DR. GEORGE Flashcards
Comparison of populations in different places at the same time or in a time series.
Ecological study
i. These studies are also called follow-up or incidence studies
ii. Begins with a group of people who are free of disease or outcome of interest and who are classified into subgroups according to exposure to a potential cause of disease or outcome
Cohort study
i. Investigates the causes of diseases
ii. Incudes people with a disease of interest and a suitable control (comparison or reference) group of people unaffected by the disease or outcome variable
iii. Compares the occurrence of the possible cause in cases and in controls
iv. Data is collected on disease occurrence at one point in time and exposures at a previous point in time
v. They are longitudinal studies/retrospective/prospective studies
Case control study
i. Measure the prevalence of disease
ii. The measurements of exposure and effect are made at the same time1. E.g. Risk factor survey for non-communicable diseases
Cross-sectional study
i. An experiment designed to study the effects of a particular intervention (clinical trial)
ii. Subjects are randomly allocated to intervention and control groups and the results are assessed by comparing outcomes
Randomized control study
i. Involve people who are healthy but presumed to be at risk
ii. Data takes place in the field among non-institutionalized people in the general population
iii. Useful to evaluate interventions aimed at reducing exposure without necessarily measuring the occurrence of health effects
1. E.g. Testing of the Salk vaccine for the prevention of poliomyelitis
Field trial
i. The treatment groups are communities rather than individuals
ii. Appropriate for diseases that are influenced by social conditions and for which prevention efforts target group behavior
Community trial
a. When a value of the sample measurement diverges – due to chance alone – from that of the true population value
b. Causes inaccurate measures of association
a. Random error
c. There are three (3) major sources of random error
i. Individual biological variation
ii. Sampling error (small sample is not representative of the population)
iii. Measurement error (reduce through protocols and quality control procedures)
a. Must be large enough for the study to have sufficient statistical power to detect the differences
Sample size
b. Parameters required for calculation of sample size:
i. Required level of statistical significance of the ability to detect a difference
ii. Acceptable error
iii. Magnitude of the effect under investigation
iv. Amount of disease in the population
v. Relative sizes of the groups being compared
a. When results differ in a systematic manner from the true values
Systematic error
b. A study with a small systematic error is said to have a
high accuracy
d. The principle biases are
i. Selection bias
ii. Measurement or classification bias
a. When patients included in the study are not representative of the population to which the results will be applied
b. When there is systematic difference between the characteristics of the people selected for a study and the characteristics of those who are not
c. E.G
i. When participants select themselves for a study
ii. When the disease or factor under investigation itself makes people unavailable for study
Selection bias
a. When the individual measurements or classifications of disease or exposure are inaccurate – they do not measure correctly what they are supposed to measure
b. E.G. Biochemical or physiological measurements are never completely accurate and different laboratories produce different results on the same specimen
Measurement bias
Relative Risk
Relative Risk = Risk in exposed/Risk in non-exposed
Determines the ratio of the risk of disease in exposed individuals to the risk of disease in non-exposed individuals
o The ratio of the odds of development of disease in exposed persons to the odds of development of disease in non-exposed persons
Odds Ratio
o How much of the disease that occurs can be attributed to a known certain exposure?
o How much of a health problem might be eliminated or prevented if exposure were eradicated?
Attributable risk
o The amount or proportion of disease incidence (or disease risk) that can be attributed (caused by) to a specific exposure
o Addresses how much of the risk (incidence) of disease can we hope to prevent if we are able to eliminate exposure to the agent in question?
Attributable risk
In exposed persons, how much of the total risk of disease is actually due to the exposure (e.g. in a group of smokers, how much of the risk of lung cancer is due to smoking)• Calculated by subtracting the rate of disease (incidence) in the population without the risk factor from the rate of disease (incidence) in the exposed population
• In the absolute form:
• EAR (abs) = exposed risk difference = Incidence (exposed) – Incidence (unexposed)
• In the relative form:
• EAR = etiologic fraction = EAR (abs)/Incidence (exposed)
Exposed Attributable risk
o Provides an estimate of how much of the disease in the population can be attributed to exposure
o In the absolute form:
• PAR (abs) = population risk difference = Incidence (population) – Incidence (unexposed)
o In the relative form:
• PAR = etiologic fraction = PAR (abs)/Incidence (population)
Population Attributable Risk (PAR)
i. Occur when each individual can only belong to one of a number of distinct categories of the variable
ii. A categorical variable is binary or dichotomous when there are only two possible categories
iii. Examples include ‘ Yes/No ‘, ‘ Dead/Alive ’ or ‘ Patient has disease/Patient does not have disease
Categorical
the categories are not ordered but simply have names
v. Examples include blood group (A, B, AB and O) and marital status (married/widowed/single, etc.)
Nominal data
the categories are ordered in some way
viii. Examples include disease staging systems (advanced, moderate, mild, none) and degree of pain (severe, moderate, mild, none).
Ordinal data
These occur when the variable takes some numerical value
Numerical
occur when the variable can only take certain whole numerical values
Eg.iii. These are often counts of numbers of events, such as the number of visits to a GP in a particular year or the number of episodes of illness in an individual over the last five years
Discrete data
occur when there is no limitation on the values that the variable can take, e.g. weight or height, other than that which restricts us when we make the measurement
Continuous data
- When the sample included in a study is not randomly selected from the population and differs in some important respects from that population
i. E.g. when doctors collect information on patients from their clinic rather than using a random sample from the population
Ascertainment bias
When those who are lost to follow-up in a cohort study differ in a systematic way from those who are not lost to follow-up
Attrition bias
a. When mortality and morbidity rates are lower in the initial stages of a cohort study than in the general population because the individuals included in the study are disease-free at its outset
Healthy entrant effect
Caused by differences in characteristics between those who choose or volunteer to participate in a study and those who do not
Response bias
When survival is compared in patients who do or who do not receive a particular intervention where this intervention only became available at some point after the start of the study so that patients have to survive long enough to be eligible to receive the intervention
Survivorship bias
- When using a Likert scale (comprising a small number of graded alternative response such as very poor, poor, no opinion) where responders tend to move towards the mid-point of the scale (usually “no opinion
Central tendency bias
- Occurs in studies assessing changes in survival over time where the development of more accurate diagnostic procedures may mean that patients entered later into the study are diagnosed at an earlier stage in their disease, resulting in an apparent increase in survival from the time of diagnosis
Lead-time bias
When systematic error is introduced by an inaccurate measurement tool
Measurement bias
When a categorical exposure and/or outcome variable is classified incorrectly
Misclassification bias
When one observer tends to under-report (or over report) a particular variable (also called assessment bias)
Observer bias
When fitting a regression model to describe the association between an outcome variable and one or more exposure variables
Regression dilution bias
When participants give answers in the direction they perceive are of interest to the researcher or under-report socially unacceptable or embarrassing behaviors or disorders
Reporting bias
Where measurements that follow particularly low measurements tend to be higher than those recorded previously, and those that follow particularly high measurements tend to be lower
Regression to the mean
Where conclusions are based solely on aggregate statistics for groups within a population
Ecological fallacy
o When a spurious association between a potential risk factor and a disease outcome or miss a real association between them because of failure to adjust for any confounding variables
o Confounding variable or confounder – an exposure variable that is related to both the outcome variable and to one or more of the other exposure variables
o May lead to a misrepresentation of the true role of the exposure variable
o As a result, there is need to adjust for confounders in a regression analysis
Confounding