Basic Epidemiology for Community Health Assessment

1
Basic Epidemiology for Community Health Assessment
Nelson Adekoya, DrPH Centers for Disease Control
and Prevention
2
Outline

• Definitions
• Fundamentals of Epidemiology
• Data
• Death Certificate
• Determine the Leading Causes of Death
• Data Presentation Tips
• Questions Self Quiz

2
3

• Define Health
• Define Public Health
• Define Epidemiology
• What are the goals of Public Health?

3
4

• Public health is the science of preventing
  diseases, prolonging life, and promoting health
  thru organized community effort

4
5

• Goals of Public Health?
• Prevent or control disease, disability and injury
• Improve quality of life for residents of state or
  community

5
6

• Define Epidemiology (write down your definition,
  check against definition in Part 2)

6
7
Part I
Epidemiology and Data are inseparable. So, why do
we need data? ..write down your answers

7
8

• Why are Data Needed?
• to determine major health problems (needs
  assessment, program development)
• to identify where to focus efforts and resources
  (asset mapping)
• to determine progress in solving health problems
  (measuring health indicators, outcomes, or
  Healthy People Objectives)
• to conduct research and grant applications

8
9
Succinctly, data are needed ? to assess the
health of a community or population ? to search
for causes of disease, injury and disability ?
to plan programs to meet community needs and ?
to measure progress in prevention and control
efforts.
9
10

• According to the National Center of Vital Health
  Statistics
• Data is required for a process that involves
  the community in identifying problems, setting
  priorities, developing an action plan, measuring
  progress, deciding whether the actions are
  effective, modifying the actions if necessary,
  and reevaluating the community's problems and
  priorities.

10
11
Types of Data Needed

• Depend on
• What is the program of interest?
• What is the focus of the program?
• What is the purpose of the program?

11
12

• Primary data (data collected directly by the
  organization).
• Secondary data (data collected by someone outside
  their own program or agency, to measure the
  outcomes of interest).

12
13
Locating and Accessing Secondary Data

• Online Secondary Data access available by such
  links from the New Hampshire Health Data
  Inventory (HDI) at www.nhhealthdata.org
• Some web sites provide data tables that can be
  used to perform calculations or create charts.
• Other web sites allow users to obtain raw data
  files, which can be used to create aggregate
  tables, generate statistics, and perform
  calculations.

13
14
Web Sites for Secondary Data Sources

• CDC WONDER (http//wonder.cdc.gov/) has access
  to a variety of public health information,
  including links to environmental and
  disease-specific data.
• CDC WISQARS (http//www.cdc.gov/ncipc/wisqars/d
  efault.htm) WISQARSTM (Web-based Injury
  Statistics Query and Reporting System) is an
  interactive database system that provides
  customized reports of injury-related data.
• US Census (www.census.gov) makes some of its
  data about the US population available for
  download.

14
15

• What is NM website address to access health data?

15
16
Importance of Data Quality

• Relevance
• Accuracy
• Timeliness
• Accessibility
• Interpretability
• Coherence

16
17
Relevance

• The relevance of statistical information reflects
  the degree to which it meets the real needs of
  clients.

17
18
Accuracy

• The accuracy of statistical information is the
  degree to which the information correctly
  describes the phenomena it was designed to
  measure.
• It may also be described in terms of the major
  sources of error that potentially cause
  inaccuracy (e.g., coverage, sampling,
  non-response, response).

18
19
Timeliness

• The timeliness of statistical information refers
  to the delay between the reference point (or the
  end of the reference period) to which the
  information pertains, and the date on which the
  information becomes available.

19
20
Accessibility

• The accessibility of statistical information
  refers to the ease with which it can be obtained
  from the Agency.

20
21
Interpretability

• The interpretability of statistical information
  reflects the availability of the supplementary
  information and metadata necessary to interpret
  and use it appropriately. This information
  normally includes the methodology of data
  collection and processing, and indications or
  measures of the accuracy of the statistical
  information.

21
22
Coherence

• The coherence of statistical information reflects
  the degree to which it can be successfully
  brought together with other statistical
  information within a broad analytic framework and
  over time.

22
23
Other aspects of data quality to consider

• Able to drive decision-making and behavior
• Can the outcome be monitored over time?

23
24
Fundamentals of Epidemiology
Part I1
24
25
Definition of Epidemiology

• The study of the distribution and determinants
  
• of health-related states or events in
  specified
• populations in a specified time period, and
  the
• application of findings to control of health
• problems.

25
26
Purpose of Epidemiology
? To provide a basis for developing disease
control and prevention measures for groups
at risk.
26
27
Descriptive Epidemiology

• ? Examine the distribution of disease in a
  population
• and observe the basic features of its
  distribution.
• Answer questions about what, who, when and
• where people get ill, injured, or disabled.

27
28
Analytic Epidemiology

• ? Test a hypothesis about the cause of disease by
• studying how exposures relate to the disease.
• Answer questions about how and why people get
• ill, injured, or disabled.

28
29
Uses of Epidemiology

• ? Determine the primary agent or causative
  factors
• ? Determine the characteristics of the agent
• ? Define the mode of transmission, and
  contributing factors
• ? Identify geographic patterns
• ? Describe the natural course of disease,
  disability, injury
• and death
• ? Help planning and developing health services
  and programs
• ? Provide administrative and planning data

29
30
Epidemiologists look For

• ? Person
• ? Time
• ? Place
• To answer questions about what, who, when and
  where, how and why people get ill, injured, or
  disabled.

30
31
31
32
Place ?Geographic place ?Urban, suburban,
rural ?Climate ?Geology ?Population
density ?Economic development ?Cultural
norm ?Medical practice ?Nutritional practices
32
33
34
34
35
35
36
36
37
37
DATA
Part III
38
38
?Quantitative Data ?Qualitative Data
39
39

• Quantitative Health DATA can focus on
• individuals, or
• entire populations.

40
40
Population-based Data versus
Individualized Data ?Example of
individualized data is a patients medical
record. ?Each record is devoted exclusively to
one person and contains information about
his or her unique illnesses, injuries,
behaviors, etc. ?The data are used primarily to
improve the health of that one individual.
41
41
Population-based Data
versus Individualized Data
(cont) ?In public health, our focus is primarily
on populations (e.g. communities, cities,
counties, states). ?Population-based data
tells us about the overall health of that
population.
42
42
Major Types of Health Data available for
analysis ? Health outcome / Health Status
Data ? Risk factor Data ? Resource
Data ? Demographic Data
43
43
Quantitative DATA ?
Measurable and tangible ? Provides answer
regarding what, who, when, and where of
health-related events.
44
44
Simply, what we are doing in
Quantitative DATA are ? Counting of
people, behaviors, conditions, or other
discrete events ? Classifying those events into
categories ? Using math and statistics to
answer questions.
45
45
Quantitative DATA Examples
(cont) ?Using numbers of deaths to identify
leading causes of death (What) ?Using
numbers of smokers and nonsmokers by gender
to determine whether men are more likely to
smoke than women (Who) ?Keeping track of the
number of people with flu can identify the
beginning of the flu season (When)
46
46
Quantitative DATA Example
(cont) ?Comparing the proportion of women who
began prenatal care after the first
trimester in various counties will provide
an indication of where access to prenatal
services may be a problem.
47
47
Qualitative
DATA ?Qualitative data can be used to explain
the why and the how of health-related events.
?Qualitative data involve observing people in
selected places and listening to discover how
they feel and why they might feel that
way.
47
48
Examples of Qualitative DATA ? A
focus group of teenage girls could provide
valuable insights concerning why they do or
dont use contraceptives. ? A visit to a
local clinic might indicate how people might
feel as they enter the waiting area.
48
49
What Analysis to Request for Quantitative Data?

• Measures of central tendency
• ? Mean is the arithmetic average of the values
  in the data
• ? Median is the middle value
• ? Mode is the most commonly occurring value

49
50

• Any Questions???

50
51
Death Certificate Data
Part IV
51
52
Introduction to Death Certificate Data
Part IV Section One
52
53

• Primary Source of Death / Mortality DATA
• Death / mortality is one of the primary measures
• of a populations health.
• Death certificate is the primary source of death
  or
• mortality.

53
54
What data are on a typical death certificate? ?
Information about the characteristics of the
decedent, ? Information about the circumstances
of death (e.g. time, date, and place), and ?
Specifics about the causes of death.
54
55
What a typical death certificate
looks like (Figure 1 2)
55
56
Information about the decedent includes ?Sex
(2), Age (5a) ?Birthplace (7) ?Marital Status
(10) ?Residence (13 a-f) ?Hispanic Origin (14)
and Race (15) ?Occupation (12a) and Education
(16).
56
57
Information about the cause of death includes
? Date of Death (3) ? Place of Death (9a) ?
Immediate Cause and Underlying Cause of Death
(27, Part I) and Other Significant Conditions
(27, Part II) ? Manner of Death (29) and
additional details for deaths due to
accidents, suicide, or homicide (30 a-f).
57
58
ICD and Diseases / Conditions

• At the state vital statistics office,
  information from death
• certificates is entered into electronic
  records.
• All of the diseases and conditions reported on
  the death
• certificate are translated from text into
  medical codes
• using the International Classification of
  Diseases (ICD)

58
59

• What are ICD codes ?
• The International Classification of Diseases
  (ICD), published by the World Health
  Organization (WHO), establishes a structure for
  translating the entries on the death certificate
  into a statistical classification.
• From 1979 to 1998, death certificates were coded
• according to ICD-9. Beginning in 1999, death
  certificates are being coded according to the
  Tenth Revision (ICD-10) .

59
60
Part IV Section Two

• How to Determine the Leading Causes of
• Death by analyzing death data from a death
• certificate

60
61
Death Data / Mortality DATA Outline ?
Death certificate (Figure 1 ) ? Data
collection based on a death certificate ?
Discuss limitations of those data
61
62
USES of Deaths Data / Mortality Data ? Calculate
general death rates ? Age-specific death rates
? Cause-specific death rates ? Sex-specific
death rates ? Describe the difference between
death counts and death rates ? Identify
leading causes of death
62
63
USES of Deaths Data / Mortality Data (cont) ?
Use rates to examine differences in causes of
death by sex ? Analyze trends in rates and
changes in underlying causes over time ?
Determine effective ways to present data
63
64
STANDARD MORTALITY MEASURES

• What is a count?
• the most basic unit of data.
• ? the number of hospital visits, the number of
  injuries,
• the number of live-born twins, the number of
  mothers who
• deliver twins, the number of deaths, and
• the number of deaths due to AIDS, etc.
• ?Counts are also useful when comparing numbers
  for the
• same population group.

64
65

• What are leading causes of death?
• Leading causes of death are diseases identified
  as being of
• public health importance based upon the
  burden of the
• disease within a population.
• The importance of each disease is ranked
  relative to other
• important diseases.
• A standard procedure for ranking leading causes
  of death
• was adopted by the National Center for
  Health Statistics
• decades ago in which eligible causes are
  ranked according
• to the number of deaths (not rates).

65
66
Eligible Causes (ICD-9)

• The eligible causes include the 37 rankable
  causes from
• the List of 72 Selected Causes of Death
  (those causes
• marked by an asterisk on Figure 6), along
  with HIV
• infection (added with 1987 data) and
  Alzheimers
• Disease (added with 1994 data).
• Link to ICD-10 is provided below for revised
  ranking http//www.cdc.gov/nchs/datawh/nchsdefs/co
  drank.htm

66
67
Counts and Leading Causes Of Death

• ? Counts are used to identify leading causes of
  death.
• ? The leading cause of death is the cause with
  the greatest
• number of deaths.
• ? Counts can be used to identify the leading
  causes of death
• in a single population, their value is
  limited.
• Counts do not help you compare the risk of death
  in one
• population with the risk in another
  population.

67
68

• Rate
• A rate is the most common way to measure the
  occurrence
• of an event in a population.
• ? The general formula for a rate
• Number of events occurring during a given
  period x 100000
• Rate
• Population at risk during the
  same time period

68
69
Three Consistent Characteristics of a RATE
? The counts in the numerator and denominator
should cover the same time period. ? The
persons who experienced the events in the
numerator should all be included in the
denominator. ? The persons in the denominator
should be at risk for the event in the
numerator. In other words, it should be possible
for them to experience the event.
69
70
Proportion

• The number of observations with a given
  characteristic (a) divided by the total number of
  observations in a given group (a b) 26. That
  is, proportion a/(ab).

70
71
Ratio

• Ratio simply relates a figure to another. For
  example, there are 36 females and 18 males in a
  restaurant. The ratio of females to males is
  3618 or 36/182.

71
72
Adjusted Rate

• An expression of the predicted number of health
  events within a standard population defined by
  one or more variables not under study and used to
  control for effects medicated by such variables.

72
73
Specific Rate

• An expression of the observed number of health
  events within a defined subgroup or stratum of
  the population at risk within a predefined time
  period (e.g., age-specific death rate of 1.2
  deaths per 1,000 persons aged 10 to 19).

73
74
Crude Rate

• An expression of the observed number of health
  events per unit of the population at
• risk in a defined time period (e.g., crude
  mortality rate of 5.7 deaths per 100,000
• persons in 1985).

74
75
What is a death rate?

• A death rate is a measure of the occurrence of
  death
• in a defined population during a specified
  time interval.

Number of deaths during a
given time period Death rate

x 1,000,000
Number of people in the population
in which the deaths occurred
75
76
What are some specific types of death rates? ?
The death rate discussed above is often called
the crude death rate. ? It measures the
frequency of deaths from all causes in an
entire population.
76
77
What are some specific types of death
rates? Example Crude Death Rate The crude
death rate for the United States in 1990 would
be Number of deaths in the
U.S. in 1990
x
100,000 U.S.
population in 1990
77
78
Example Age-specific Death
Rate ? The death rate for a particular age
group. ? The death rate for older adults, age
65-80, in NM in 1983 would be
Number of deaths among 65-80 year olds in NM in
1983
x 100,000
Number of 65-80 year olds in NM, 1983
78
79

• Sex-specific death rate
• The death rate for a particular sex, either
  males or
• females.
• Formula
• Number of deaths among males (or females)
• during a given time period
• 
  x 100,000
• Population of males (or
  females)
• during the time period

79
80

• Example Sex-specific Death Rate
• The death rate for females in NM in 1991 would
  be
• Number of deaths among females
• in NM in 1991
• 
  x 100,000
• Number of females in NM, 1991

80
81

• Cause-specific death rate
• The death rate from a specific cause for a
  population.
• The sum of all cause specific mortality rates for
  a population equals the total mortality rate for
  that population.
• Formula
• Number of deaths from a
  specific cause
• during a
  given time period
• 
  x 100,000
• Population during the
  time period

81
82

• Example Cause-specific Death Rate
• The death rate from homicides in NM, 1990-1994
• would be
• Number of deaths from homicide
• in NM, 1990-1994
• 
  x 100,000
• Population in NM, 1990-1994

82
83

• Example Age-sex-cause-specific Death Rate
• The death rate from homicides for 65-80 year old
  females in NM, 1990-1994 would be
• Number of deaths from homicide among 65-80
  
• year old females in NM, 1990-1994
• 
  x 100,000
• Number of 65-80 year old females
• in NM, 1990-1994

83
84

• What happens when the numbers are small?
• Occasionally, your numerators will be so small
  that the
• resulting rates become unstable or
  unreliable.
• One rule of thumb is to have at least 20 deaths
  in each
• cause category that you are analyzing. You can
• accomplish this by combining several years of
  data.

84
85

• How are rates used to make comparisons?
• The real value of death (mortality) rates is that
  they enable us to compare the risk of death
  between different causes, different age/sex
  groups, different time periods, etc.

85
86

• How are rates used to make comparisons?
• Quite often, these comparisons are expressed in
  terms of a rate ratio.
• The formula for calculating a rate ratio
• X Y or X/Y
• where
• X death rate in one
  population
• Y death rate in another
  population

86
87

• Example Comparing Death Rates
• Suppose that in 1993 the overall death rate for
  males, ages
• 15-24 years, had been 141.8 per 100,000.
• Further suppose that the death rate for males,
  ages 25-44,
• that same year had been 258.3 per 100,000.
• ?Then the ratio of the death rate for younger
  males to the
• death rate for older males would be
• 141.8
  258.3 0.55
• ?This means that the death rate for younger males
  is half
• the death rate for older males.

87
88

• What are some common mistakes
• in calculating rates?
• When calculating a rate, always use the same
  time
• period in the numerator and denominator.
• When the numerator includes deaths for more
  than one
• year, the denominator must include the
  population for
• the same years.
• If the deaths in the numerator are added
  together
• for 5 years, the population in the
  denominator must also
• be added together for the same 5 years.

88
89

• What are some common mistakes in
• calculating rates (cont)?
• Be sure to use the same population, age group,
  and/or sex in both the numerator and denominator.
  
• When calculating cause-specific rates, the
  denominator should include the entire population
  while the numerator should include only deaths
  due to one specific cause.
• Be sure to indicate what scale was used usually
  results are per 1,000 or per 100,000.

89
90

• How are percents helpful?
• Use a percent () to describe the proportion of
  deaths due to a specific cause (that is, those
  deaths assigned a specific ICD code).
• The formula for calculating this percent is
• Number of deaths from a specific
  cause
• during a given time
  period
• 
  x 100
• Number of deaths due to all
  causes
• during the time
  period

90
91
Example Using
Percents Suppose that in a certain population
from 1992 to 1996, the number of deaths from
suicide was 525, and the total number of deaths
was 2,625. The percent of deaths due to suicide
would be calculated using the formula
525 2,625 0.20 This means
that 20 of the deaths from 1992 to 1996 were due
to suicide.
91
92
How Should These Data Best be Presented?

• Deciding which results are most important to
  present
• Choosing the most appropriate format for
  presenting those results.

92
93
A Presentation on the Leading Causes
of Death
The results you present should answer ? What
are the leading causes of death? ? Do the
causes differ by age group? By sex? By race
or ethnic group? ? Are there any significant
trends over time? ? What can be learned from
comparisons with other states or with the
United States as a whole?
93
94
What are some tips for presenting
quantitative data? (1/2)

• Develop clear messages.
• Limit your data points to those that are most
  important.
• Display data in colorful, interesting graphics.
• ? Make all of the graphics relate to the message.

94
95
What are some tips for presenting
quantitative data? (2/2)

• ? Avoid too much data on one graph.
• Intersperse data with the human element to
• personalize the statistics.
• ? Keep the so what? in mind to relate all data
  to your
• main points.

95
96
Are there any special rules for
preparing tables? (1/3)

• Effective tables have the following
  characteristics
• as simple as possible.
• self-explanatory.
• ? use a clear and concise title.

96
97
Are there any special rules for
preparing tables? (2/3)
Effective tables have the following
characteristics ? have clear and concise
labels for each row and column, and include
the unit of measurement for the data (e.g.,
years, rate per 100,000).
97
98
Are there any special rules for
preparing tables? (3/3)

• Effective tables have the following
  characteristics
• show totals for rows and columns.
• explain any codes, abbreviations, or symbols in a
  footnote.
• note the source of data in a footnote (if not
  original data).

98
99
What about charts? (1/2)

• Arrange the categories that define the bars in a
  natural
• order, such as alphabetically or by increasing
  age, or in
• an order that will produce increasing or
  decreasing bar
• lengths.
• Position the bars either vertically or
  horizontally as you
• prefer.
• ? Make all the bars the same width.

99
100
What about charts? (2/2)

• Make the length of bars proportional to the
  frequency
• of the event.
• Leave a space between adjacent bars, to make the
  bar
• chart easier to read.
• Code different variables by differences in bar
  color,
• shading, cross-hatching, etc., and include a
  legend that
• interprets your code

100
101
101
102
102
103
103
104
104
105
105
106
Summary
106
107
Framing epidemiologic questions?

• Know the elements of question

107
108
Elements of the question

• Who what are the characteristics of the
  population you want to know about?
• Age
• Gender
• Race/ethnicity

108
109
Elements of the question (cont.)

• What About what indicator or outcome are you
  seeking information?
• Behavior
• Health/illness - particular ICD-10 codes can
  be found at
• http//www3.who.int/icd/vol1htm2003/fr-icd.htm
• ICD-9-CM codes can be found at
• http//www.cdc.gov/nchs/icd9.htmRTF
• Injuries
• Chronic diseases

109
110
Elements of the question (cont.)

• Where What areas are you interested in learning
  about?
• Geographic area (e.g. HSA, SAU, county)
• Know the town, or group of towns in the
• geographic area
• Residents v. occurrences

110
111
Elements of the question (cont.)

• When What time period are you considering?
• Trends over time
• Is a certain time frame necessary (is the
• indicator changing)?

111
112
Examples of Questions

• What is the rate of cardiovascular disease in
  New Mexico for the most recent year for
• which data is available? Five years ago? Ten
  years ago? For ages 65?
• What is the prevalence of diabetes in New
  Mexico for different age groups?

112
113
Release of PublicHealth Data

• Confidentiality
• Stability and
• Reliability

113
114
End !
137