, o z v o Ç ] w , } Á v t / v ] ( Ç s ] ] } v d z ^ µ } z...

19
Copyright 2018. College of Healthcare Information Management Executives (CHIME) All rights reserved. Reprints with permission only. 1 Healthcare Analytics: How Can We Identify Variations That Support Change? James Jose, MD Children’s Healthcare of Atlanta October 16, 2018 Analytics and the “Big Picture”

Upload: others

Post on 11-Oct-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 1

Healthcare Analytics: How Can We Identify Variations

That Support Change?

James Jose, MDChildren’s Healthcare of AtlantaOctober 16, 2018

Analytics and the “Big Picture”

Page 2: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 2

Getting to “All Better”

Use of visual analytics to map variations in asthma care in the emergency department

From: Understanding variations in pediatric asthma care processes in the emergency department using visual analytics, J Am Med Inform Assoc. 2015;22(2):318-323. doi:10.1093/jamia/ocu016. Rahul Basole, Mark Braunstein, Vikas Kumar, Hyunwoo Park, Minsuk Kahng, Duen chau, Acar Tamersoy, Daniel Hirsh, James Bost, Burton Lesnick, Beth Schissel, Michael Thompson.

Page 3: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 3

Realizing Value from Analytics

From: Understanding variations in pediatric asthma care processes in the emergency department using visual analytics, J Am Med Inform Assoc. 2015;22(2):318-323. doi:10.1093/jamia/ocu016. Rahul Basole, Mark Braunstein, Vikas Kumar, Hyunwoo Park, Minsuk Kahng, Duen chau, Acar Tamersoy, Daniel Hirsh, James Bost, Burton Lesnick, Beth Schissel, Michael Thompson.

“Analytics Maturity”Descriptive AnalyticsPredictive AnalyticsPrescriptive AnalyticsDiscovery Analytics

What about Implementation Science?

Questions and ChallengesHow can analytic tools make a real difference for patient care?

1. How can we configure partnerships to design effective analytic strategies?

2. What must we know about a process to make analytic insights meaningful?

3. How can we promote lasting value?4. Beyond “analytic and presentation tools” – What are

we missing?

Page 4: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 4

Thinking about variation

Walter Shewhart’s 1924 Memo to Western Electric Employees

Page 5: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 5

Process Control Chart - Rules for Identifying “Special Cause” Variation

Why does variation matter?

Standards are of little service if there is no “process control” to administer them.

Studies have shown that simple elimination of variation can reduce cost and improve quality.

Transforming an organization or building an effective culture, requires elimination of “useless variation.”

Page 6: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 6

Approach to Variation• Determine your strategic focus to reduce variation • Set measurable goals • Acquire and analyze data • Understand your data • Identify areas of focus • Implement improvements

“Health Care Leader Action Guide: Understanding and Managing Variation” report of the American Hospital Association’s Task Force on Variation in Health Care SpendingAccessible at: http://www.hret.org/quality/projects/healthcare-leader-action-guide-understandingmanaging-variation.shtml

Approach to Managing Variation• Determine a strategic focus to reduce variation based

on value to organization and individuals. Engagement starts here.

• Set measurable goals that fit sustainable operational activities.

• Acquire and analyze data to understand and influence. • Understand your data. Understand the roles and

workflows that produced your data. • Identify areas of focus. Understand how your data

motivates clinicians and staff. • Help your clinical partners implement improvements.

Page 7: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 7

Bronchiolitis Guideline and Analytics • Purpose: Reduce unnecessary use of bronchodilators

and diagnostic testing • Process Design:

Clinical Practice Guideline updated to reflect practice changes of unnecessary use of treatments

Qlik Application designed to monitor resource use.

13

The “Bronchiolitis App”

Bronchiolitis analytics prior to new program

Previous Analytic Tools• Hard to interpret• Reports not real time• Took months to drill

down on data.

14

Page 8: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 8

Chest X-ray & albuterol use decreased – Why?ED Chest x-Ray usage 2012-2016

IP Albuterol Use 2012-2016 IP ALOS 2012-2016

ED Albuterol Usage 2012-2016

Startup: 2014 – Compliance with Bronchiolitis Pathway Goals

Page 9: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 9

Year 1: StartupCompliance with Bronchiolitis Pathway Goals

Year 2: Post InterventionCompliance with Bronchiolitis Pathway Goals

Page 10: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 10

Year 3: Post InterventionCompliance with Bronchiolitis Pathway Goals

Year 4: Post InterventionCompliance with Bronchiolitis Pathway

Goals

Page 11: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 11

2012-2018: Sustained ChangeCompliance with Bronchiolitis Pathway Goals

1. Reassuring clinicians interventions are safe2. Elimination of variation has therapeutic value

ED Chest x-Ray usage 2012-2016

ED Albuterol Usage 2012-2016

IP Albuterol Use 2012-2016

IP ALOS 2012-2016

Page 12: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 12

The “Burnout Quadrant:”Provider ( ) with Low Efficiency and High Workload

WorkloadIncreasing Workload

Increasing Efficiency

Efficiency Score

Specialty AverageProvider

Provider Example #1“Efficient Provider” Reputation – Why Low Score?

Workload

Increasing Efficiency

Efficiency Score

Increasing Workload

Specialty AverageProvider

Page 13: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 13

Provider Example #1Physician Efficiency Metrics: Guiding Optimization

Minutes Use/Hrper Day

EMR Activity by Hour of Day

7am 6pmHour of Day

Provider Example #1 Provider Improvement Over 6 Months

Workload

Increasing Efficiency

Efficiency Score

Increasing Workload

Specialty AverageProvider

Page 14: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 14

Provider Example #2: High Work Load / Low Efficiency Provider

Workload

Increasing Efficiency

Efficiency Score

Increasing Workload

Specialty AverageProvider

Provider Example #2: High Work Load / Low Efficiency ProviderBaseline Data

Minutes Use/Hrper Day

EMR Activity by Hour of Day

Hour of Day7am 6pm

Page 15: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 15

High work load / low efficiency provider Provider Example #2: Results after 6 months

Workload

Increasing Efficiency

Efficiency Score

Increasing Workload

Specialty AverageProvider

Provider with Excessive Notes Activity Effort

Minutes Use/Hrper Day

Hour of Day

EMR Activity by Hour of Day

Increasing Efficiency

Increasing Workload6pm7am

Page 16: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 16

Minutes Use/Hrper Day

Hour of Day

Minutes Use/Hrper Day

Hour of Day

Increasing Efficiency

Increasing Workload

Increasing Efficiency

Increasing Workload

6pm7am

6pm7am

EMR Activity by Hour of Day

EMR Activity by Hour of Day

Provider with Excessive Notes Activity EffortResult of 3 meetings, 6 months at-elbow support

Time in Notes Activity During Intervention with Low Efficiency Providers

Time in Notes Activity

Minutes/Month2017 Average

2018 March April May June July Aug Sept

(Lower numbers reflect higher efficiency)

Page 17: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 17

After Hours Charting Time During Intervention with Low Efficiency Providers

Self-service analytics workflow:Probing for Excessive Variation

“Non-standardization” as probe for sub-optimal care

POINT

TASK

Quality manager looks at distribution of diagnostic tests, medications to determine if wide variation in care patterns

WORKFLOW

1) Effective in Identifying need for guidelines2) Helpful tool in driving consensus

RN Manager Quality and Clinical Effectiveness

Who?

Page 18: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 18

“Pop-Disco” Self-Service Analytics Tool”

12

79

38

29

64

31

42 4348

25 24

62

20

60

10

20

30

40

50

60

70

80

Tota

l Num

ber

of V

isit

s

Guideline Off guideline

Variation as a Probe for Sub-optimal Care

Medications used for behavioral health

Getting Value from Healthcare Analytics for 0perational/ClinicalChange

1. Effective Analytics require comprehensive knowledge of a process: who are the people, what is the goal?

2. It’s all about the “personal mission” … as perceived by each partner in the process you are controlling.

3. Think sustainable. Is your product embedded in a standard process? Used by an aligned partner?

4. Are the metrics getting traction? See step #1? • Do your metrics prove the change’s value? • Do they recruit and influence people who will act?

Page 19: , o Z v o Ç ] W , } Á v t / v ] ( Ç s ] ] } v d Z ^ µ } Z ...ga.himsschapter.org/sites/himsschapter/files...o o ] P Z À X Z ] v Á ] Z u ] ] } v } v o Ç X ï Z o ] Ì ] v P s

Copyright 2018. College of Healthcare Information Management Executives (CHIME)

All rights reserved. Reprints with permission only. 19

AcknowledgementsWe gratefully acknowledge material contributions by:

James Bost, PhDDaniel Hirsh, MDEric Housel, RNBurton Lesnick, MDRobert Palmer, PhDEdwin Ray, RNBeth Schissel, MDJavier Tejedor-Sojo, MDNatalie Tillman, RNMichael ThompsonChildren’s Healthcare of Atlanta Outcomes Center