Session 2: Descriptive Epidemiology

Epidemiology

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

What does descriptive epidemiology tell us?

- What is the problem/frequency

- Who is affected

- Where is affected

- When does it occur

What does analytical epidemiology tell us?

Why it occurs in this specific population

Why is descriptive epidemiology important?

1. Evidence based medicine

2. Research - understanding

3. Clinicians = reports, audits, exams

Why is epidemiology important, specifically to qualifying doctors?

- Writing reports = Special Selected Components (SSCs)

- Audits (compulsory part of foundation training)

- Postgraduate exams

Types of data

Categorical (discrete) and continuous

Categorical (discrete) data, with examples of this data

Data which can be divided into groups

Nominal e.g., male/female

Ordinal e.g., mild, moderate, severe

Interval e.g., distance between measures has meaning

Continuous data, with examples

Continuous data can take on any value

Interval e.g., distance between measures on scale

Ratio e.g., distance/ratio between measurements defined

Example of a frequency distribution of continuous data

Histogram

Example of a frequency distribution of discrete (categorical) data

Bar chart

Three factors which summarise a frequency distribution

1) Shape

2) Location or central tendency

3) Spread

What shape is found in this histogram?

Unimodal

Skewed to the left

What shape is found in this histogram?

Unimodal

Skewed to the right

What shape is found in this histogram?

Unimodal: symmetrical

Normal distribution

Normal distribution

Symmetrical bell-shaped curve

What shape is found in this histogram?

Bimodal: symmetrical

Causes of outliers

• Data entry errors

• Measurement errors

• Experimental errors

• Intentional

• Data processing errors

• Sampling errors

• Natural

Location or central tendency

Tell us about the focal point or central point of the data.

This includes the mean, median, and mode.

Mode

Value or group of values which occur most often

Median

Middle value after arranged in order of size (ascending order)

Mean

Arithmetic average of observations

Relationship of mean to median in a symmetrical distribution...

Mean ≈ Median

Relationship of mean to median in a skewed distribution (right)

Mean > median

Relationship of mean to median in a skewed distribution (left)

Mean < median

What is the mode of this frequency distribution

The most common score which was 66 (highest frequency)

What is the median of this frequency distribution

Median = [Total number of participants + 1] / 2

n = 229

[(229 + 1)]/2 = 115th value = 64

What is the mean of this frequency distribution

Mean = sum of all data points / number of data points

Mean = 14, 198 / 229 = 62

Two ways of assessing the spread of data in a frequency distribution

1) Range = difference between highest and lowest value

2) Interquartile range = difference between first and third quartile

Type of plot used to examine the spread of a frequency distribution

Box and Whisker plot

How to calculate the interquartile range

IQR = Q3 - Q1

How to calculate variance (s^2) of a data set and what it means...

Variance is a measure of the variation shown by a set of observations, defined by the following formula

Standard deviation (s) formula and what it means...

Standard deviation (s) is a summary of how widely dispersed the values are around the mean, defined by the following formula

Why are scatter plots useful?

1) Show relationship between two continuous variables

2) Form = linear or non-linear

3) Strength = strong or weak

4) Positive or negative correlation (or zero)

Categorical data - examples of how it can be presented and what it shows

How it can be presented

Frequency table

Bar chart

Pie chart

What it shows

Percentages

Modal (most frequent)

Discrete data - examples of how it can be presented and what it shows

How it can be presented

Frequency table

Bar chart

What it shows

Median, SD (symmetrical)

Median and IQR (skewed)

Continuous data - examples of how it can be presented and what it shows

How it can be presented

Frequency table

Histogram

Box and whisker plot

What it shows

Mean and SD (symmetrical)

Mean and IQR (skewed)

Categorical data - examples of how it can be presented and what it shows

How it can be presented

Frequency table

Bar chart

Pie chart

What it shows

Percentages

Modal

Some examples of measures of frequency

Ratio = division of two unrelated numbers

Proportion = division of two related numbers

Rate = measure of frequency of occurrence 'per unit time'

Ratio

Division of two unrelated numbers

Proportion

Division of two related numbers

Rate

Measure of frequency of occurrence 'per unit time'

What is incidence?

The number of new events e.g., new cases of disease in defined population within specific period of time

What is prevalence?

The number of cases of disease in a given population at a designated time

Incidence rate

Incidence risk

Point prevalence

Point prevalence is a variation which represents the number of persons who were a case at any time during a specified (short) period as a proportion of the total number of persons in that population

Why might the prevalence of a condition change over time?

1) Chance

2) Ascertainment

3) Demography

4) Treatment effects

5) True changes in prevalence

Epidemiologist's Bathtub

Relationship between incidence, prevalence, and mortality

Prevalence = (incidence rate) x (average duration of disease)

Incidence = prevalence / duration

Duration = prevalence / incidence

Mortality rate is an example of an ___ rate

Incidence

Three measures of mortality

1) Infant Mortality Rate (IMR)

2) Crude Death Rate (CDR)

3) Age-Specific Death Rate (ASDR)

Infant Mortality Rate (IMR)

A figure that describes the number of babies that die within the first year of their lives in a given population.

Crude Death Rate (CDR)

The total number of deaths in a year for every 1,000 people alive in the society.

Age-Specific Death Rate (ASDR)

The number of deaths to persons in a specific age group per 1,000 persons in that age group

Risk

Probability or likelihood of an outcome

Association

Correlation between exposure and outcome

The strength of an association between a putative Risk Factor and an outcome can be measured in terms of...

- Relative risk (RR)

- Odds ratio (OR)

- Attributable risk (AR)

- Population attributable risk (PAR)

- PAR%

Measures of Absolute Risk

Incidence Risk (IR) =

number of new cases of disease in specified time period / number of individuals at risk during same time period

Measures of Relative Risk

Risk Ratio (RR) =

A comparison of the risk of some health-related event such as disease or death in two groups.

Risk ratio (RR) > 1

Positive association

Risk ratio (RR) < 1

Negative association

Epidemiology is the science which informs ___ ___

Public health

Incidence is defined as...

Measure of new cases or events in a population within specified time

Prevalence is defined as...

Measure of the burden of disease in a specified population at a particular time (this is not a rate!)