Epidemiology - Block 2

davidwurbel7's version from 2015-06-25 20:48

Cohort Studies

Question Answer
Individuals are divided on the basis of presence or absence of risk factors. Individuals are then followed over time to determine if they develop a specific outcome or diseaseCohort Study
Individuals sharing same experience and followed up for specific period of timeCohort Study
A factor of no interest in the study that can interfere with the results of the studyConfounder Factor
Initiation of study occurs before occurrence of disease / outcome of interestProspective Cohort Study
Initiation of study occurs after occurrence of disease / outcome of interestRetrospective Cohort Study
Bias is more likely in which study typeRetrospective Cohort Study
Best study type for assessing the effects of rare exposuresCohort Study
Free from the exposure, but otherwise similar to the case groupComparison Group
When individuals participating in a study change behavior (mainly towards positive) for a temporary period of time, due to the fact that they are being observedHawthorne Effect
Usual sources e.g medical / death certificates, interviews, questionnairesSources of Data
Death certificates, Hospital records, Periodic health exams of cohort, QuestionnairesSources of Outcome Data
A study that loses greater than this percentage of participants is flawed> 25%

Case Control Studies

Question Answer
Starting point is divide the group into two groups based on those with disease and those without diseaseCase Control Studies
Rare diseases can not be analyzed easily using another approach. Chronic diseases (e.g. cancer) have long latency periods. Time and money issues: cost efficient and less time to complete. Many exposure factors can be studied at the same time are benefits of this study typeCase Control Study
Difficulties choosing appropriate controls. Cannot get true estimate of risk. Issues with temporal association. Strong potential for bias. Confounding are some issues with this study typeCase Control Studies
the most difficult issue in case-control designControl Group Selection
Investigators usually use 2 to 4 control groups, selected in different waysMultiple Controls
Create strata of the confounding variableStratified Analysis
The estimate of relative risk can be calculated in a case control studyOdds Ratio
To calculate odds ratio - ratio of odds the cases that were exposed divided by ratio of odds the controls that were exposedOdds Ratio
(A)(D) / (B)(C)Alternate Odds Ration Calculation
Can lead to an over or underestimation of riskRecall Bias
The persons in the case group might be able to recall exposures better than those in the control groupRecall Bias
Interviewers who are aware of the study hypothesis are likely to question cases and controls differentlyInterviewer Bias
Experimenter’s expectations are communicated to the subjects, unintentionally or intentionally. The subject then produce the desired effectPygmalion Effect
Subjects may be erroneously categorized with respect to exposure or disease statusMisclassification Bias
Method of collecting information was flawedMeasurement Bias
Controls picked from same source as cases, use motivated individualsControl of Selection Bias
One hospitalized control groupControl of Recall Bias
Highly trained personnel, blinded to study hypothesis, standardized questionnaires etcControl of Interviewer Bias
Use standard sources to validateControl of Misclassification Bias
Common example of these include age, gender, tobacco, alcohol, socio-economic statusConfounders
A way to reduce the effect of these in a study is to place restrictions on the study subjectsConfounders

Clinical Trials

Question Answer
Researcher have full control over the exposure received by the participantsInterventional Studies
Evaluate new drugs or to evaluate new treatment proceduresRandomized Clinical Trial
The subjects get both treatments in sequence. Each subject serves as his / her own control. A subject is randomly assigned to a specific treatment orderCrossover Design
The general group to whom the results will be applicable toReference Group
Actual group in whom the study is conductedExperimental Population
Reduces bias due to known and unknown confoundersRandomization
Only the experimenter knows the assignment of subjects, subjects do notSingle Blind
Neither the experimenter nor the subjects know the assignmentsDouble Blind
The subject, the experimenter and data analysts are blind to the assignmentsTriple Blind
Preferred control groupConcurrent randomized
Preserves random allocation, simulates real world experience. Analysis of all people that were in the trial regardless of if they completed the trial or not.Intention to Treat
Only analyzes those who actually take treatment. Those that did not complete the trial or were non-compliant are excluded from the analysis.Explanatory Analysis
Analyses in the medical literature (drug trials) should always be analyzed and reported by this methodIntention to Treat Method
The ability of a treatment to work in the ideal study settingEfficacy
Efficacy used this type of analyzesExplanatory
The ability of the treatment to work under realistic circumstancesEffectiveness
Effectiveness used this type of analyzesIntention to Treat
Clinical trial done with individual patients. Patient is given treatment or placebo randomly at different timesN Equal to 1 Trail
Animal studiesPreclinical Phase
Initial testing in healthy human volunteers following animal studies. Identify dose limiting toxicities, tolerated doses, describe pharmacology (metabolism, excretion)Phase I
Testing in subjects with disease to determine activity and therapeutic efficacy, validate toxicity, side effects and dosage dataPhase II
Randomized trials for comparison with Standard therapyPhase III
Studies are done after the drug or treatment has been marketed to gather information on the drug's effect in various populations and any side effects associated with long-term usePhase IV
Informed consent must be obtained from all participants involved in human experimentsDeclaration of Helsinki 1964
Patients must be aware the study hypothesis. They must understand that they can be assigned to treatment or placebo arms. They must be told all possible consequences of participation. Minors can only enroll with guardian’s consent. An ethics board must oversee the studyInformed Consent
Over time knowledge about a disease treatment may come to light. New treatments may become available. If the results show a sustained statistical association, it is unethical to withhold treatment from the placebo armStopping Rules


Question Answer
A strategy used to identify disease in an unsuspecting populationScreening
These types of tests should be simple, rapid and preferably inexpensiveScreening
Basic purpose is to detect disease from a large group of apparently well persons earlyScreening Test
Whether a lab test consistently gives the same value when multiple tests are conducted on the same sampleReliability
Kappa value is at least 75% indicatesReliable Test
The screening test’s ability to do what it is supposed to doValidity
If the disease is present how often does the test detect itSensitivity
If the disease is not present how often does the test correctly gives a negative resultSpecificity
True positives divided by true positives + false negatives times 100Sensitivity Formula
True negatives divided by true negatives + false positives times 100Specificity Formula
True positives divided by true positives + false positives times 100 Positive Predictor Value (PPV)
True negatives divided by true negatives + false negatives times 100Negative Predictive Value (NPV)
Proportion of all subjects who were correctly classified by the testAccuracy
True Positives + True Negatives / Total ScreenedAccuracy
An increase in the specificity will cause an increase inPositive Predictor Value (PPV)
An increase in the prevalence will cause an increase inPositive Predictor Value (PPV)
Screening in high risk groups is used forRare Diseases
Lead time bias, length bias and selection biasScreening Bias
The perception that the case has a longer survival simply because the disease was identified earlier in the natural course of diseaseLead Time Bias
Misperception that screening itself leads to better outcomesLength Bias
Individuals who are motivated to participate in screening programs may have a different probability of disease than individuals who refuse to participateSelection Bias

Causality and Association

Question Answer
If the interval does not contain 1.0 then the association between the variables isSignificant
Study design, Strength of Association, Consistency, Correct Temporal Relationship, Dose Response Relationship, Plausibility, Specificity and AnalogyHill’s Criteria
Cause must always precede effectTemporal Relationship
A case-control study can be inserted into a cohort studyNested Case-control Studies
Formulate a valid question, identify references, critical appraisal, applying the evidence, re-evaluationEvidence Based Medicine