From Data to Implementation and Back Again: Lessons from VAs - - PowerPoint PPT Presentation

▶

Jan 14, 2023 135 likes •342 views

From Data to Implementation and Back Again: Lessons from VAs Emerging Learning Health System Maria Souden, PhD Associate Director, VA Information Resource Center AcademyHealth Annual Research Meeting 27 June 2017 A Learning Health System

SLIDE 1

From Data to Implementation and Back Again: Lessons from VA’s Emerging Learning Health System

Maria Souden, PhD Associate Director, VA Information Resource Center AcademyHealth Annual Research Meeting  27 June 2017

SLIDE 2

A Learning Health System is one where… “science, informatics, incentives, and culture are aligned for continuous improvement and innovation, with best

practices seamlessly embedded in the delivery process

and new knowledge captured as an integral byproduct of the delivery experience.”

SLIDE 3

Iterative cycles of a Learning Healthcare System

SLIDE 4

Elements of the Supporting Infrastructure

1) well-developed digital infrastructure to capture and store administrative and clinical information

2) mature data utility that can turn this diverse set of information into usable

knowledge to guide decisions at every level of the health system 1) considers science, informatics, incentives, and culture 2) supports continuous improvement and innovation

SLIDE 5

Today’s panel

Lessons Learned about Supporting Data Literacy for a Learning Healthcare System Maria Souden, PhD Big Data Phenomics in the VA Mary Whooley, MD From Learning to Doing: Identifying, Implementing, and Spreading Promising Practices in the VA Healthcare System George L. Jackson, PhD, MHA The Partnered Evidence-based Policy Resource Center: Building the VA Learning Healthcare System Steven D. Pizer, PhD Discussant Steve Fihn, MD, MPH

SLIDE 6

RDW

V20 V19 V18 V22 V21 MOSS Farm

RDW

V12 V15 V16 V17 V23 MOSS Farm

BI SharePoint (MOSS) Farm

Performance Point Services
Excel Services
Reporting Services
Analysis Services
Collaboration Services
Team Foundation Services

CDW

SAS Grid VINCI Apps PMAS

GIS

MOSS Farm

Enterprise (AITC)

RDW

V1 V2 V3 V4 V5 MOSS Farm

RDW

V6 V7 V8 V9 V10 V11 MOSS Farm

Region 1 Region 2 Region 3 Region 4

Annually:

1.6 billion outpatient encounters, 9 million hospital admissions
3.2 billion clinical orders; 5.6 billion lab tests, 1.5 billion prescriptions filled
165 million radiology procedures
2 billion text notes (unstructured), hugely enriched with a lot of information

Typical day:

420,000 patient encounters, 2.4 million lab results , 553,000 pharmacy fills

VA Data

SLIDE 7

The VA’s Corporate Data Warehouse (CDW)

Relational database containing VHA administrative and clinical data
Most data dates to fiscal year 1999
Central to business management, population health management,

quality improvement, research, and health system innovation

Comprises 50+ data “domains,” including:

– Inpatient and Outpatient Records – Inpatient and Outpatient Pharmacy – Lab Tests and Results – Enrollment information

Units of analysis vary by table, ranging from drug to diagnosis to inpatient

stay to consult to prescription refill

7

SLIDE 8

Opportunities Challenges

8

SLIDE 9

SLIDE 10

SLIDE 11

Supporting “data literacy”

Work with and document the data
Educate data users through:

– Factbooks, Technical briefs, User Guides, Statistical Snapshots – Cyberseminars & tutorials

Disseminate resources through:

– Websites & Communication portals – Listservs – HelpDesk – Newsletter: Data Issues Brief

SLIDE 12

Lesson 1: Context is critical, and not always obvious

Research and quality improvement use of data is secondary to

its original purpose

Requires in-depth understanding of how the data were

generated and transformed

Know what you have and what it means

SLIDE 13

13

SLIDE 14

Lesson 2: Need for a broad knowledge pipeline

Beyond “institutional” knowledge
Support diversity of questions, uses, and applications of data
Keep up with complexity, change, growth

14

SLIDE 15

15

SLIDE 16

LHS Lesson 3: People are part of the infrastructure

Explicate tacit knowledge- expertise residing in people
Data reuse can be nuanced- don’t know what you don’t know
Knowledge distributed across organization

16

SLIDE 17

HSRData Listserv

Email-based question-answering community of ~1,300 VA

data users

Membership spans disciplines, functions, offices
Posts have increased fourfold since CDW came into broad use

(30120/month)

Average 2.9 answers per question
Thoughtful and valued community resource
But lacking knowledge organization and affordances for reuse
f information shared.

17

SLIDE 18

Next-generation HSRData Forums - better supporting the Learning Healthcare System

SLIDE 19