Behavioral Science Flashcards
What are types of observational studies?
1) Cross sectional
2) Case controlled
3) Cohort
4) Twin Concordance
5) Adoption
What is cross sectional study?
1) Collects data from groups to assess frequency of disease
2) Measures prevalence
What is case controlled study?
1) Compares group of people with a disease, to those without a disease
2) Will have an odds ratio
What is a cohort study?
1) Compares group with the exposure, compared to those without (see if exposure will affect the liklihood of disease)
2) Can be prospective or retrospective
What is a twin concordance study?
1) Frequency that monozygotic twins get same disease
2) Heritability, and influencd of the envrionment
What is adoption study?
1) Mesures heritability and influance of environmental factors.
What are the phases of clinical trials, and what is purpose?
Phase 1: Small healthy individuals (is it safe)
Phase 2: Small number of patients with diease (does it work, dose, and adverse effect)
Phase 3: Large number of patients assigned to treatment or best available treatment (is it good or better)
Phase 4: Postmarketing surveillence after approval (can it stay, detect long-term adverse effects.
How to evaluate the diagnostic tests? How to calculate sensitivity? How to calculate specificity? Positive predictive value? Negative predictive value?
2 x 2 table
Disease Test + -
+ TP FP
- FN TN
Sensitivty TP/ (TP+FN) (proportion of sick people who are identified as such)
Specificity TN/(TN+FP) (healthy people correctly identified as not having the disease)
Positive predicitve (number of positive tests, with patients have the disease) TP/ (TP+ FP)
Negative predictive (number of negative tests, with patients without diease (TN/ (TN+FN)
What does high sensitivity mean?
High sensitive tests, means when negative, then rules out
What does high specificity mean?
High specificity, when the test is positive, diease is ruled in.
what affects the PPV?
The prevalence of the disease
If have high prevalence, have high pretest probability, and high PPV
What affects the NPV?
If have high pretest probability, then have low NPV
What is the incidence?
What is the prevalence?
Incidence: Number of new cases/ number of people at risk
Prevalence: Number of existing cases/total number of people
What is the effect of prevalance and incidence depending on the disease?
Prevalence and incidence approximately equal when short duration of disease.
Prevalence > incidence when have chronic disease
What are ways of quantifying risk?
1) Odds ratio
2) Relative risk
3) Attibutable risk
4) Relative risk reduction
5) Absolute risk reduction
6) Numbers needed to treat
7) Numbers needed to harm
What is the table to quantify risk?
Disease
+ -
Risk of intervention
+ a b
- c d
What is odds ratio?
Case controlled studies
What is relative risk?
Cohort studies (risk of developing disease if exposed or not)
What is attributable risk?
The difference in risk between the exposed and unexposed group
What is relative risk reduction?
Risk reduction due to an intervention
Absolute risk reduction?
The difference attributable to an intervention (controlled study)
What is number needed to treat?
Number of patients treated for 1 patient to benefit
What is the number needed to harm?
Number of patients that need to be exposed to a risk factor for 1 patient to be harmed.
What is precision?
Consistency and reproducible of a test
How does precision affect random error?
How does it effect standard deviation?
How does it affect statistical power?
Decrease random error
Decreases standard deviation
Increases statistical power
What is the accuracy of the test?
The trueness of the measurement
What does systemic error do to accuracy of test?
decreases
What are types of selection bias?
Berkson bias: study population from hospital is less healthy
Healthy worker population effect: study population is healthier then normal population
Non response bias: participating subjects differ from non respondants
How to prevent selection bias?
Randomization
What is recall bias?
Patients with disease more likely to remember exposure
Measurement bias?
Association between the disease and the exposure not seen in non-standardized tests
What is procedure bias?
Subjects in different groups not treated the same
what is observer-expectancy bias?
Reasercher believes in the treatment
What is confounding bias?
Factor related to exposure and outcome, but not causal
How to reduce the confounding bias?
1) Mutliple studies
2) Crossover studies
3) Matching
4) Restriction
5) Randomization
How to reduce recall bias?
Decrease time from exposure to followup
What is a lead time bias?
Early detection, confused with increased survival
survival has not increased, earlier detection