Developed by: Annie Gjelsvik, PhD,

1
Web Access to Rhode Island Public Health Data

• Developed by Annie Gjelsvik, PhD,
• Karine Tolentino, MPH, Halima Ahmadi, and
  Christin Giordano

In collaboration with the Rhode Island Department
of Health Asthma Control Program This program is
a part of the Public Health Training Program
(PHTP) A joint activity of the Brown University
Program in Public Health, the RI Department of
Health and funded by an education grant from
Pfizer, Inc Center for Health Data and Analysis,
Rhode Island Department of Health
Funding source Center for Disease Prevention and
Control Cooperative Agreement U82/CCU122380-01
2
Introduction to Public Health Data

• Valid, reliable data on the communities we serve
• can help us to
• assess the health of a community or population
• search for causes of disease, injury and
  disability
• plan programs to meet community needs
• measure progress in prevention and control
  efforts

Adapted from Public Health Data Our Silent
Partner, http//www.cdc.gov/nchs/data/training/int
roduction.pdf
3
Individual and Population-based Data

• Quantitative health data can focus on individuals
  or entire populations.
• Classic example of individualized data is a
  patients medical record. The data are used
  primarily to improve the health of that one
  individual.
• In public health focus is primarily on
  populations (communities, cities, counties,
  states).
• Population-based data tell us about the overall
  health of that population.

Adapted from Public Health Data Our Silent
Partner, http//www.cdc.gov/nchs/data/training/int
roduction.pdf
4
Quantitative Data

• Provide what, who, when, where
• Measurable
• Involves the counting of people, behaviors,
  conditions, or other discrete events.

Adapted from Public Health Data Our Silent
Partner, http//www.cdc.gov/nchs/data/training/int
roduction.pdf
5
Quantitative Data

• Examples
• Numbers of deaths can be used to identify leading
  causes of death (what).
• Numbers of smokers and nonsmokers by gender can
  be used 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).
• 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.

Adapted from Public Health Data Our Silent
Partner, http//www.cdc.gov/nchs/data/training/int
roduction.pdf
6
Behavioral Risk Factor Surveillance System (BRFSS)
It enables the Centers for Disease Control and
Prevention (CDC), state health departments, and
other health and education agencies to monitor
risk behaviors related to chronic diseases,
injuries and death.
The BRFSS is the largest continuously conducted
telephone health survey in the world.
7
Behavioral Risk Factor Surveillance System
Rhode Island undertook the BRFSS for the first
time in 1984.
For more information, please visit
www.cdc.gov/brfss/
8
Behavioral Risk Factor Surveillance System
The BRFSS includes data about health care access,
nutritional facts, alcohol and tobacco use,
exercise habits, injury prevention, vaccinations,
womens health issues and much more.
9
Youth Risk Behavior Survey (YRBS)
The YRBS was developed in 1990 and includes
national and state school-based surveys of
representative samples of 9th through 12th grade
students, conducted every other year.
The schools and classrooms are randomly selected.
10
Youth Risk Behavior Survey
The YRBS surveys teens on weight and dietary
practices, sexual practices, tobacco and alcohol
use, exercise habits, violence, depression and
much more.
For more information, please visit
www.cdc.gov/yrbs
11
BRFSS and YRBS Strengths

• Representative

For example, the BRFSS samples approximately 4000
adults and these adults represent the entire RI
adult population via weighting.
Female
Male
12
BRFSS and YRBS Strengths

• Comparable

Vs.
Since the surveys are conducted using the same
methodology nationally, we can compare RI
estimates to national, regional, or other state
estimates.
13
BRFSS and YRBS Strengths

• Timely

The BRFSS is conducted all year, every year and
the YRBS is conducted in the spring every other
year.
14
BRFSS and YRBS Strengths

• Standard surveys

The BRFSS is conducted by trained interviewers
using computer assisted telephone interview
software. The YRBS is a self-administered
paper-based survey.
15
BRFSS and YRBS Limitations
The BRFSS and YRBS are limited in the populations
they are able to survey. For the BRFSS, a small
percentage of the population do not have any
phones, young and poor citizens only use cell
phones and there is a greater rate of refusals
for phone interviews than for face-to-face
interviews. For the YRBS, private school and
special education classrooms are excluded. In
addition, not every student in the selected
classroom will be attending school on that
particular day.
16
BRFSS and YRBS Limitations
150.0
Versus
Another limitation is the validity of answers
because people are self-reporting. Memory and
attempts to normalize can affect ones answers.
For example, someone might actually weigh 150
pounds but reports 128.3 pounds. Finally, the
surveys have restricted times. The BRFSS is
limited to about 20 minutes and the YRBS is
limited to one class period. Therefore, not all
topics can be covered.
17
This is the main page Click on the HEALTH Web
Query System-BRFS link to get to the BRFSS Web
Query System.
18
You can go to the YRBS Web Query by clicking
here.
19
Note which browsers to use.
20

Step One Select the year (s) that interest you.
You can click here for variables definition and
to find out which year (s) questions are
available.
21
Step 2 Here you select your row topic. Once you
highlight a specific topic, the questions asked
for that topic will appear in the blank space.
You must select a topic AND variable within that
topic.
Step 3 Here you select your column topic. Once
you highlight a specific topic, the questions
asked for that topic will appear in the blank
space. You must select a topic AND variable
within that topic.
22
Step Four allows you to filter your results by
specific populations. If you place too many
restrictions, it is possible that you may get
Statistically Unreliable results because of
small sample size.
Note that you can make multiple selections or
deselect selections by holding down the control
key.
Step Five allows you to limit your results to a
certain county or region.
23
Step Six allows you to check a box requesting
population estimates or an estimated number of RI
adults in the group.
24
Click here to clear the form completely.
Click here to submit your query.
25
Case-Study Scenario
Public Health Questions-BRFSS

• A public health educator is interested in
  developing a community based awareness program
  about asthma. She is particularly interested in
  differences of asthma prevalence among Rhode
  Island adults.
• In 2004, what is the percent of Rhode Island
  adults with lifetime asthma?

26
We are interested in 2004 so we check off the box
corresponding to 2004.
Then we select the row topic, in this case,
Year.
link
27
Next we select year of interview. You must
highlight the variable even though it is the only
one in the box.
28
Then we select the column topic, Asthma.
29
Finally, we select the column variable Ever told
have asthma by health care professional which
corresponds to the question we want to ask.
30
We do not want to filter any populations so we
leave this section blank.
31
Now we can press Submit Query in order to
submit our query.
32
This is the summary page. Check to make sure that
you have selected the desired year (s), row and
column variables and press submit. If something
is incorrect, you can press back to return to the
query page.
33
These are your results. Your search criteria will
always appear on top.
This is where your row topic appears. In this
case the year you selected.
This is where your column topic appears. In this
case, whether or not the respondent has ever been
told they have asthma by a health care
professional.
34
604 respondents said Yes, that they had been
told that they have asthma by a health care
professional.
n is the number of people who were interviewed.
35
In 2004, among Rhode Island adults ages 18 and
older, an estimated 14.6 have been diagnosed
with asthma.
This is the approximate actual percentage of the
RI population that has this characteristic.
36
This shows the 95 confidence interval (95CI),
which we will discuss later.
37
(No Transcript)
38
(No Transcript)
39
Public Health Question

• In 2004, were there differences based on
  race/ethnicity in percent of current asthma?
• What happens when we combine years?

40
We are interested in 2004 so we check off the box
corresponding to 2004.
41
(No Transcript)
42
Now we can press Submit Query in order to
submit our query.
43
Note that data is not available for Current
Asthma in 1998 and in 1999.
Now we can press Submit Query in order to
submit our query.
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
(No Transcript)
48
(No Transcript)
49
To increase the sample size, we can add more
years. This time we study years 2001 through 2004
while leaving everything else the same.
50
Note that data is available for all of the years
that we selected in Step 1.
Press submit.
51
(No Transcript)
52
Confidence Intervals

• An estimated range of values usually to 95
  confidence Were 95 confident that the
  confidence interval contains the true but unknown
  value.
• A very wide interval suggests that many possible
  values of the estimate are possible.
• If the intervals overlap for two different
  populations, you cannot say that the populations
  are different.

53
(No Transcript)
54
(No Transcript)
55
Interpreting Results -
Combining 2001 to 2004, among Rhode Islanders age
18 and older the proportion of asthma among
Black, Non-Hispanics and Whites, Non-Hispanics
were higher than in Hispanics. -The 95 CI do not
overlap for Whites and Hispanics or for Blacks
and Hispanics but they do overlap for Whites and
Blacks.
56
Case Study Scenario

• A health clinic is applying for funds to increase
  the availability of routine check-ups for women.
• They would like to demonstrate the connection
  between routine check-ups and mammograms.
• What question should they ask to generate data
  that may be helpful in advocating for increased
  funds for routine check-ups?

57
Mammograms

• Do you want to know
• what percent of women ages 40 and older who ever
  had a mammogram also had a routine check-up in
  the past 12 months
• or
• what percent of women ages 40 and older who have
  had a routine check-up in the past 12 months have
  ever had a mammogram?

58
Mammograms

• Use what percent of women who got a routine
  check-up in the past 12 months have ever had a
  mammogram?

59
We are interested in 2004 so we check off the box
corresponding to 2004.
60
Press Submit Query.
61
Once again, check your selections and press
submit.
62
(No Transcript)
63
This time we check off all years.
Next we select Health Care Access and then
Time since last routine check-up.
Finally, you select Womens Health and Ever
Mammogram-Age 40.
64
Press submit.
65
(No Transcript)
66
Interpreting Results - Combining 1998-2001 and
2004 an estimated 94.1 (95 CI 93.5 to 94.8)
of women age 40 who had a checkup in the last
year also had a mammogram in their lifetime -If
we look at the other intervals since last
check-up, it is important to note that the
percent of women who get a mammogram
significantly decreases as the time since last
check-up increases. -What does this mean for the
clinic?
67
Case Study Scenario

• All women age 65 have mammograms covered as a
  part of Medicare.
• For 1998-2001 and 2004 combined, what percent of
  women age 40-64 who got a routine check-up in the
  past 12 months have ever had a mammogram?

68
We check off years 1998-2001 and 2004.
Next we select Health Care Access and then Time
since last routine check-up.
We once again select Womens Health and Ever
Mammogram-Age 40.
69
To get results for only women ages 40-64, we can
select these age groups in Step 4. Hold down the
control button to select multiple items or to
deselect an item.
Press submit.
70
(No Transcript)
71

• Interpreting Results
• Combining 1998-2001, and 2004, an estimated 94.0
  (95 CI 93.2 to 94.9) of women age 40-64 who
  had a checkup in the last year also had a
  mammogram in their lifetime.
• -If we look at the other intervals since last
  check-up, it is important to note that the
  percent of women who get a mammogram
  significantly decreases as the time since last
  check-up increases.

72
Public Health Questions- YRBS

• In 2005, what percent of Rhode Islanders grades
  9-12 were depressed (Past 12 months, ever felt so
  sad or hopeless almost every day for 2 wks in a
  row that stopped doing some usual activities?)?
• In 2005, what percent of Rhode Islanders grades
  9-12 were in a physical fight that caused injury
  (Past 12 months, times in physical fight in which
  injured and had to be treated by doctor or
  nurse?)?
• In 2005, what percent of Rhode Island students
  grades 9-12 who were depressed had gotten in a
  physical fight that caused injury?

73
Lets start by answering the first question In
2005, what percent of Rhode Islanders grades 9-12
were depressed (Past 12 months, ever felt so sad
or hopeless almost every day for 2 wks in a row
that stopped doing some usual activities?)?
We are interested in 2005 so we check off the box
corresponding to 2005.
74
(No Transcript)
75
(No Transcript)
76
Now lets answer the second question In 2005,
what percent of Rhode Islanders grades 9-12 were
in a physical fight that caused injury (Past 12
months, times in physical fight in which injured
and had to be treated by doctor or nurse?)?
We are interested in 2005 so we check off the box
corresponding to 2005.
77
(No Transcript)
78
(No Transcript)
79
(No Transcript)
80
Depression and Violence
Lastly, lets answer the third question In 2005,
what percent of Rhode Island students grades 9-12
who were depressed had gotten in a physical fight
that caused injury?
81
(No Transcript)
82
(No Transcript)
83
(No Transcript)
84
Appendix Interpreting Results

• Use the statement An estimated X of row
  variable are column variable.
• For example, if the row variable is those with
  asthma and the column variable is those who
  smoke, what would the statement be (if X were
  40)?
• An estimated 40 of those with asthma are
  smokers.
•