Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
78 Cards in this Set
- Front
- Back
Primary prevention
|
Preventing initialization (immunization)
|
|
Secondary Prevention
|
Early detection (cancer screening)
|
|
Tertiary Prevention
|
Reducing impact - rehab
|
|
Ignaz Semmelweis
|
Childbred fever/washing hands
|
|
Edward Jenner
|
Smallpox/vaccination
|
|
John Snow
|
Cholera/water pump
|
|
Wade Hampton Frost
|
Epi of infectious diseases/life tables
|
|
Clinical vs. non-clinical disease
|
Clinical shows symptoms
|
|
4 non-clinical disease types
|
Preclinical - before symptoms, going to have symptoms
Subclinical - Not apparent, won't become apparent Persistent - Can't shake off Latent - No active multiplication |
|
Epidemiological Triad
|
Host
Agent Environment (Vector) |
|
Endemic
|
Routine appearance of disease
|
|
Epidemic
|
Occurrence of disease in a certain region, in excess of normal, derived from a common source
|
|
Pandemic
|
World-wide epidemic
|
|
Herd immunity
|
Group resistance due to a large portion being immune
|
|
Incubation period
|
Interval from infection to the time of onset of clinical illness
|
|
3 variables of epidemiological investigation
|
- Time of exposure
- Time of disease onset - Incubation period |
|
Steps of epi investigation
|
1. Define
2. Examine 3. Look for relevant variables 4. Develop hypothesis 5. Test hypothesis 6. Recommend control measures |
|
Attack rate
|
# of cases among people at risk / Total number at risk
|
|
Prevalence equation
|
Prevalence = Incidence * Duration
|
|
Incidence
|
Number of new cases in a defined population and time / people at risk
|
|
Incidence/Prevalence/Attack rate - Rate or proportion?
|
Incidence - rate
prevalence - proportion attack rate - proportion |
|
Prevalence
|
Number of cases present at time / number of person at that time
|
|
Crude mortality rate
|
Total deaths from all causes / Population at midyear
|
|
Case Fatality Rate
|
People who died from a specific disease in a time period / people with that specific disease
(proportion) |
|
Proportionate mortality
|
Number of deaths from a specific disease / Total number of deaths
|
|
Direct Age Adjustment
|
find death rate for each age group, apply to standard population
|
|
Indirect Age adjustment
|
Standardized Mortality Ratio = Observed # of deaths per year / expected # of deaths per year
|
|
Cohort Effect
|
Plot prevalence by birth group, not age group
|
|
YLL
|
years of life lost due to death
|
|
YLD
|
years of healthy life lost due to disability
|
|
DALY
|
YLL + YLD = years of healthy life lost due to disability/death
|
|
Two assumptions in life tables
|
- No changes in survivorship
- Those lost to follow have the same experience |
|
Observed survival rate
|
Use life tables, Kaplan-Meier, or person-years
|
|
Relative survival rate
|
Observed survival rate / Expected survival rate
|
|
Validity measures
|
Sensitivity and specificity
|
|
Prevalence vs. Predictive Value
|
Higher prevalence = higher positive predictive value
|
|
Biggest factor of PPV
|
Specificity
|
|
Simultaneous vs. Sequential: Spec/Sens
|
Simult: Sens increases, Spec decreases
Seq: Sens decreases, spec increases |
|
Reliability
|
How well an observer classifies the same individual under different circumstances
|
|
Validity
|
How well a given test reflects another test of greater accuracy
|
|
Kappa
|
(% agree - % chance) / ( 1 - % chance)
|
|
Kappa grades
|
.4 - .75 moderate
.7+ good |
|
Advantages of Case study vs. Cohort study
|
Case is less expensive
Case is better for infrequent diseases Case is quicker |
|
Nested case-control studies
|
Controls are a pool of people at risk when the disease develops. When an individual gets sick, he goes into the case cohort, and someone simultaneously goes into the control group.
|
|
Case-cohort studies
|
Defined cohort at beginning, # of cases that are developed over time are matched to a control group all at the end.
Can use multiple diseases |
|
Advantages of nested case-control and case-cohort studies
|
All samples are taken before, so no selection bias and no "chicken or egg" problem
Only need to test those infected and matching controls, not all samples More comparability |
|
Case-crossover study
|
Studies etiology of acute outcomes.
Environmental exposure at time of case is compared to a certain time before case. Adv - Cheap, same person is control Dis - Recall bias, disregards other factors |
|
Cross-sectional study
|
Study both exposure and outcome simultaneously
Compare prevalence of those exposed to those not exposed Adv - Cheap, easy Dis - Chicken or egg AKA prevalence study |
|
Confounding by indication
|
The bias to systematically pick people for treatment who are most ill
|
|
Stratification
|
Allocating randomization by certain characteristics
|
|
Masking/blinding
|
Patient doesn't know what they're on
|
|
Double blinding
|
Patient and investigator both don't know what treatment the patient is on
|
|
Levels of data analysis
|
Primary - Intention to treat
Secondary - observing those already treated Subgroup - planned in advance |
|
ITT
|
intention to treat - analyze purely on randomization, ignore non-adherence, switched, ineligibility, etc.
Most conservative approach |
|
How to deal with non-compliance
|
- pilot studies
- active monitoring |
|
Cross-over design
|
Each patient does both treatments
Adv - Serves as own control Dis - generalizability |
|
Factorial design
|
Testing two treatments at once
|
|
Group allocation design
|
Randomization is group of individuals, i.e. school, neighborhood
|
|
Relative Risk (efficacy)
|
Rates of bad events: (rate in treated) / (rate in untreated)
|
|
Relative risk reduction
|
Bad events: (rate in untreated - rate in treated) / (rate in untreated)
Also: 1-relative risk = 1-(treated/untreated) |
|
Absolute risk reduction
|
(bad events): Rate in untreated - rate in treated
|
|
Number needed to treat
|
bad events: 1 / (rate in untreated - rate in treated)
|
|
Generalizability
|
Validity of test to reflect population
|
|
Clinical phases of new drugs
|
Phase 1: Determine level of toxicity
Phase 2: Clinical investigation on a small amount of people Phase 3: Clinical trials on large population after safety is shown Phase 4: Post-marketing |
|
Types of error
|
Type 1: Treatments are same, conclude they are different (alpha)
Type 2: Treatments are different, conclude they are the same (beta) |
|
Power of study
|
Ability to correctly identify that treatments are different (1 - beta)
|
|
Cohort study
|
Start with exposed/not exposed to environmental factor (smoking), divide into disease or no disease
|
|
Relative risk in cohort study
|
(incidence among exposed) / (incidence among non-exposed)
|
|
Attributable risk in cohort study
|
(incidence rate for exposed) - (incidence rate for non-exposed)
|
|
Bias in assessment of outcome
|
Unmasked investigator might skew results towards hypothesis
|
|
Information bias
|
Those treated might have more in-depth information on treatment/disease, especially in retrospective study
|
|
Non-response bias
|
Healthy are less apt to respond
|
|
Analytic bias
|
Investigator bias while analyzing data, may use prior biases to help hypothesis
|
|
Selection bias
|
Who gets into the program - healthy people volunteer more
|
|
When to use cohort study
|
- Good evidence of association of disease/exposure
- Incidence of disease among exposure is high - Time between exposure and disease is short |
|
Adv/Dis of cohort study
|
Adv: Can assess several outcome at once
Adv: Control of time and outcome measurements Adv: Less potential for bias Dis: Large samples Dis: long follow up Dis: not efficient for rare outcomes |
|
Case control study
|
Start with disease/no disease, see who was exposed
|
|
When to do case control
|
- Rare disease
- Need quick answers about exposure - Limited funding - Not able to follow exposed/non-exposed easily |