CPM Test 3 Flashcards
(19 cards)
Best study design for screening test
Cohort
Sensitivity
Specificity
TP/ TP+FN
TN/TN+FP
Screening (snOUT)
Confirmatory (spIN)
Maximize sensitivity
SnOUT (Rule out)
Maximize specificity
Rule in
Sequential Testing
Confirmatory test after + screening test
Dec sensitivity
Inc specifity
PPV
NPV
TP/TP+FP, want high
TN/TN+FN
Prevalence and test interpretation
High prev
Low prev
Higher PPV (+ indicates dz) Lower NPV (Less likely to indicate no dz)
Low PPV (less likely to indicate dz) High NPV (more likely to indicate no dz)
Likelihood ratio
Likelihood that given test result would occur in pt w target disorder / likelihood that given test result would occur in pt wo target disorder
P LR
N LR
Sens/ (1-Specificity)
(1-Sens)/Specificity
Likelihood ratio values
1= unhelpful
moderately above/below 1= not that great
Well above/below 1= good test
Dz that can be screened
Important health problem (high inc or mortality)
Treatment is + beneficial in early stages
Early dx improves survival/QOL
Appropriate screening tests
Simple, safe, precise
Detect latent/early stage
Benefit outweighs harm
Screening benefits
Know dz inc/mortality
NNS
Inc average life expectancy (limited usually)
Screening harms
Adverse effects (discomfort/pain/radiation)
Overdiagnosis (yrs before manifestations/anx)
FP (anx, stigma, testing)
Screening progam
Cost is balanced to HC expenditure
Systematic plan for monitoring program for quality assurance
Pt education about +/-
Ratio of cost effectiveness
Incremental cost / health benefits
Overdiagnosis bais
Overestimation of survival from screening caused by inclusion of pts w asymp dz that would not affect pt before they die of another cause
Lead time bias
Overestimate survival from screening when survival when measured from time of dx
Length bias
Overestimate survival caused by excess of slow progressive cases
Screening Test Study Designs vs RCT
Good= observational (CC or cohort)
Volunteer bias tho
RCT- evaluation
challenging
