Chapter 2 Flashcards
Validity
does the test accurately id whether a pt has a disease?
- gold standard test
- 2x2 table format
Sensitivity
probability that a person with disease has a positive test
- the true positive rate
Specificity
probability that a non-diseased person has a negative test
- the true negative rate
Positive Predictive Value
probability that a person w/ a + test has disease
Negative Predictive Value
probability that a person w/ a - test does not have disease
Prevalence of Disease
proportion of pts in the disease present column
Likelihood Ratio
probability of obtaining a given test result in a diseased patent divided by the probability of obtaining a given test result in a non-diseased patient
- tells us how much a test result changes the pre-test disease probability (prevalence) to the post-test disease probability
higher likelihood ratio for positive test
indicates that a positive test is more likely to be coming from a diseased person than from a non-diseased person, increasing confidence that a person w/ a + result has disease
lower likelihood ratio for negative test
negative test is much more likely to be coming from a non-diseased person than from a diseased person, increasing our confidence that a person w/ a - test does not have disease
Bayes Theorem
one way to use likelihood ratios to revise probabilities for disease
Ways to use Likelihood Ratios
Bayes Theorem
Fagan Nomogram
Natural Frequencies
Natural Frequencies
represent the joint frequency of 2 events, such as # of pts w/ disease and the # who have a + test result
Kappa Score
(reproducibility)
measures amount of agreement that occurs beyond chance (higher = better)
Precision
(reproducibility)
being able to apply the same test to the same unchanged person and obtain the same results
coefficient of variation
statistical test used to characterized precision
- the standard deviation divided by the mean value (lower = greater precision)
Primary Prevention
interventions designed to prevent disease
Secondary Prevention
screening tests designed to find disease or disease processes at an early, asymptomatic stage
Evidence Pyramid
- Systematic Reviews
- Randomized Control Trials - strongest health promotion recommendations are based on these results
- Cohort Studies
- Case-Control Studies
- Case Series, Case Reports
- Editorials, Expert Opinion
Selection Bias
- occurs when comparison groups have systematic differences in their baseline characteristics what can affect the outcome of the study
- creates problems interpreting observed differences in outcomes n/c they could result from the interventions or the baseline differences between groups
- randomly allocating subjects to the intervention is the best approach to minimize this
Performance Bias
- occurs when there are systematic differences in the care received between comparison groups (other than intervention)
- creates problems in interpreting outcome differences
- Minimize by: blinding subjects and providers to the intervention
Detection Bias
- Occurs: systematic difference in efforts to diagnoses ot ascertain outcome
- Minimize by: blinding outcomes assessors (ensuring that they are unaware of the intervention received by h\the subject)
Attrition Bias
- Occurs: systematic differences in the comparison groups in the number of subjects who do not complete the study
- failing to account for these difference can lead to incorrectly estimating the effectiveness of an intervention
- Minimize by: using an intervention-to-treat analysis, where all analyses consider all subjects who are assigned to a comparison group, regardless of whether they received to completed the intervention
statistics used to characterize the performance of a treatment of prevention intervention:
- relative risks
- relative risk difference (reduction or increase, reflecting benefit or harm)
- absolute risk difference (reduction or increase, reflecting benefit or harm)
- numbers needed to treat
- numbers needed to harm
Experimental Event Rate
probability that an intervention subject had the outcome
Control Event Rate
probability that a control subject had the outcome
Relative Risk
probability of an outcome in the intervention group compared to the probability of an outcome in the control group
Relative Risk Difference
proportion of baseline risk is reduced/increased by the therapy
Absolute Risk Difference
difference in outcome rates between the comparisons groups
Reciprocal of the absolute risk difference
number of subjects who need to be treated over a specific period of time to prevent one outcome
- if intervention increases risk of a bad outcome this statistic becomes the number needed to harm
guideline recommendations: US Preventative Services Task Force Approach
assigns 1-5 ratings and a level of certainty regarding net benefits
guideline recommendations: Grading of Recommendations, Assessment, Development, and Evaluation (GRADE)
rates quality of the evidence and grades the strength of recommendations in clinical guidelines
- primary goals (1) clearly separate quality of evidence and strength of recommendations [A,B,C] (2) provide clear, pragmatic interpretations of strong v. weak recommendations [1,2]