Input Your Data

1
Input Your Data Output Your Website
A
Web-Based Tool for Quality and Utilization
Reporting

• Anne Elixhauser
• Mamatha Pancholi
• Jenny Schnaier

December 4, 2008
2
Development Team
Leveraging the Expertise of Four Organizations

• Produce HCUPnet-like utilization queries
• Test the tool

• Convert existing
• WinQI Wizard
• Produce QIs
• Adapt PHC mapping tool
• Test the tool

• Lead web development
• and presentation
• Coordinate design
• efforts
• Test the tool

3
Overview
AHRQ EQUIPS you with the tools you need to
generate a website transforming your health care
data into information about health, costs, and
quality of care.

• What is EQUIPS?
• Why is it important?
• Who is the audience?
• How will EQUIPS work?
• When will it be deployed?

4
WHAT The Concept
A web-based tool that provides local information
on
HEALTH CARE USE, COSTS, VOLUME, AND
QUALITY
AHRQ creates and distributes software programs to
generate web-based query system
Local organizations use programs to host a
web-based query system
Local users access the web-based query system to
obtain health care information
?
THE BIG QUESTION WAS IS THIS FEASIBLE
Project presented and approved as part of Value
Portfolio Supports CVEs
5
WHY Rationale
CONSUMERS NEED HEALTH CARE INFORMATION TO MAKE
INFORMED CHOICES
What hospitals in my area perform heart
surgery? How many bypass procedures are
performed at hospital x versus hospital y? What
is the quality of care for cardiac procedures at
hospitals in my area? I dont have insurance,
how much should I expect to pay?
Currently, there is very limited health care data
available to answer
such questions
6
Rationale

• State/local organizations have resource
  limitations
• Unable to develop comprehensive reporting system
• AHRQ can provide support for these organizations

7
From Data to Information to
Action
Hospital discharge data can generate valuable
healthcare information
Consumers can use this information to make
decisions
Hospital discharge data are already being
collected
8
WHO Two User Groups
Host and End User

• HOST USER Organization building the website
• State and local data organizations
• Chartered Value Exchanges
• Hospitals and health care systems
• Health plans
• Anyone with access to provider-level data

END USERS People using the website
Consumers Health Planners Policy
Makers Media Data Analysts/Researchers
9
HOW The Making of EQUIPS...
Iterative rapid development
Continuous Alpha Testing
Storyboard
Beta Test
EQUIPS code
EQUIPS code distributed to organizations with
access to hospital discharge data
Host organizations implement code and load
discharge data to create local EQUIPS website on
their own servers
End user website
10
General Considerations as Development Proceeds
USING A SMART DEVELOPMENT AND DESIGN APPROACH USING A SMART DEVELOPMENT AND DESIGN APPROACH
Efficient development Adapt existing programs to feed into the EQUIP system
Minimal burden Develop web system based on most commonly used programs and software platforms Ensure methods can be understood by wide range of host users
Customizable features Modular Customizable user interface
Comprehensibility Ensure information adheres to standards for conveying statistics to public
Documentation Document methods throughout for host and end users
Accuracy Iterative testing of results
11
Customizable Features

• Modular
• User can choose which pieces to implement
• Generate user-defined reports
• Create maps of various measures
• Drill down on specific issues by patient and area
  characteristics
• Customizable Interface
• Users can insert their logo and organization name

12
EQUIPS HCUPnet QI More
HCUPnet
QIs
Query Paths, Output Tables, Interface, Help and
Definitions
WinQI Tool, AHRQ QI Measures, Reporting Template,
Mapping Tool
13
A Brief Technical Overview
EQUIPS Site Builder (Wizard) Moves host user
through the process
Raw data from user
Programs that load data, map values, transform
data, and build data cubes
Website query interface seen by end user

• Page writers
• Graphics
• generators
• (for Utilization and QI data)

HTML and .JPGs for QIs and Utilization information
SQL Server Database stores Meta-data,
transformed data, and data cubes

• Web System
• Context and
• navigation
• page writers
• Script writers
• cascading
• style sheet
• writers

Census data Comparison data Labels

• Navigation
• Formats

14
EQUIPS Paths
15
Example Questions
16
A Peek at EQUIPS
How will it work?

• The host user (e.g., local/state organization, an
  individual hospital) will download the AHRQ
  EQUIPS Tool and generate a website

The end user (e.g., health consumers, analysts)
will access the website created by the host user
to obtain health care data
17
A Peek at EQUIPS A Working Draft
What the Host Sees
Based on the existing AHRQ Quality Indicators
Wizard. Expanded to build EQUIPS. Import Data
Wizard initial screen.
18
Web Builder
Selected Import and Analyze Data Wizard
19
Web Builder
Select input file.
20
Web Builder
Select options for loading and mapping data. Map
input file variables to variables used by
software. This file stores all of the mappings
for future use.
21
Web Builder
Summary of variables
22
Web Builder
Select which groupers to apply.
23
Web Builder
Scan file and check for errors.
24
Web Builder
Data errors report choose to exclude
problematic data.
25
Web Builder
Crosswalk Map the values from the input data to
standard values required by EQUIPs.
26
Web Builder
Define your groupings of hospitals choose
Hospital Service Areas (HSAs, from Dartmouth) or
build your own groupings.
27
Web Builder
Edit hospital lists.
28
Web Builder
Data are loaded into an SQL Server.
29
Web Builder
Generating measures that will be displayed in
website.
30
Web Builder
Preview your data.
31
Web Builder
Previewing utilization measures.
32
End User Website Interface
EQUIP Site Builder (Wizard) Moves host user
through the process
Part of Builder Which parts of EQUIPS does the
host want to create?
Raw data from user
Programs that load data, map values, transform
data, and build data cubes

• Page writers
• Graphics
• generators
• (for Utilization and QI data)

Website query interface seen by end user
SQL Server Database stores Meta-data,
transformed data, and data cubes
HTML and .JPGs for QIs and Utilization information

• Context and
• navigation
• page writers
• Script writers
• cascading
• style sheet
• writers

Census data Comparison data Labels

• Navigation
• Formats

33
A Peek at EQUIPS A Working Draft
What the End User Sees
Select your path QUALITY INDICATORS AVOIDABLE
HOSPITALIZATIONS RATES UTILIZATION
34
Quality Path

• Select one of ten
  quality health care topics
• Heart conditions
• Brain and nervous system
• Childbirth
• Hip replacement and hip fracture
• Operations for cancer of the esophagus and
  pancreas
• Other surgeries
• Other health conditions
• Medical complications of patients having an
  operation
• Medical complications of all patients
• Medical care for children

35
Quality Path
Select all indicators within a topic or specific
indicators
36
Quality Path
Select a specific hospital or up to four
hospitals in an area Allows easy access to the
data through view data now feature Allows
download to Microsoft Excel of summary tables
37
Quality Path
Click on any indicator and get detailed
information in graphic form.
38
Quality Path
Get details for each hospital in rank order,
compared to the state average.
39
Potentially Avoidable Hospitalization Path
Potentially Avoidable Hospitalizations based on
AHRQs Prevention Quality Indicators (PQIs
ambulatory care sensitive conditions)
40
Preventable Hospitalization Costs Mapping Tool
Better understand geographical patterns of
hospital admission rates. Calculate potential
cost savings if admission rates are reduced.
40
41
Estimates of Potential Cost
Savings
Potential cost savings if number of admissions
were reduced by specified percentage
QI Name
County name, mean cost of admission for
indicator, number of cases, and total cost
41
42
Rates of Conditions and Procedures Path
Rates of health conditions and procedures use
county populations as the denominator in
prevalence rates
43
Rates of Conditions and Procedures Path
Choose among diagnoses and procedures, grouped by
CCS categories
44
Rates of Conditions and Procedures Path
Choose specific counties or examine all counties
45
Rates of Conditions and Procedures Path
For all counties in the state numbers of
discharges and rate per 1,000 population Will
include a mapping function
46
Utilization Path
Utilization statistics (like information from
HCUPnet)
47
Utilization Path
Select MDCs (body system or etiology) DRGs (dx
and pr combined) CCS diagnoses (groups of
ICD-9-CM codes) CCS procedures (groups of
ICD-9-CM codes)
48
Utilization Path
Select a specific DRG
49
Utilization Path
Select specific hospitals or examine all
hospitals in Hospital Service Areas.
50
Utilization Path
Detailed patient characteristics for a specific
DRG for one hospital.
51
Utilization Path
Information on numbers of discharges, charges,
costs, length of stay, percent died for all
hospitals individually. National, regional, and
state benchmarks included.
52
Utilization Path
Hospitals grouped by Hospital Service Area
(Dartmouth Atlas)
53
What makes EQUIPS unique?
EMPOWERS
organizations and consumers to
use data to make informed decisions

• Provides organizations with the ability to
  create/host their own website and upload their
  own data
• Enables local organizations to do their own
  reporting using a standard, validated method
• Allows users to draw together multiple data
  sources that provide information at the local
  level

54
WHEN Timeline
9 months from Concept to Product
Concept First Build Continuous Builds
Beta Test
Incremental Builds
APRIL 2008 JULY 2008 JULY - DEC
2008 JANUARY 2009
Incremental build throughout Summer and Fall
using an iterative rapid application development
methodology TARGET COMPLETION DATE OF PHASE 1
JANUARY 30, 2009
55
Challenges

• Aggressive timeline
• Variety of users will want to use the system
• Different system capabilities
• Different resource availability
• System needs to have minimum requirements
• Need to keep the cost of implementation low
• Limited nature of static model versus dynamic
  model

56
Phase 1 versus Subsequent Phases
EQUIPS is an evolving tool the current version
is a prototype

• Phase 1
• 3 Paths
• Utilization
• Rates
• Quality
• Mapping capability
• National and regional benchmarks
• Static design

• Subsequent Phases
• Additional Paths
• Hospital Compare
• HCAHPS
• Evaluate other data sources
• Provide links to other resources
• Dynamic design
• Consider adding other federal data sources
• Many more ideas...soliciting user suggestions

57

• Questions?
• Comments?
• Suggestions?
• anne.elixhauser_at_ahrq.hhs.gov