Using and Interpreting Data

1
Using and Interpreting Data

• Community Health Assessment Program
• Epidemiology and Response Division
• New Mexico Department of Health

2
Todays Objectives

• Learn basic definitions and concepts fundamental
  to understanding public health and
  epidemiological data
• Learn how to interpret data to make it relevant
  to your context
• Become familiar with major limitations of
  frequently used data sources
• Learn how to use different types of data to talk
  about different populations
• Learn how and where to get assistance with
  accessing, using, and interpreting data

3
How these skills can benefit YOU !

• Using data has become essential for most health
  work
• Knowing where you are now and if you have reached
  your goals
• If you,
• Monitor or evaluate programs
• Write grants or seek funds
• Report to granting organizations, communities,
  leadership
• Want to measure change
• Improve a program or services
• then using data well benefits you !

4
Common Data Myths

• All data by the same name are equal
• Rates and numbers can be used interchangeably
• Morbidity and Mortality are pretty much the same
  thing
• Birth rate and pregnancy rate are the same thing

• Age adjustment is just epi mumbo jumbo

• Small numbers can NEVER be used
• Really, these things matter, were not just being
  picky!

5
Learning to Speak Epi

• Understanding terms to understand what data
  represent
• Know basics so you know which data are
  appropriate for your use

6
Review of Common Terms

• Community Health Assessment
• A process

• of gathering and interpreting
  information from multiple and diverse sources in
  order to develop a deep understanding of the
  health of a community.
• The process comes full circle when assessment
  results are used to improve the health status of
  the community.

7
Common Terms (2)

• Data
• Information or facts. Usually in numeric format.
  Most data have arithmetic properties (such as
  population counts).
• Some are simple categories (like gender or
  ethnicity) or arbitrarily assigned numbers (1 for
  no, 2 for yes)
• Frequency
• Count or number of people in group

8
Common Terms (3)

• Percent
• a number out of 100 ()
• portion of a whole
• Number of events X 100
• Total number of events

9
Common Terms (4)

• Rate
• Measure of frequency of an event for a
  defined population during a specified period of
  time.
• RATE Number of events in specified time period
  X K
• Population at risk of these events in
• specified
  time period

10
Calculating Rates Example 1

• Death Rate for NM 2002
• Numerator the number of deaths in NM 2002
  14,114
• Denominator the total population of NM 2002
  1,855,059
• Time Period 2002
• Constant 100,000
  

Death Rate NM 2002 _14,114_ X 100,000
760.8/100,000
1,855,059
11
Calculating Rates Example 2

• Infant Mortality Rate NM 2002
• Numerator Number of Infant Deaths NM 2002 170
• Denominator Number of Live Births NM 2002
  27,100
• Time Period 2002
• Constant 1,000

IMR NM 2002 _170_ X1,000 6.3/1,000 27,100
12
Commonly Used Rates

• Prevalence Rate (burden)
• Measures number of people in a population who
  have the disease at a given time. Includes new
  cases and previously contracted cases.
• Number of existing cases of a disease (at a
  specified time period) X K
• Total Population at mid-point of
• time period

• Incidence Rate (risk)
• Measures the probability that healthy people
  will develop a disease during a specified period
  time. New cases of a disease in a population
  over a period of time.
• Number of new cases of disease (at specified
  time period) X K
• Population at risk at mid-point
• of time period

13
Prevalence Rate Example

• Prevalence rate of diabetes in Community A 2003
• Numerator Number of diabetics in Community A
  34
• Denominator Population of Community A mid-2003
  2256
• Time period 2003
• Constant 1,000

Diabetes prevalence rate __34__ X1,000
15.1/1,000 In Community A 2003
2,256
14
Incidence Rate Example
Incidence Rate of Diabetes, Community A 2003
Numerator Number of new cases of diabetes
Community A in 2003 6 Denominator Population
of Community A not already diagnosed with
diabetes 2003 2,228 Time period 2003 Constant
1,000

• Diabetes incidence rate __6__ X 1,000
  2.7/1,000
• Community A 2003 2,228

15
Common Data Analyses

• Data can be analyzed in a variety of ways
• Most commonly used
• Age (common age groups lt1, 1-4, 5-9, 10-14,
  etc.)
• Sex (Male, Female)
• Race (White, African American, AI/AN, Asian,
  Pacific Islander)
• Ethnicity (Hispanic, Non-Hispanic)
• By cause (Cause of death, cause of disease, etc.)

16
Commonly Used Data Sources

• Population/ Demographic data from Census or
  Bureau of Business and Economic Research
• Birth and Death data (Vital records) from NM
  Office of Vital Records and Health Statistics
• Youth Risk and Resiliency Survey (YRRS) and
  Behavioral Risk Factor Surveillance System
  (BRFSS) from NM DOH, Office of Epidemiology

17
Commonly Used Data Sources

• Hospitalization In-patient Discharge Data (HIDD)
  from Health Policy Commission
• Infectious Disease data from Office of
  Epidemiology and Bureau of Infectious Disease.

18
Making data meaningful

• Data in itself is really not so meaningful.
• Usefulness comes from interpreting,and putting
  it into a context
• so it becomes relevant and meaningful
• to you

19
Population and Demographic Data

• Most common source is Census
• Population based data
• Use population data for calculating rates
• Use demographic data to describe a community or
  population

20
Population and Demographic Data

• General Characteristics of a Population
• Total count of people
• Age distribution ( people in each age group)
• Sex distribution (how many males and females)
• Education
• Employment/ Income
• Languages spoken
• Household characteristics (count of
  characteristics by household instead of
  individual people)
• Family characteristics (count of characteristics
  by family instead of individual people)

21
Population and Service Data

• Population
• Users
• People
• Population Based
• Patients

• Utilization
• Encounters
• Visits
• Services

They differ and are used differently!
22
Using Population Data in Calculating Rates

• Heart Disease death rate for people over 50
  years of age in NM 2000
• Numerator Number of people gt50 years of age who
  died of heart disease in NM 2000
• Denominator Population of NM in 2000 over 50
  years of age (from Census)
• Time Period 2000
• Constant 100,000

23
Heart Disease Death Rate NM 2000
Number of deaths from heart disease in NM 2000
among ____________people gt 50 years_____________ N
umber of people gt 50 years in NM in 2000
X 100,000
24
Birth and Death DataVital Records

• Includes data collected via birth and death
  certificates through the NM OVRHS
• There are some standard ways of analyzing birth
  and death data

25
Birth Data

• Standard analyses of birth data
• Percent of live births with
• Low birth weight
• Prenatal Care Began in 1st Trimester or No PNC
• Age of mother
• Education level of mother
• Fertility rate
• Birth rate
• Infant Mortality Rate

26
Fertility Rate

• Number of live births in a population during a
  ______specified time period______
• Number of female 15-44 year olds in
• mid-point of that time period

X 1,000
Crude Birth Rate
__Number of births during a given period of
time__ Mid-point total population of the given
time period
X 1,000
Infant Mortality Rate (IMR)
__Number of Infant Deaths in a given time
period__ Number of Live Births in a given time
period
X 1,000
27
Death Data

• Standard analyses of death data
• Infant Mortality Rate
• Cause specific death rates
• Age specific death rates
• Age adjusted death rates
• Leading Causes of Death

28
Cause-specific Death RatesExample
2000 Diabetes Death Rate for County A
Numerator Number of deaths due to diabetes in
County A during 2000 62 Denominator
Population of County A in 2000 174,682 Time
period 2000 Constant 100,000
___62___ X 100,000 35.5/ 100,000
174,682
29
Age- specific Death RatesExample
Death rate for children under 10 years, 2000,
County A
Numerator Number of deaths to people under age
10 in County A in 2000 22 Denominator
Population County A lt10 in 2000 277,773 Time
period 2000 Constant 100,000
__22__ X 100,000 7.9/ 100,000 277,773
30
Population Pyramid (age structure)
31
Age-adjusted Death Rates
Death rates are per 100,000 population and were
age-adjusted to the 2000 U.S. standard population
32
Leading Causes of Deathfor McKinley County 2001
33
Leading Causes of Deathfor McKinley County 2001
34
Hospitalization In-Patient Discharge Data (HIDD)
Death

• Includes data from non-federal hospitals in NM
• People admitted and discharged from hospital

Hospitalization
Ambulatory
Not in any system
35
Survey DataBehavioral Risk Factor Surveillance
System and Youth Risk Resiliency Survey

• YRRS-
• school based survey
• In schools that agreed to participate
• sample includes high school students who were
  present that day and class period
• BRFSS-
• telephone survey,
• sample includes people over 18 years with a land
  line phone

36
BRFSSCounty Level Data

• Counties must have at least 75 respondents.
• -Even with 75 respondents, there are
  limitations.
• When a county does not have 75 respondents in one
  year, two or more years of data can be combined.
• -Not all questions are asked every year.

37
Youth Risk and Resiliency Survey
38
BRFSS and Confidence Intervals
39
Confidence Intervals
40
Trends, Trends, Trends

• What is a trend?
• Looking at the same data over time
• 1) Data collected the same way over a
  period of time
• 2) Numbers large enough to calculate stable
  percentages or rates for each time period
• 3) Same length of time in each period used as a
  data point

41
Trends.

• When to use trends?
• - Multiple points in time
• - Data collected the same way over those points
  in time
• Use your context to determine how you present
  your data

42
Trends within a population
43
Trends between populations
44
USING small NUMBERS

• Concerns of the Epidemiologists
• Statistical Reliability
• Confidentiality
• Ways to Cope
• Aggregate time periods
• Aggregate populations
• Lose some detail
• Seek counsel with an Epidemiologist

45
Using Data to Tell Your Story

• Purpose of your story
• Know your audience
• Find a balance between tables, graphs, and text
• Save complex details for appendix

Its your story. Data can help to tell it, if you
know how .
46
So, I know how to use it, where do I get it?

• Data you can access on your own
• Call a DOH Epidemiologist

47
Data you can access on your own

• A few websites with data
• http//wonder.cdc.gov CDC Wonder
• Query death data and links to many other data
  on the web
• http//dohewbs2.health.state.nm.us/VitalRec/ New
  MICA
• An interactive query system of birth and
  death data
• http//www.unm.edu/bber/ BBER
• Population and demographic data
• http//www.health.state.nm.us NMDOH
• Reports and link to New MICA
• http//www.nmihi.com NM DOH IHI
• Query based health indicators
• http//factfinder.census.gov U.S. Census Bureau
• Data tables containing Census data

48
Call a DOH Epidemiologist(theyre people too!)

• DOH Epidemiologists
• District 1,Tom Scharman, 505-897-5700
• District 2, Vacant (call Corazon Halasan)
• District 3,Lisa Roth-Edwards 505-528-5074
• District 4,Sue Champagne 505-347-2409
• Community Epidemiologist, Corazon Halasan,
  505-476-3676
• Tribal Epidemiologist, Dawn McCusker 505-476-3073
• There are other, topic-specific
  epidemiologists at DOH
• we can refer you to too if necessary
• You are NOT Alone