ISI Flashcards
Population
Characteristics of participants in study and their disease. May include setting and defined inclusion and exclusion criteria.
Age, sex, disease, referred for RT by MDT for example
Intervention
The novel therapy being evaluated.
Exposure (in cohort studies)
What have they been exposed to, of which effect is unknown
Comparator
Therapy or diagnostic with which novel intervention is being compared. Usually standard care for ethical reasons
Outcome measures
Outcome measures that are defined prospectively - blood pressure, mortality rate etc.
End points in US studies
Usually one primary and several secondary
Timing
Timing of the intervention, comparator, follow up etc.
Bias examples
Admission rate
Attrition
Change over time
Deduction
Exclusion
Financial
Intervention
Lack of blinding
Lengthy treatment: bias due to drop out
Measurement
Operator
Performance
Selection
Spectrum
Selection bias
Difference between groups in comparative study
Spectrum bias
Influence of population choice on effect size
Examples of limitations
Confounding
Underpowered/overpowered
Hawthorne effect
Single-centre
Not randomised?
Data collection process
Confounding limitation
A variable that influences both dependent and independent variable
Problem with cohort studies
Hawthorne effect
When individuals modify their behaviour because they are being observed
Performance bias
Difference in care other than the intervention
Exclusion bias
The drop out of one group more than the other which could lead to a bias in results
Deduction bias
No blinding in assessor - shouldn’t know which group patient is in
Case reports
Only have one patient
Case series
Report a series of patients - no comparator group
Cross-sectional survey
Survey hospitals to see what number of people they are treating with a disease at one point for example
Case-controlled studies
Match case with control, intervene in one but not other
Cohort study
Follow group over time that has had exposure or intervention
RCT
Randomly put into groups - one has intervention and one doesn’t
Meta analysis/systematic review
Pull together results from RCTs etc
Intervention bias
Differences in the way the intervention was applied
Measurement bias
Were they measured independetly
Change over time bias
If patients were treated a long time ago the standard of care may have changed
Attrition bias
Patients not seen through to the end may not be respresentative
Admission rate bias
Occurs when only hospitalised participants are used - could be higher rate of issues or worst case of disease if they are hospitalised than general population
Apprehension bias
When study participants responds differently because they are being observed
(can be behaviour, can be anxiousness)
Lead time bias
Overestimating the survival time due to earlier screening than usual
Volunteer bias
Participants volunteering to take part in a study intrinsically have different characteristics from the general population of interest.
