Utah

1
Utahs Indicator-Based Information System for
Public Health (IBIS-PH)

• Utah Department of Health
• Office of Public Health Assessment
• Lois M. Haggard, Ph.D., Director

2
UDOH Organizational Structure
The UDOH executive director takes a personal
interest in IBIS. The Center for Health Data is
located in the Executive Directors Office. It is
important to have high-level sponsorship for the
project to be successful.
3
http//ibis.health.utah.gov
Utahs IBIS-PH homepage provides access to all
areas of the website and information on new
features. This is a standard home page template,
but states adopting IBIS can either modify the
content only, or the whole page.
4
IBIS Information Dissemination Model Most
frequently-asked questions are answered in paper
(.pdf) publications and in our Indicator Profile
reports. Users whod like more detailed analysis
or sub-populations can use the interactive query
system. If none of those systems answer their
question, they must go to the old system of
calling us on the phone.
5
IBIS has over 100 indicator profiles currently
published. They are indexed in an alphabetical
index (pictured here) and also in a categorical
index that allows indicators to be placed in any
of several categorization schemes, such as by
program, by data source, or HP2010 focus area,
etc.
6
Each indicator has a default graph that comes up
on the indicator main page.
7
The main page also allows navigation to a table
of numbers for the current graph, as well as
access to additional graphs (if they exist). We
also include the measure definition and a message
on why it is an important construct in public
health.
8
IBIS graphs are displayed using SVG (scalable
vector graphics, a standard form of XML). They
use commercial software from Elansoft that allows
for interactive mouse-overs. Users must have an
SVG-Viewer from Adobe (similar to Adobe Acrobat)
installed on their computer to use the mouse-over
capability. However, if a user does not have the
SVG-Viewer installed, the IBIS system will
produce a jpeg image and display the image in
place of the SVG.
9
Important facts about each measure include a
number of fields. Not all fields are mandatory
mandatory fields are shown here with an asterisk.
Definition
Data Interpretation Issues
Why Is This Important?
HP 2010 Objectives
Other Objectives
How Are We Doing?
Utah Versus U.S.
What Is Being Done?
Other Program Information
10
The IBIS indicator profiles allow users to see
how the current measure might be related to other
measures.
11
For instance, an important health status outcome
for binge drinking is motor vehicle crash
injuries.
12
Programs can use IBIS to inform the general
public of services and resources.
13
IBIS Indicator Reports Architecture
IBIS System
14
Admin DB, Relational DB Entities (Tables)
This is a (gross) oversimplification of the IBIS
database structure. Certain information is
stored at the indicator level, such as definition
and HP2010 objectives. Each indicator may have
one or more graphical data views. Each graph
view has multiple data points (indicator view
values).
15
Indicator Attributes (Table Fields)
Here are some examples of fields in the Indicator
table of the IBIS DB.

• Name, Title
• Definition
• Numerator Denominator
• Important Facts (PH Context)
• Related Indicators
• Owner
• Status (open, completed, published)
• Etc

16
Indicator View Attributes (Fields)
Here are some examples of fields in the Indicator
View table of the IBIS DB.

• Name
• Indicator Name
• Graph Title Sub-Title
• Y-axis titles
• Data Notes
• Data Sources
• Category (X-axis) Series (Legend) Labels

17
Indicator View Value Attributes (Fields)
Here are some examples of fields in the Indicator
View Value table of the IBIS DB.

• Category
• Series
• Value
• Lower Confidence Limit
• Upper Confidence Limit
• Label
• Active Flag

18
IBIS Admin Site
Program staff from across the department use a
website, called the IBIS Admin site to access
the database and modify data records. Once an
indicator has been set up, the maintenance is
minimal, generally limited to annual update of
numeric data and modification of text to reflect
those changes.

• GUI Interface for DB Access
• Used by busy folks who do not necessarily have
  website publishing or DB data entry skills (with
  no interest in acquiring them)
• Must be user-friendly, follow good IS
  rules/heuristics

19
Here are the first 18 records in the admin
system. Since I am a super-user, my screen
shot shows all 130 indicator records. But if I
were a program person, I would only see those
indicators that I have permission to modify. The
Admin site is pretty simple to use, which is
necessary if we are to get the cooperation we
need.
20
This is the indicator main page, where we have
provided text boxes to maintain some of the
fields at the indicator level.
21
The four buttons near the bottom, Graph List,
Graph Specifications, Graph Data Sources, and
Graph Values, provide the ability to modify
different graph views. This slide shows the list
of graph views available for the Binge Drinking
indicator. Notice that not all views are active
(box on right). The indicator owner is still
working on these views and has not made them
available to the public yet.
22
Here is the data entry page for the Binge
Drinking by Age and Sex graph view. Eventually,
we want to have a system that populates these
data values from our query system. Right now,
the data must be entered by hand.
23
Custom Data Query

• Allows users to query health data sets directly
• General audience
• User-friendly GUI interface
• Sound epidemiology
• Help descriptions instructions
• Bullet-proof cant confuse or give misleading
  answers, cant allow users to do stupid things

The IBIS Query system allows members of the
general public to access several public health
data sets and generate query results based on
their specifications.
24
The custom query tab takes users to the IBIS-Q
interactive data query system.
25
From the query entry page, users can go to
tutorials, or request query modules for a variety
of data sets, such as the BRFSS survey data.
26
Context-specific help buttons provide insights to
questions we can anticipate. We can provide help
text wherever we think it may be useful. These
buttons can link to a help message, such as this
one, or a help webpage.
27
The user goes through the steps and makes
choices. Weve found that our users really like
the task to be organized into steps, so they can
clearly see what is required of them.
28
The last step before submitting the request is
choosing how to display the results. The system
can do more than two dimensions, but,
interestingly enough, our users told us they
didnt want that complexity!
29
After submitting the query request, the system
returns a confirmation screen that allows the
user to proceed, or go back and modify their
query. This is one of the optional elements that
experienced users may want to turn off to improve
system performance.
30
The results page includes documentation of the
query parameters. Here, the results have been
displayed by age group for Salt Lake County. By
clicking the Output to Excel button, users may
also output the data directly to Excel.
31
IBIS can graph the results in several graphical
formats.
32
A 2-D table provides results if two dimensions
have been selected. We like to always present
survey data with the percentage and its
complement, to avoid user misinterpretation.
33
A data list (fact table) provides information in
more detail than the 2-D table.
34
Meta-data (information about the data) can be
customized for each query module and measure.
35
The IBIS query system can also produce map views
of data. Maps provide mouse-over information and
zooming.
36
Thank you!

• Lois M. Haggard, Ph.D.
• Office of Public Health Assessment
• Utah Department of Health
• loishaggard_at_utah.gov
• [phone]

37
Custom Data Query

• IBIS interface (view system)
• Creates request URL from user selections
• Passes is to the C program
• C program
• Parses parameters
• Loads config, function, and label files and
  builds SAS code file
• Calls SAS

38
Query Architecture
IBIS System
39
IBIS-Q Module Structure
40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
The Trade-Off

• More Module.xml files
• Shorter, simpler .xml files
• More files to maintain
• Fewer Module.xml files
• Longer, more complex .xml files
• Fewer files to maintain