove implausible, exposure IS associated with disease
ODDS =/ probability
odds calculation
number of times something happens/something DOES NOT happen
probability calculation
number of times something happens/number of times something COULD happen
case control calculations
odds
Cohort calculations
risks
odds ratio =
odds of disease among people exposed/odds ff disease among people not exposed
if odds ratio >1
exposure and disease are positively associated - more diseased people exposed)
if odds ratio = 1
exposure and disease are not associated
if odds <1
exposure and disease are negatively associated, more non-diseased people are exposed
