POPM 3240 FINAL

Definition of Epidemiology

Study of frequency, distribution and determinants of health and disease in populations. Disease distribution is not random.

primary prevention

preventing the initial development of a disease
ex. immunization, condoms, lab coat

Secondary Prevention

Early detection of an existing disease
ex. cancer screening, mammogram

Tertiary Prevention

Reducing impact of the disease. Treating signs and symptoms
ex. stroke rehabilitation, tylenol, losing weight

Who developed a hand-washing policy to make sure medical students didn't transmit childbed fever to pregnant women

Semmelweis

Who discovered that cowpox is a way of preventing small pox?

Jenner

Who theorized that cholera comes from contaminated water?

Snow

Who found the cause and cure of scurvy aboard ships?

Lind

Who observed that chimney sweepers were more likely to get carcinogenic tumours?

Pott

agents

bacteria, parasites, viruses, fungi, poison, alcohol, smoke

environmental factors

temperature, housing, water, air pollution, altitude

host factors

age, sex, genetic profile, family history

horizontal transmission

transmission between members of the same species

vertical transmission

transmission between mother to embryo/fetus/baby

direct transmission

person-to-person contact

indirect transmission

ingestion, aerial, inanimate object, latrogenic

vectors

- indirect route of transmission from animal to human
- 75% of all diseases
- zoonotic (between animals and humans)

Subclinical Disease

when you don't show symptoms of illness
ex. cases of polio

Clinical Disease

shows symptoms
- preclinical aren't yet apparent but will become clinical

Persistent Chronic Disease

Infection persists for years

Latent Disease

infection w no active multiplication of the agent

Infected Vs Diseased

infected: host is invaded by organisms and hosts immune system responds
diseased: infection causes clinical signs and symptoms

Carrier

harbours the organism but isn't infected
- can still infect others but at a lower risk than other infections

endemic, epidemic, pandemic

Endemic: presence of a disease within a given area (flu in ON)
Epidemic: occurence in a region of similar illnesses (malaria in canada)
Pandemic: worldwide epidemic (SARS, small pox)

Incubation Period

from the time of infection to the time of onset clinical illness
- you don't show symptoms of disease during this time

Latent period

time between exposure to the cause of the disease and being able to detect the disease state once its underway

Development of a Disease

1. susceptibility
2. pre-symptomatic disease
3. clinical disease
4. disability or recovery

Innate Immunity

- born with this immunity
- immediate response to invaders by WBCs, skin, acid

Acquired Immunity

naturally occurs when a person is exposed to a live pathogen and primary immune response creates antigens against it

herd immunity

- balance between susceptible and those who are immune
- resistant individuals protect the entire population
- 3 qualifiers: infectious agent is restricted to 1 host species, infection has to induce solid immunity, random mixing of individuals, direct transmission

Error vs Bias

error: due to sample or chance
bias: systemic error due to flaws in study design. affects validity.
types of bias: selection, information, confounding

Association

sequence of studies: clinical observations, available data, case-control studies, cohort studies, randomized trials
1. find out if theres an association
2. if theres an association, find out if the association is causal or confounding

Direct Causation

factors directly causes a disease without any intermediates

Indirect Causation

factor indirectly causes a disease through intermediate steps

Necessary or Sufficient

- Necessary and Sufficient: without the factor the disease never develops
- Necessary but not Sufficient: each of the multiple factors is necessary but none alone are sufficient in developing the disease
- Sufficient but not Necessary: factors alone can produce the disease and so can others
- Neither Sufficient or Necessary: a factor by itself is not sufficient or necessary to produce the disease

selection bias

- case, controls, exposed, non-exposed individuals were selected so that an apparent association is observed
- types of selection bias: exclusion, sampling, non-response, confirmation, loss-follow-up, healthy worker effect

Information Bias

- when the means for getting information about subjects is inadequate and the disease outcomes are incorrect
- surrogate interviews, recall bias, reporting bias, wish bias contribute

missclassification bias

when an individual is classified into the wrong category

confounding bias

determining if factor A is the cause of the disease B, factor X is a confounder if:
1. factor X is a known risk factor for disease B, be associated with the exposure
2. factor X is associated with factor A but it isn't a result of factor A. Be associated with the outcome
3. not be a result/caused by the exposure

Screening Tests

population approach on 'healthy' individuals to detect subclinical diseases. indicates diseases in their early days

diagnostic testing

used on sick sick individuals with clinical disease to confirm, classify, inform treatment, provide prognosis

Validity and Reliability on a Graph

- narrow curve indicates its reliable
- clustering around the true value (middle) means its valid

Sensitivity (Sn)

- ability of a test to correctly identify those who have the disease
- a/(a+c)
- as sensitivity increases, the # of false negatives decreases

Specificity (Sp)

- ability of a test to correctly identify those who do not have the disease
- d/(b+d)
- as specificity the # of false positives decreases

SNOUT

if the test is highly sensitive (SN) and you get a -ve result you can confidently rule the disease OUT

SPIN

if the test is highly specific (SP) and you get a +ve, you can confidently rule the disease IN

Type I error is a ________

false positive

Type II error is a ________

false negative

n =

a + b + c + d

True Prevalence

- actual proportion of disease within a population determined by the results of the gold standard test
- TP = (a+c)/n

Apparent Prevalence

- observed proportion of the disease determined by the result of the new test
- AP = (a+b)/n

Positive Predictive Value (PPV)

- The probability that a person with a positive test result is truly positive
- PPV = a / (a+b)
- PPVs over 90% are needed
- 1 - PPV = prevalence

Negative Predictive Value (NPV)

- proportion of those that test negative that are truly negative
- d/(c+d)

Changing Prevalence Affects Predictive Value

- if prevalence values increase, PPV increases, NPV decreases
- changing prevalence values does not affect Sp or Sn

Sequential Testing (2 stage)

- less invasive, less expensive, less uncomfortable
- 2nd test is more expensive and has greater Sn and Sp
- reduces the amount of false positives, but miss false negatives
- you gain net Sp, lose Sn
- individual must test +ve on both tests to be considered diseased

Simultaneous/Parallel Testing

- multiple tests given at the same time. Results are combined for interpretation.
- only if both results are -ve is the individual considered negative
- good at preventing false negatives
- prevent false-negatives
- gain net Sn
- more expensive and invasive

Reliability

the ability of test accurate results when repeated over time

Intrasubject Variation

- change in an individual over time. related to aging, attitudes and taste buds

intraobserver variation

different interpretations by the same person that can be influenced by training, context, experience

interobserver variation

this is related to 2+ people disagreeing on the interpretation of a test

chance agreement

situation where 2 individuals make the same assessment due to chance alone rather than true agreement

Kappa Statistic

- used to measure agreement with multiple observers or the agreement between 2 tests
- beyond chance
> 0.8 means excellent
- (observed agreement - expected agreement)/(1 - expected agreement)

Mortality

refers to death

Morbidity

diseased state

incidence rate per 1000

- defined as the new # of cases occurring in the population in a period of time / # of persons who are at risk of developing the disease during that period of time

attack rate

time associated with the denominator of the incidence

prevalence

= # of affected persons present in the population at a specific time /
# of persons in the population
- this measures the existing cases
- = incidence x duration
- "being diseased"

point prevalence

prevalence of disease at a point in time

period prevalence

how many people have had the disease at any point during a certain time period.

Incidence risk

Measure of new cases of a disease or injury occurring in a population during a specified period of time.
- new cases only in the numerator
- "probability of becoming diseased"
- = (# of new cases)/(NAR - 1/2 withdrawals)
- (# of new cases)/(NAR x time period)
- true rate

NAR

initial number at risk

Crude Morbidity Risk

(# developing any disease/dying) / (NAR - 1/2 withdrawals)
- true rate

crude morbidity rate

(# developing disease)/(NARavg x ITC)
- true rate

Cause-specific mortality rate

(# dying from specific disease) / (NARavg x ITC)
- true rate

relative risk

- (risk of disease or outcome in the exposed group) / (risk of disease or outcome in the non-exposed group)
- RR = [a / (a + b)] / [c / (c + d)]
- RR=1 means theres no association between the exposure and the outcome

Odds Ratio

(odds of disease in the exposed group)/(odds of disease in the non-exposed group)
- OR = (a*d)/(b*c)

Relative Risk Statement

If RR = 8.48. The risk of brain tumours among those who use cell phones is 8.48 times the risk in those who don't use cell phones

Odds ratio statement

if OR = 18.8 the odds of brain tumours among cell phone users are 18.8 times the odds of brain tumours among people who don't use their cell phones

Incidence Rate Ratio

- IRR = (IR in the exposed group)/(IR in non-exposed group)

Risk Difference

- (risk/rate in exposed) - (risk/rate in non-exposed)
- RD = [ a / (a + b)] - [c / (c + d)]
- if RD = 0 the exposure has no effect on the outcome
- if RD > 0 - possibly causal
- if RD < 0 - possibly protective

Attributable Proportion

- proportion of risk of the outcome in the exposed group that is attributable to that specific exposure
= RD / (risk/rate of outcome in exposed)
= RD / [ a / (a+b)]

Population Attainable Risk/Rate

= [(prevalence of outcome in the population) - (prevalence in non-exposed group)] / (prevalence of outcome in the population)

target population

population where it might be possible to extrapolate results

source population

Population from which study subjects are drawn.

sampling frame

list of individuals in the source populations where subject are drawn from

sampling unit

actual unit of measurement

sampled population

group of sampling units that was actually selected from the sampling frame

Internal Validity

- degree to which the observed findings in a study lead to correct inferences about the outcome of interest in the source population
- ability of a study to provide accurate info

External Validity

- degree to which inferences drawn from the suited can be generalized to the broader pop. of interest

Non Probability Sampling

- choosing study subjects in a non-random way. every member doesn't have an equal chance of getting chosen
- ex. convenience, judgment and purposive sampling
- easy, cheap, biased, prevents extrapolation

Convenience sampling

sampling where units are chosen because they are easy to get

judgment sampling

sampling where the investigator chooses what he or she deems to be units that are representative of the population

purposive sampling

- sampling where units are chosen on purpose because of their exposure and disease status

Probability Sampling

- random process so all individuals have a non-zero probability of being selected for the study
- representative sample, time-consuming, requires large samples
ex. simple, systemic, stratified sampling

simple random sampling

fixed % of the source population being chosen using a random process. everyone has an equal chance of being included

system random sampling

- uses sampling interval
- 1st person is chosen using random process then the sampling interval picks the next nth person

stratified random sampling

break down the sampling frame into strata, sub categories

cluster sampling

sampling unit is group of subjects with characteristics in common
ex. a household, herd of animals

Variance

- p * q
- p is the proportion of individuals that have a disease
- q is the proportion of individuals that don't have the disease

Level of Confidence (Z)

level of confidence that the estimate will obtain the true value