Transcript Treatment engagement in substance use disorders

Using VistA for Research
Joseph Glass, MSW
VA Ann Arbor Healthcare System
Ann Arbor, MI
The contents of this presentation are mine personally and do not
reflect any position of the Government or the Department of
Veterans Affairs.
Acknowledgements



VA Ann Arbor Healthcare System
Ken Baker and John Wilkerson, VA Office of
Information and Technology
Stephen Chermack, PhD, Chief, Mental Health
Clinic and Substance Use Disorders Clinic
University of Michigan
Dale Fitch, PhD, Asst. Prof, School of Social Work
Presentation overview




Potential that VistA has for conducting
research
Methods used to harvest and collect data
using VistA
Examples of studies that have used VistA
Brainstorming / Discussion
VistA’s research potential



Utilizing data already in VistA
VistA as a data collection tool
VistA as a research data storage tool
Utilizing data already in VistA

Demographics

Age, gender, race, marital status, income, disability status,
military status
Predictors and/or outcome variables:
 Interventions


Health factors


Medical and psychiatric diagnoses, weight, glucose
Screening results


Medications, specialty care, clinic appointments
Diabetes, depression, PTSD, alcohol
Outcomes

Inpatient hospitalization, resource utilization, mental health
outcome measures, Addiction Severity Index
VistA as a data collection tool

Clinical Reminders



Use to identify research participants
Reminders can trigger need for assessment
Reminder templates



Avoids narrative data storage (typically bad
for research)
Implement short questionnaires
Data stored VistA, rather than Excel
spreadsheet or MS Access
continued: VistA as a data
collection tool

Mental Health Assistant



Self-assessment, clerk entry
Stores data in VistA
Version 3: Add your own measures!
VistA as a data storage tool


Centralized storage: eliminate hundreds of
spreadsheets floating around your clinic
Query one database for all the information
you need
Summary:
VistA’s potential for research

Reuse, collect, and store data in one
integrated system
Harvesting VistA data
Myth:
There is a lot of data put into VistA …
… But not many ways to get it out.
Harvesting Methods




Mumps Data Extractor
Cache Shadowing
FileMan
Others
Mumps Data Extractor






Commercial: Strategic Reporting Systems
Quick installation
Natively reads VistA files and dumps to
SQL
Maps variables correctly
Manages data types
Batch dumps may take a while, should
schedule at night
http://mde.srs-inc.com/solution.html
Cache Shadowing


Real-time mirroring of live VistA data
Stored on a separate server



Avoid taxing production server
Cache has SQL support
Built-in ODBC access
Others



Fileman
RPC
Mumps programming
Comparison: Extraction Utilities
Pros
SQL
Mumps Data
Extractor
Cache
Shadowing
Fileman
Cons
ready
Maps variable names automatically
Translates variables types, escape
codes
Quick implementation
Extra
Real-time
Requires
SQL
ready
Cost Included in Cache license
variable mapping
and translation
Hard to implement
Free
No
Gets
data: Clinically useful
cost
Taxes production server
Older data
Frequent IT intervention
the job done
“dump to SQL” option
Every query uses
production resources
Requires end user to learn
a specialized language
Recommendation: Use Cache Shadowing
Extracting data: Case Study
VA Ann Arbor
MS SQL Server
Researcher
with SPSS
VistA
Mumps Data Extractor
(Every 2 weeks)
Future Direction: VA Ann Arbor
VistA
Researcher
with SPSS
Cache Server Pages
Cache Shadow
Server
Email alert to clinician
regarding a significant
clinical event
Potential discussion items

Implementations must meet user ability


Disconnect between IT and
clinicians/researchers


Cost, specialized knowledge
Communication barriers
How can we improve VistA research?
Resources
Finding the data you need

Database/files list


http://openforum.worldvista.org/~forum/index.php?title=VistA_Database_Globals
Vista Documentation Library

http://www.va.gov/vdl/