Epidemiology test 1

Definition of epidemiology
the study of how health related states are distributed in populations and the factors that influence or determine this distribution in order to reduce health problems

Goals of epidemiology Look in the book
1)to identify the etiology or cause of a disease and relevant risk factors that increase a person’s risk for disease

2)determine the extent of disease found in the community

3)study the natural history and prognosis of disease

4)evaluate both existing and newly developed preventative and therapeutic measures and modes of health care delivery

5)provide foundation for developing public policy relating to environmental problems, genetic issues, and other considerations regarding disease prevention and health promotion

Iceberg concept
we only know what is on the surface; example HIV
– 20% of HIV in US are unaware that they are infected
-unaware HIV transmit approximately half of the new HIV infections
-important to not only count the clinical cases because more people could have it that don’t know

Endemic/Epidemic/Pandemic
endemic: habitual presence of a disease in a region
epidemic: excess of normal expectancy
pandemic: world-wide epidemic

Exposure types (common-vehicle single/multiple periodic/continuous)
common vehicle: all cases develop among people exposed to something
single exposure: spoiled egg salad at a picnic
vs. multiple: contaminated air supply in a building
periodic vs continuous

Herd immunity
-relies on having enough immune hosts to interrupt the chain of transmission

-a communicable disease is reliant on the chance that a person who is infected will transmit the infection to another susceptible host
-but not all hosts need to be immune to stop the spread
-depends on how effectively the disease is transmitted
-exposures are not necessarily random!

Incubation period
how long it takes for infection to lead to clinical illness
can vary based on: disease, rought of exposure/invasion, dose of exposure,
-many diseases spread during the incubation period

Attack rate*
equation:

AR = (# of at-risk who get sick) / (Total # at risk)
Similar to incidence, but used more frequently for infectious disease
Food-specific AR
Time is implicit (diseases with known incubation period)

Cross-tabulation
determining which factor caused illness
A table or matrix that shows the relationship between two or more variables

Person-time (e.g. person-years)*
different people may have been at risk for different lengths of time aka incidence density
-accounts for people who were only following through part of the study
-1 person observed for one full year= 1 person-year
– use number of people time number of years each is at risk/observed (but each year is assumed to be the same)

Morbidity measures*:
incidence (person-time), cumulative incidence, point prevalence, period prevalence)

incidence
# new cases in population during a specified time period)/total person-time (sum of time periods of observation of each person who has been observed for all or part of the time period)

cumulative incidence
or incidence proportion is a measure of frequency, as in epidemiology, where it is a measure of disease frequency during a period of time.

point prevalence
is the number of persons with disease in a time interval (eg, one year) divided by number of persons in the population; that is, prevalence at the beginning of an interval plus any incident cases.

period prevalence
is the proportion of a population that has the condition at some time during a given period (e.g., 12 month prevalence), and includes people who already have the condition at the start of the study period as well as those who acquire it during that period.

Know the relationship (conceptual and mathematical) between incidence and prevalence
In “incidence”, there is a notion of dynamic that is absent in “prevalence”. Incidence refers to the number of new cases occurring during a given period of time (ex: in my village there were 12 new cases of flu during the winter), while prevalence corresponds to the proportion of cases at a specific point in time (on 3rd Feb, there was only one person who had flu in my village).

Mortality measures*:
all-cause, group specific, cause-specific, case fatality rate, proportionate mortality)

all-cause
measures rate of deaths.

(#deaths from all causes in 1 year)/ (#persons in the population at mid-year)

group specific
(#deaths from all causes in 1 year in defined group)/ #persons in the population in defined group at mid-year)

cause-specific
(#deaths from specified cause in 1 year) / (#persons in the population at mid-year)

case fatality rate
(#ind. dying during specified period of time after disease onset or diagnosis) / (#ind with the specified disease)

proportionate mortality
(#deaths from particular disease in a year) / (total # deaths in a year)

Also know the concept & purpose of age adjustment (not the calculation)
no calculations

Cohort effect Concept
people may seem older over the study

Screening tests*:
validity, reliability, sensitivity, specificity, positive predictive value, negative predictive value

validity
ability to distinguish b/w who has a disease and who does not

reliability
repeatability

sensitivity
ability to correctly identify those with the disease

positive predictive value
proportion who test positive actually have the disease

negative predictive value
proportion who test negative who do not have the disease

Note that 4×4 table layouts may or may not match that in the book – know how to apply the calculations based upon the concepts

Survival*:
5-year survival, median survival, relative survival

5year survival
percentage of indivs alive 5 years after diagnosis
proportion/%alive after 5 years
limitation: lead time, no info <5 years, trajectory may be different

median survival
the length of time that half of the study population survives

-time at which 1/2 have died
limitation: there is a high level summary measure that does not capture differences over time

relative survival
(observed survival in people with the disease) / (expected survival if disease were absent)

-: Observed Survival / Expected Survival
-Useful when comparing groups where survival is not close to 100%

Survival: Life tables, Kaplan-Meier Be able to distinguish between these
life tables: using fixed time intervals (ie year 1, year 2, year 3 etc)
– more complex appraoch but can use more data
– uses year by year probability of survival (given that you survived the preceding year)
– allows us to view the overall probability of survival by year of follow-up
-accommodates losses to follow-up by estimating the number at risk

assumptions:
1)No change in survivorship over study
2)Withdrawals survivorship is the same
as those retained
3)People lost to follow-up are lost uniformly over the interval
4)The event occurs uniformly over the interval

-rates calculated at pre-specified intervals (e.g., years)
-Limitations: changes in detection/Tx, LTFU (must make assumption that survivorship is the same)

Kaplan-Meier
using all event times ; calculate survival every time someone dies

-: Calculate survival at each time of death
Limitations: Similar to life tables, but LTFU is handled at calculation times

Generalizability (external validity)
is the study representative of the population in a way that is meaningful ;
“common” experiences with diease (eg signs, symptoms, prognossi) vary according to population and clinic based samples

Exchangeability
everything should be the same except the independent variables (exposures of interest) should be different so you can compare the dependent variable (outcome)

Randomization
participants randomly assigned to one group or another which will remove the potential that researcher or self-selection will result in differences b/w groups beyond exposure

stratified randomization
a two-stage procedure in which patients who enter a clinical trial are first grouped into strata according to clinical features that may influence outcome risk. Within each stratum, patients are then assigned to a treatment according to separate randomization schedules [1].

crossover (planned vs unplanned)
Time when subjects may switch between study groups: Planned crossover, and Unplanned crossover

planned crossover
Planned crossover
Randomly assign subjects to study groups
Observe the subjects over time
Switch the subjects to the other study group
Observe them over time
Compare the changes between the subjects with one treatment versus the other treatment

unplanned crossover
Unplanned crossover
May occur if someone in one treatment group needs to be medically managed like the second treatment group instead because of health concerns
You may have randomized someone to one treatment group but they decide they don’t want to be in that treatment group

factorial design
Use one study population to study two drugs
Drugs must be different and mode of action of each drug (how it works in the body) must be different
Allows you to use a study population most economically

type 1 error
type 1: alpha treatment differs when they really don’t

type 2 error
type 2: beta treatments don’t differ when they really do

power
For example, 80% power in a clinical trial means that the study has a 80% chance of ending up with a p value of less than 5% in a statistical test (i.e. a statistically significant treatment effect) if there really was an important difference (e.g. 10% versus 5% mortality) between treatments

p-value
A p value of 0.5 suggests that there is a 50-50 chance that the findings of the study are significant.

alpha
With respect to estimation problems , alpha refers to the likelihood that the true population parameter lies outside the confidence interval . Alpha is usually expressed as a proportion. Thus, if the confidence level is 95%, then alpha would equal 1 – 0.95 or 0.05.

beta
Beta is the probability of Type II error in any hypothesis test-incorrectly concluding no statistical significance. (1 – Beta is power).

Efficacy*
Efficacy can be defined as the performance of an intervention under ideal and controlled circumstances, whereas effectiveness refers to its performance under ‘real-world’ conditions.

Efficacy = ((rate in those who received old therapy) – (rate in those who received new therapy)/ rate in those ho received old therapy)

Number needed to treat*
The NNT is the average number of patients who need to be treated to prevent one additional bad outcome (e.g. the number of patients that need to be treated for one to benefit compared with a control in a clinical trial). It is defined as the inverse of the absolute risk reduction.

Phases of a drug trial (US/FDA)
Phase I – small studies that look at potential toxic effects and pharmacologic effects of new drugs
Phase II – A larger study to determine how effective the drug is and how safe it is
Phase III – Large-scale trial of the drug for effectiveness and safety (limited time)
Phase IV – Long-term surveillance of safety of a drug

Ethics-Core concepts including beneficence, respect for human dignity, and justice; informed consent; vulnerable groups
beneficence: Doctors and researchers should act in the
best interest of their patients

respect for human dignity: people’s ability to
make their own decisions

justice: People should equally benefit from
research

informed consent:

vulnerable groups: Additional protections for vulnerable
populations including pregnant women,
children, and prisoners.

Manuscript Review: purpose of study, independent variable, dependent variable, population studied, measurement methods, statistically significant results, conclusions, limitations-Be able to identify these from a peer-reviewed manuscript