slide deck 5, exam 2

time place and person

time

aspects of when the disease occurs/occurred

place

conditions of the location where the disease occurs/occurred

person

characteristics of the people who have/had the disease

why is using person/place/time good for epi analysis

it can identify trends of mortality and morbidity

identification of clusters (grouping) of morbidity and mortality

what do clusters help

to identify etiologic (cause of disease) agents and risk facts in respect of time, place and person

what did john snow do

he developed key descriptive techniques to investigate the source of an 1850s cholera outbreak that was said to be a miasma caused by bad air

what was snows method of investigating place

he did mapping as an epi technique to make fatal cases on a map to show a cluster in a neighborhood

what was snows method of investigating person factors

he used surveys, stats and data analysis to determine that the common person factor was where the cases got their water (the broad street pump)

secular trends

changes in health related states or events that take place over more than one year

short term trends

changes in health related states or events that take place over hours, days, weeks or months

seasonal trends

changes in health related states or events that take place during a particular season of the year

cyclic trends

changes in health states or events that take place in a recurrent manner over time

example is with measles, it comes back up every few years

examples of place

landforms, bodies of water, climate, ecosystems

examples of social determinants of health

employment, housing, education, access to healthcare, discrimination, quality of care

examples of person

age, sex, gender, race, occupation, educational status, family structure

examples of political determinants of health

income, debt, transportation, language, hunger, health coverage, stress

what does cluster analysis allow for

can help to define target populations for prevention strategies

time clustering

health related state or event occurred more frequently during a given time frame than it did in other time frames

example is the opioid overdose that sparked during 3-4pm

place clustering

health related state or event occurred more frequently in a particular place that it did in other places

example is cluster in southern ohio with drug overdose deaths

person clustering

health related state or event occurred more frequently in a group of people who share a certain person characteristic than it did in other groups

example is with black women who are 40 and over had the greatest maternal mortality versus other age and race groups

example of how clustering the wrong thing can give the wrong results

a study used clustering of people who died from typhus and tracked them back to where they lived

another cluster was the people who died from typhus and where they worked

the cluster came from a warehouse that had a mouse problem

absolute change (subtract)

it is the difference between two values

example, a study found an increase of 250 cases of tuberculosis between january 2024 and july 2024

if the absolute change is positive,

there has been an overall increase in deaths/cases

if the absolute change is negative,

there has been an overall decrease in deaths/cases

relative change (divide)

the more recent value divided by the older value

the example is there was 3 times more cases of chlamydia in july 2024 compared to july 2023 and there were ¼ as many cases of strep throat in july 2024 compared to january 2024

if the relative change is greater than 1,

there has been an increase of x fold

2 fold= double

if the relative change is less than 1,

there has been a decrease of x fold

0.5 fold= half

percent change

is a combination of absolute (subtract) and relative (divide)

is the percent change is positive

there has been a percent increase

ex. 100% increase means the value has doubled

if the percent change is negative,

there has been a percent decrease

ex. 100% decrease means the value is gone, not halved