Align module Flashcards
What is an experimental research design
When the investigator specifies the intervention as part of the research design. People are observed, outcome is measured
What are the concepts of a RCT
Equipoise Randomization Placebo Masking (blinding) Bias
Wat is an observational study
Investigator does not specify intervention of research purposes. People are observed over time, outcomes are measured
Can be retrospective or prospective
What kind of study has more risk of bias
Observational
What types of studies are considered “controlled observational”
Case control (look at pts with and without condition at a certain point in time) Cohort (follow pt w/ or w/o disease over a certain amount of time)
What is considered an “uncontrolled observational study”
Case report
Why are cross sectional studies difficult to infer
Because we aren’t observing patients over time
Which study type is not good for studying multiple outcomes
Case-control, because it is a retrospective study
What are problems in research design
outcomes and endpoints
We want answers ASAP
short term outcomes may not predict long term benefits
short cuts lead to negative consequences
What are solutions in research design
consider future patients and get right answer for long term
Make sure RCT has adequate follow up
Make sure RCT is carried through completion
What are the levels of evidence (ascending)
- Unsystematic clinical observation (stories of people trying to deal with healthcare issues)
- Physiologic studies (vitals, imaging, lab values)
- Single UNcontrolled obs study/case series (not good for finding cause/effect)
- Single controlled obs study (still affected by bias)
- Systematic review of controlled obs (doesn’t correct for bias in design)
- Single randomized trial
- Systematic review of RCT (BEST- confirms that results of singe RCT weren’t coincidence)
What is the primary purpose of research
to benefit future patients
What are important lessons learned for consumer advocates (ERT and CVD)
There was an over reliance on observational studies
Systematic review (meta) of RCT contradicted observational findings
Clinical guidelines encouraged prescription of hormones for prevention w/o conclusive evidence
Pharmaceuticals emphasized benefit over risk and underwrite med ed for physicians
Natural aging was transformed into a disease in need of treatment
What is an ecologic study
a type of observational study that can find correlations of disease in groups (not individuals) but can’t show cause
ex: public smoking
What is an ecologic study good for
generating hypothesis for future studies
What is a cross sectional study good for
finding prevalence of a condition
What are the problems with a case-control
masking is not possible
BIAS
not randomized
How can you fix the problem of selection bias associated with a cohort study
RCT’s!
Compare controlled usb uncontrolled study
Controlled: more than one group studied (comparison/control Vs. Experimental)
Uncontrolled: single group studied, no comparison made
What is selection bias
not true randomization
selection of people into a group is biased
What is attrition bias
exclusion of people entered into a study
one side has fewer or more participants, affecting statistical conclusion
How do you adjust for attrition bias
Use intention to treat analysis
What is performance bias
difference in care provided and exposure to other factors
How can you correct for performance bias
Use masking/ double blind
What is information bias
difference in how data is collected and assessed between groups
(use double blind)
What is a confounding variable
factor associated with intervention and outcome of a treatment
What is a surrogate end point
outcome measures not of direct importance, but believed to reflect outcomes
What is equipoise
uncertainty whether experimental intervention is better than comparison
What is Intention to treat analysis
All patients enrolled in the study are included in the analysis, regardless of dropping out?
