Test 2 LEcture Flashcards
primary, secondary and tertiary prevention
Primary - avert the occurrence of a dz or injury…might be the best way
seoncdary - early detection and intervention to slow or reverse
Tertiary - dz management to improve patient function
RCT things that still might not be good
Outcome assessment or follow-up
Random error and how to reduce
Larger population
Graph of 1 and 2 erorrs
1 will be closer to the non-control group
Radndomization
Each pt has equal chance to be assigned
Also random allocation
Stratified randomization
Balance randomization across a confounder
Allocation concealment
Person enrolling patients i unaware of the assignment…occurs prior to or time of randomization
May use a sealed envelope
Lack of concealment is a source of selection bias bc characteristics may change allocation
Patient outcomes
Patient recall (most subjective)
Discharge dx (better)
Case definition (best)
Intention to tx
Once randomized, always analyzed
Preserves baseline comparibility
Simplifies dealing with missing outcomes and poor adherence
Addresses withdrawl
Reflects real owrld outcomes from a tx
Need to enroll more pts to account for those that don’t finish…may underestimate effect size
Assessing directness
DIff bt pop and study participants (Table)
Diff in tx of pop and study part
Use of surrogate outcome measures instead of outcome of interest
INdirect comparioson bt presence/absence of exppsore or risk factor (combinatorial appraoch)
How to assess consistency
Effect estimates across subgroups and outcome measures
Heterogeneity is minimal
Assessing precision
Use the CI
