
Upon, on top of
people or population
study of



The study of distribution and determinants of healthrelated states and events in specified populations, and the application of this study to control health problems 

Assumptions in epidemiology 

disease and ill health are not randomly distributed
factors that determine this distribution are knowable and mostly modifiable
modification = prevention and control of disease



D – concerned with distribution of disease by persons, places, and times. Without a hypothesis – hypothesis generating
A – Concerned with identifying disease determinants. Hypothesis testing. Casual inference is ultimate goal.



Hypothesis (the link)
Description
Analysis



Describes changing patterns of community health problems
Provides clues
Health services planning
Public health policy



Upon, on top of people or population study of 


prevent initial development of disease (risk factor reduction, immunization) 


Early detection of existing disease to reduce morbidity (screening for cancer, high blood pressure, etc.) 


Reduce the impact of acute worsening of disease 

Epi Triad of infectious disease 

Host – Susceptible Agent – infectious or noninfectious Environment – promotes exposure Middle – vehicle
*or vector organism for indirect transmission 

Triad of Noninfectious disease 

Host – genetics, stress, psychological factors Agent – behavior, diet, tobacco Environment – neighborhood, parks, food outlets 

Modes of disease transmission 

1. escape of the agent from a source or reservoir 2. Conveyance 3. entry into host 




1. common vehicle a. single exposure (potato salad) b. multiple exposures c. continuous exposure (telephone receiver) 2. vector (organism) 

Definition of “carrier” status 

A host “harbors” the infectious agent but appears not infected and there are no detectable signs or symptoms 

effect of nonapparent infection on disease control and disease statistics 

when the proportion with nonapparent infection is high, statistics will underestimate rates of infection and overstate overall disease severity 


habitual presence of a disease in a geographical area or population 


occurrence of a disease in a geographical area that is unusually frequent 




the first disease case that appears in the population 


the disease case that brings the group or population to the attention of public health personnel 


time between infection and appearance of signs or symptoms 


resistance of group or population to invasion and spread of infectious agent based on the immunity of a high proportion of the group 


from a single brief and essentially simultaneous source of exposure
– all of the cases usually result in a single incubation period 

propagated (“progressive”) 

transmission from one individual to another
– epidemic extends over many incubation periods 

general approach to outbreak investigation 

– verify the diagnosis – perform clinical studies – verify existence of epidemic – compare with past levels – person, place and time – calc. attack rates 


new cases rapidly occuring in a welldefined population over a short period of time —————————————— total number of people at risk during that same period of time 


*All times 100
new cases among contacts of initial cases occurring during a short period of time ————————————— (total number of people at risk during same period of time) – (initial cases) 


1. a diseased state or symptom 2. the incidence of disease: the rate of sickness 


a ratio where time is included in the denominator 


special type of proportion that includes time – represents the probability of disease in a defined population – the basic measure of disease occurence 


number of new cases of disease occurring during a defined period of time ———————————— number of people at risk for disease during defined period of time 


all observed throughout a defined period of time – closed cohort – denominator = sum of all people at risk 


all NOT observed for the full defined period of time – persontime units – denominator = sum of person observation time for each person at risk 

cumulative incidence formula 

number of new cases of disease occurring during a defined period of time ————————————– initial population at risk of the disease 

incidence density formula 

number of new cases of disease occurring during a defined period of time ————————————– persontime at risk for the disease 

Incidence rate based on group level data 

number of new cases of disease occurring during a defined period of time —————————————– number of people in population at risk at the midpoint of the period 


International Statistical Classification of Diseases and related health problems – used for coding and classifying mortality data from death certificates 


International Statistical Classification of Diseases – Clinical Modification – Used to code and classify disease morbidity data from inpatient and outpatient records 

problems determining who goes into the numerator 

1. what is the best case definition 2. detection (under or over diagnosis) 3. prevalent or incident (existing cases or new cases) 

problems determining who goes into the denominator 

1. undercounting of certain population subgroups 2. not everyone in the denominator is at risk for disease 3. population studied may have limited external applicability 

Limitations of hospital data 

1. admissions are selective 2. Hospital records are not designed for research 3. Populations at risk are not generally well defined 


number of existing cases AT a point in time ———————————– number of people in the popualtion AT that point in time 


Number of existing cases DURING a defined period of time ———————————————— number of people in the population DURING that defined period of time 

Relationship between incidence and prevalence 



number of all new events during a period of time ————————————— number of people at risk for these events during same period 

Problems with crude rates 

Crude rates can obscure the fact that subgroups in the population at risk exhibit significant differences in risk. Thus, rates may not be comparable across populations unless there is adjustment for populationspecific subgroup composition. 

Why look at mortality quantitatively? 

– can indicate disease severity and determine if treatment has become more effective over time serve as a surrogate for incidence when the disease is severe – pinpoint difference in the risk of dying from a disease between population subgroups 


number of people who die during a period of time ————————————— size of population during midpoint of the period 

allcause annual mortality formula 

total number of people who die from any cause during 1 year ————————————————– population at midpoint of that year 

disease specific annual mortaity 

number of people who die of a specific disease during 1 year ———————————————— population at risk at midpoint of that year 


* proportion, so a %
number of people who die of a disease ——————————– number of people with the disease 


*proportion, so a %
number of people who die of a disease —————————– total deaths 


1. age is associated with the disease or health state of interest 2. age distribution structure of the populations being compared are different 

Direct method for ageadjustment 

compute the overall mortality rate that would be expected if each population had the same age structure as the standard population 


– holds age constant – gives a more accurate picture of differences in rates across populations 


– fictional rates are calculated – suppresses details of how subgroups differ across the variable used for adjustment 


– used when age groupspecific rates are not available 


probability of death vs probability of survival 

Case fatality rate formula 

*rate, so a %
number of people who die of a disease —————————— number of people with the disease 


– good index of severity of a shortterm, acute disease – can measure benefit of a new therapy 


percentage of people with a disease who are alive at the end of a specified time interval 

5year survival rate formula 

*rate, so a %
number of cases of the disease who are alive 5 years later ——————————————– number of cases of the disease 

limitations of 5year survival rate 

– assumes the probability of surviving is equal throughout – does not use actual observed survival – cannot account for survival affect 

creator of the first life table 


limitations of classic life tables 

– assumes the probability of surviving is equal throughout the year or fixed interval 

Numerators in KaplanMeier life tables 

– an event is instantaneous – numerator = the number of people who died at that instant – denominator = the number of people alive up to that instant, including those who died – minus withdrawals that occurred before that instant 

KaplanMeier life table assumptions 

– no secular(temporal) changes in effectiveness of treatment or tendency toward survival over time – censoring must be independent of the probability of survival 


the length of time that half of the study population survives 


the sum of the individual survival times divided by the total number of people in the study population 


observed survival in people with the disease ———————————— expected survival without the disease 

cumulative incidence in an open cohort or closed cohort with incomplete followup 

number of cases of disease occurring during a defined period of time ———————————————— initial population at risk of the disease 


the probability of an event occurring – difficult to measure over time – applies to a population, not individuals – values range from 01, unitless – also expressed as a % – defined for a certain time period 


– special form of proportion that includes a specification of time – probability of disease in a defined population over time – the basic measure of disease occurrence – a risk in disguise – the risk of acquiring a disease, and death from a disease is the best approx. by a measure of its incidence 

Which measures of disease occurrence below gives you the most accurate information about risk? 



Risk = (incidence rate * time) 


Risk = 1 – [exp ^(rate * time)] 

Canons of Inductive Reasoning – Method of Difference 

examines differences among groups for clues as to why disease rates or other health problems vary across groups 

Canons of Inductive Reasoning – Method of Agreement 

looks for commonality in groups that manifest the same health problem 

Canons of Inductive Reasoning – Method of Concomitant Variation 

traces how exposure to a hazard varies in relation to disease or other health problem 




