The Imperative of Linking Clinical and Financial Data to Improve Outcomes - HAS Session 17

Charles G. Macias M.D., M.P.H.

The Imperative of Linking Clinical and Financial Data to Improve Outcomes

Chief Clinical Systems Integration Officer, Texas Children’s Hospital

Learning objectives

Assess the effectiveness of an organization’s quality gaps to ensure organizational readiness, drive efficiency and leverage opportunities to improve quality.

Illustrate how a blend of clinical and financial data informed by analytics from an enterprise data warehouse can improve outcomes.

Describe how an EDW and care process implementation can encourage a culture of quality and safety, providing physicians with the necessary tools to integrate financial relevance into the practice of delivering high-quality healthcare.

Discuss how strategy for integration of science, data and predictive analytics and operational improvement through improvement science can transform a system towards the triple aim.

Jenny Jones and the Challenges of a Fragmented System

Within six months, Jenny had visited: One PCP Two Hospitals Three ERsLeading to: Six different Asthma Action Plans with

conflicting discharge instructions

Quality DefinedInstitute of Medicine domains:

Safe Effective Efficient Timely Patient centered Equitable

The degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge.

– Lohr, K.N., & Schroeder, S.A. (1990). A strategy for quality assurance in Medicare. New England Journal of Medicine, 322 (10):707-712.

Importance of minimizing unintended variation in health care delivery

1 2

3

The Healthcare Value EquationIn an environment where cost is marginally increasing, healthcare must markedly improve quality.

Adoption of EMRs and clinical systems should help push the quality agenda but alone may not be enough to deliver data intelligence.

5

Value =Quality

Cost

In Second Look, Few Savings from Digital Health Records

New York Times: January 10, 2013 2005 RAND report forecasts $81 billion annual U.S. savings. “Seven years

later the empirical data on the technology’s impact on health care efficiency and safety are mixed, and annual health care expenditures in the United States have grown by $800 billion.”

Disappointing performance of health IT to date largely attributed to: Sluggish adoption of health IT systems, coupled with the choice of systems that

are neither interoperable nor easy to use; The failure of health care providers and institutions to reengineer care processes

to reap the full benefits of health IT.

EHRs, Red Tape Eroding Physician Job Satisfaction Most physicians express frustration with the failure to provide efficiency. 20% want to return to paper

6

Variation in Care

7

Describing variation in care in three pediatric diseases: gastroenteritis, asthma, simple febrile seizure Pediatric Health Information System database (for data from 21 member hospitals) Two quality-of-care metrics measured for each disease process Wide variations in practice Increased costs were NOT associated with lower admission rates or 3-day ED

revisit rates

Implications? Optimal care may be delivered at a lower cost than today’s care!

Kharbanda AB, Hall M, Shah SS, Freedman SB, Mistry RD, Macias CG, Bonsu B, Dayan PS, Alessandrini EA, Neuman MI. Variation in resource utilization across a national sample of pediatric emergency departments. J Pediatr. 2013

Consumer Care/Cost UncertaintyConsumers:

Trust their physicians

Hope for the best

Struggle to understand cost and care

Don’t often know what they are getting

Don’t always get great outcomes

Value is what they want

8

Challenge of HealthcarePhysicians are:

Driven by science and key values

Overwhelmed with medical literature

Not well trained to turn that experience into high quality patient outcomes

Transparency of local data is part of the solution!

9

Poll Question #1

In your organization, what percentage of patient visits are your physicians talking about cost and care tradeoffs at the bedside?

10

a) 0-19%

b) 20-39%

c) 40-59%

d) 60-79%

e) 80-100%

f) Unsure or not applicable

Source: SAEM. Evidence Based Medicine Online Course 2005

Physicians and Care Cost

Clinical Expertise

Evidence Patient values and preferences

Physician preferences

Resource issues

Once taboo, physicians should take cost into consideration:

12

No Money No Mission No Expansion No Innovation

And so providers must….. Understand what creates improvements Understand the story that their data tells.

About Texas Children’s Hospital

Statistics

Number of Beds 469

Annual Inpatient Admissions

21,744

Annual Outpatient Visits

1.44 million

Emergency Room Visits

82,049

Inpatient Surgeries 8,655

Outpatient Surgeries 14,439

A data management strategy to improve outcomes

IMPROVED OUTCOMES from high quality of care

SOURCE SYSTEMS (e.g. EMR, Financial, Costing, Patient Satisfaction)

ANALYTIC SYSTEMData analytics and collaborative data

CLINICALCONTENT SYSTEM

Science and evidence

DEPLOYMENTSYSTEM

Operations

Informatics, Electronic Data Warehousing

Advanced Quality Improvement course, QI curriculum, Care process teams

Evidence Based Guidelines and Order sets, Clinical Decision

Support, patient and provider materials

Patient centric outcomes and institutional

outcomes achieved

Creating a foundation for EB practice





DEPLOYMENTSYSTEM

Operations

Evidence Based Guidelines and

Order sets, Clinical Decision Support,

patient and provider materials


Evidence-Based Guidelines: EBOC

Deep Vein Thrombosis

Diabetic Ketoacidosis

Fever and Neutropenia in Children with Cancer

Fever Without Localizing Signs (FWLS) 0-60 Days

Fever Without Localizaing Signs (FWLS) 2-36 Months

Housewide Procedural Sedation

Hyperbilirubinemia

Neonatal Thrombosis

Nutrition/Feeding in the Post-Cardiac Neonate

Rapid Sequence Intubation

Skin and Soft Tissue Infection

Status Epilepticus

Tracheostomy Management

Urinary Tract Infection

Acute Chest Syndrome

Acute Gastroenteritis

Acute Heart Failure

Acute Hematogenous Osteomyelitis

Acute Ischemic Stroke

Acute Otitis Media

Appendicitis

Arterial Thrombosis

Asthma

Bronchiolitis

Cancer Center Procedural Management

Cardiac Thrombosis

Central Line-Associated Bloodstream Infections

Closed Head Injury

Community-Acquired Pneumonia

Cystic Fibrosis – Nutrition/GI >12 y/o

Autism Assessment and Diagnosis

C-spine Assessment

Intraosseus Line Placement

IV Lock Therapy

Postpartum Hemorrhage

Poll Question #2

In ambulatory settings, what is the best estimate for the percentage of questions for which evidence exists to answer clinical questions that affect the decision to treat?

17

a) 5%

b) 10%

c) 15%

d) 25%

e) 50%

f) Unsure or not applicable

Creating a foundation for data use





DEPLOYMENTSYSTEM

Operations

Informatics, Electronic Data Warehousing


Metadata: EDW Atlas Security and Auditing

Common, Linkable Vocabulary; Late binding

FinancialSource Marts

AdministrativeSource Marts

DepartmentalSource Marts

PatientSource Marts

EMR Source Marts

HRSource Mart

FINANCIAL SOURCES (e.g. EPSi,)

ADMINISTRATIVE SOURCES

(e.g. API Time Tracking)

EMR SOURCE (e.g. Epic)

DEPARTMENTAL SOURCES

(e.g. Sunquest Labs)

PATIENT SATISFACTIONSOURCES

(e.g. NRC Picker,

Human Resources(e.g. PeopleSoft)

TCH’s EDW Architecture

Operations• Labor

productivity• Radiology• Practice

Mgmt• Financials• Patient

Satisfaction• + others

Clinical• Asthma• Appendectomy• Deliveries• Pneumonia• Diabetes• Surgery• Neonatal dz• Transplant

Creating a foundation for QI deployment





DEPLOYMENTSYSTEM

Operations

Advanced Quality Improvement course, QI curriculum, Care process teams

Avenues for Dissemination

QUALITY LEADERS

National Programs and Partnerships

ADVANCEDClassroom (e.g. AQI Program, Six Sigma Green Belt)• Project Required

INTERMEDIATE

Online and Classroom (IHI Educational Resources, PEDI 101, EQIPP, Fellows College)• Project Required

BEGINNEROnline and Classroom (e.g. Nursing IMPACT (QI Basic). OJO Educational Resources, Lean Awareness Training)

NEWClassroom and Department (e.g. New Employee Orientation, e-Learning, Unit/Department-based training)

Changes that result in process improvement

IdeasAdapted from: The Improvement Guide: A Practical Approach to Enhancing Organizational Performance, 2nd Ed. Gerald J. Langley, Ronald D. Moen, Kevin M. Nolan, Thomas W. Nolan, Clifford L. Norman, and Lloyd P. Provost; Jossey-Bass 2009

Act

Plan

Study Do

Act

Plan

Study Do

Act

Plan

Study Do

Improvement

Pareto 80/20 Principle in Healthcare

TCH’s Care Process Analysis

Asthma

Amou

nt o

f Var

iatio

n

Size of Clinical Process

Bubble Size = Case Count

Improvement Opportunity:

Large processes with significant variation

Driving clinical care improvement: linking science, data management, operations

Permanent, integrated teams composed of clinicians, technologists, analysts and quality improvement personnel drive adoption of evidence-based medicine and achieve and sustain superior outcomes.

Clinical ProgramGuidelines centered on evidence-based care

#5 Care Process

#4 Care Process

#3 Care Process

#2 Care Process

#1 Care Process

MD Lead

MD Lead

MD Lead

MD Lead

MD Lead

Data Manager

Outcomes Analyst

BIDeveloper

DataArchitect

Application Service Owner

Clinical Director

Domain MD Lead

Operations Lead

Balanced scorecard-expanded visualizations

1. Care Process Defined

2. Current Literature Research

3. Individual Ratings

4. Aggregate Ratings5. Group Creates Final Scorecard

Severity Adjusted Variation

Option 1: Focus on Outliers – the prescriptive approach

Strategy eliminate the unfavorable tail of the curve (“quality assurance”)

Result Ithe impact is minimal

# of Cases

Excellent Outcomes Poor Outcomes

1.96 std

# of Cases

Mean


1 box = 100 cases in a year

Data Drives Waste Reduction:Alternative Approaches

28


# of Cases

Mean1 box = 100 cases in a year

Excellent Outcomes

# of Cases

Poor Outcomes

Option 2: Focus On Inliers – improving quality outcomes across the majority

Strategy Evidence and analytics applied through EBP clinical standards targets inlier variation

Result Shifting more cases towards excellent outcomes has much more significant impact

Alternative Approaches to Waste Reduction

29

Improving Cost Structure Through Waste Reduction

30

Ordering Waste Workflow Waste Defect Waste

Ordering of tests that are neither diagnostic nor contributory

Variation in Emergency Care wait time

ADEs, transfusion reactions, pressure ulcers, HAIs, VTE, falls,

wrong surgery

Care Redesign Methodology

31

Evidence against

CXR utilization in patients with known asthma, steroids in

bronchiolitis

Evidence equivocal

Hypertonic saline and bronchodilators in select patients with bronchiolitis

Evidence Supports

Quicker steroid delivery for status asthmaticus, goal

directed therapy for septic shock

32

51%

35%

0%

10%

20%

30%

40%

50%

60%

70%

80%O

ct.

10

No

v. 1

0

Dec. 1

0

Jan

. 1

1

Feb

. 1

1

Mar.

11

Ap

r. 1

1

May

. 1

1

Jun

. 1

1

Jul.

11

Au

g. 1

1

Sep

. 1

1

Oct.

11

No

v. 1

1

Dec. 1

1

Jan

. 1

2

Feb

. 1

2

Mar.

12

Ap

r. 1

2

May

. 1

2

Jun

. 1

2

Jul.

12

Au

g. 1

2

Sep

. 1

2

Oct.

12

No

v. 1

2

Dec. 1

2

Jan

. 1

3

Feb

. 1

3

Mar.

13

Ap

r. 1

3

Perc

en

tage

Month year

Asthma: Care Process Team Cohort, Percentage of Chest X-rays Ordered* (Oct. 2010 - Apr. 2013)

Feedback of rates to hospitalists and Emergency Center clinicians Order set

revisions

* Inpatient, Emergency Center (EC) and observation patients (Care Process Team cohort), P-Chart based upon EDW data extraction of 5/14/2013 (M& W).


33





wrong surgery

Patient presents to Emergency Dept (ED).

Patient registers

Patient waiting

Patient evaluated by triage nurse

Does patient have vomiting &/

or diarrhea

Triage nurse does the following:· Vitals

What is the patient’s level of

dehydration?

Severe dehydration

Mild or Moderate

dehydration

Put patient in ED room

Triage nurse does the following:· Give Zofran· Provide gatorade/pedialyte

Is the patient vomiting?

Evaluate per clinical symptoms

Follow TCH AGE clinical algorithm

Triage nurse does the following:· Nothing or give patient gatorade/

pedialyte

BEGIN

Patient waiting

Patient put in ED room

Patient evaluated by

nurse


Medical student

Patient evaluated by ED resident


ED fellow

Patient evaluated by ED attending

Is the patient ok for discharge?

Decision to admit patient

MD does admission

orders

ED secretary requests bed

Bed approved

Nurse-Nurse checkout occurs

Decision to discharge

patient

MD does discharge

orders

PCA checks vital signs

Nurse discharges

patient


Fellow/Attending does pre-

transfer check

Patient discharged home1

Patient transferred to inpatient bed2

Key:___ solid arrow indicates “yes”_ _ broken arrow indicates “no”

1 Outcome: Time in ED2 Outcome: Time to inpatient bed3 Outcome: Length of stay (LOS)4 Outcome: Revisit from ED discharge4 Outcome: Revisit from inpatient discharge

Flow chart of a patient with acute gastroenteritis through the TCH Emergency Department: Existing process

34

Modified: 7/21/2009Process map before EBG

Patient presents to Emergency Dept (ED).

Patient registers

Patient waiting

Patient evaluated by triage nurse

Does patient have vomiting &/

or diarrhea

Triage nurse does the following:· Vitals· Assess dehydration (Gorelick score)**

What is the patient’s level of

dehydration?

Severe dehydration

Mild or Moderate

dehydration

Put patient in ED room

Triage nurse does the following:· Give Zofran· Provide patient education on ORT· Initiate ORT· Give ORT tracking sheet**

Is the patient vomiting?

Evaluate per clinical symptoms

Follow TCH AGE clinical algorithm

Triage nurse does the following:· Provide patient education on ORT· Initiate ORT· Give ORT tracking sheet**

BEGIN

Patient waiting

Patient put in ED room


nurse


Medical student

Patient evaluated by ED resident


ED fellow

Patient evaluated by ED attending

Bedside nurse does the following:· Assesses dehydration (Gorelick score)**· Monitors progress on ORT tracking sheet**· Reemphasizes patient education on ORT

ED Fellow does the following:· Assesses dehydration (Gorelick score)**· Monitors progress on ORT tracking sheet**· Reemphasizes patient education on ORT· Determines patient disposition

Is the patient ok for discharge?

Decision to admit patient

MD does admission

orders

ED secretary requests bed

Bed approved

Nurse-Nurse checkout occurs

Decision to discharge

patient

MD does discharge

orders


Nurse discharges

patient


Fellow/Attending does pre-

transfer check

Patient discharged home1

Patient transferred to inpatient bed2

Key:___ solid arrow indicates “yes”_ _ broken arrow indicates “no”

** New process1 Outcome: Time in ED2 Outcome: Time to inpatient bed3 Outcome: Length of stay (LOS)4 Outcome: Revisit from ED discharge4 Outcome: Revisit from inpatient discharge

Flow chart of a patient with acute gastroenteritis through the TCH Emergency Deparment

34

Collect ORT tracking sheet

Modified: 5/9/2009 Process map after EBG


37





wrong surgery

38

Clinical Decision Support to Minimize Errors

Streamlining and Improving Processes and Operations to Minimize Errors

Value =

EC: Early administration of Dexamethasone

Expanding evidence based practice

-Provider and staff inservicing-Clinical decision support-Bridging a continuum for home care: second dose

10% decrease in TID

Inpatient: prolonged LOS

Evidence based approach to early medication weaning

12110997857361493725131

200

150

100

50

0

Observation

Indiv

idual V

alu

e

_X=27.4

UCL=70.9

LCL=-16.1

Baseline Improvement 1Improvement 2

12110997857361493725131

200

150

100

50

0

ObservationM

ovin

g R

ange

__MR=16.4

UCL=53.4

LCL=0

Baseline Improvement 1Improvement 2

5

1

11

I-MR Chart of CT - 1st q3h to d/ c by Phase

• 35% reduction in LOS• No change in 7 or 30 day readmission rate• No change in days of school/days of work missed• Direct variable cost ($60/hr)

The continuum: improved patient experience and outcomesImproved time to first beta agonist (ED or inpatient arrival) Increase chronic severity assessment

Improve accuracy Increase appropriate controller prescriptions Clinical decision support

Increase influenza vaccination rateIncrease number of culturally sensitive education encountersIncrease number of social work/ legal support encounters

• AAP use went from 20% to 44% in first cycle to 52% in second

• ACT use went from 0% to 30% in first cycle to 41% in second

• Severity classification went from 10% to 35% in first cycle to 54% in second

Registry Financial Score Card

43

Asthma CareOutcomes Dashboard

Financial conversations

48

2011 Q1 2011 Q2 2011 Q3 2011 Q4 2012 Q1 2012 Q2 2012 Q3 2012 Q4 2013 Q1 2013 Q2 2013 Q3 2013 Q4 2014 Q1 2014 Q2

-$5,000

-$4,000

-$3,000

-$2,000

-$1,000

$0

$1,000

$2,000

$3,000

$4,000

$5,000

f(x) = 291.61978021978 x − 3062.50549450549f(x) = 106.481318681319 x − 855.252747252747

Margin

Examples Demonstrating ROI

Improved clinical care Decreases in LOS Decrease in readmission rates Decreased unnecessary test utilization Millions in savings across several disease processes

Reducing waste by systematizing reporting EDW reports cost 70% less to build

Clinical operations tools allow global views for increased operational efficiency

49

Organizational direction for data

Data reporting

Data analytics

Decision support

Predictive analytics

Organizational evolution over time

-EMR clinical reports-Financial reports

-Shortening event to reporting time-Transforming data and translating to action

-Integrating best evidence into delivery system infrastructures-EMR based recommendations and alerts-Integrated plans of care across continuums

--Linking likelihood of outcomes to care decisions driven with realtime data-Predicting financial outcomes and linking to clinical decisions for populations of patients-Linking outcomes across infrastructures

Improved outcomes for our patients and our enterprise

Predictive analytics: High risk asthma

SHORT ACTING BETA AGONISTS6 to 9 SABA = 1 point≥ 10 SABA = 2 points

EC UTILIZATION1-2 ER = 1 point> 2 ER = 2 points

HOSPITALIZATION1 hospitalization = 1 point>= 2 hospitalizations = 4 pointsNUMBER PRESCRIBING PROVIDERS>= 3 different prescribing providers in 12 monthsone of above criteria met, add 1 point

PRIMARY CARE VISITSLast PCP visit > 6 months + one of above criteria met = add 1 point

INHALED CORTICOSTERIOD >= 6 ICS low dose canister equivalent refills, subtract 1 point

Age 1-5 , 4 of 5 belowGovernment insurance (Medicaid or CHIP): Q2 under health insurance information Financial barrier to meds :Answered Yes to Q4 under health insurance information Previous asthma hospitalization: Yes to Q2 under past history of asthma care Chronic Severity= Mild persistent Acute Severity= Mild Age 6+ All 3 of the following Government insurance (Medicaid or CHIP): Q2 under health insurance information Chronic Severity= Mild persistent Acute Severity= Mild Or All 3 of the following Government insurance (Medicaid or CHIP): Q2 under health insurance information Exercise induced asthma: Answered yes to exercise page 3 of TEDAS. Acute Severity= Mild

Targets: reduce ED visits, hospitalization, albuterol overuse, ICS non adherenceCritical data source: TCHP, TDSHS data

Lieu TA et al Am J Respir Crit Care Med. 1998 Apr;157(4 Pt 1):1173-80 Farber HJ, et al. Ann Allergy Asthma Immunol. 2004 Mar;92(3):319-28. Farber HJ. J Asthma. 1998;35(1):95-9Spitzer WO, et al. N Engl J Med 1992 Feb 20;326(8):501-6Suissa S, et al. Thorax. 2002 Oct;57(10):880-4.

Targets: reduce ED visits/ unscheduled PCP visitsCritical data source: TCH ED, PCP

Diabetes Pregnancy Asthma Transplant Pneumonia Appendicitis Newborn

Hospital Acquired Conditions

Sepsis and septic shock

Obesity

Transitions of care

Survey explorer

Care Process Teams

Additionally, completed a gap strategy for 38 “registries”

Assuring an excellent patient experience

QI education and culture change

Data/predictive analytics: measuring through meaningful metrics

Content System

Measurement System

Deployment System

ImprovedPopulation

Health Deployment

strategy—Care Process Teams

Evidence Integrated practice via

guidelines, order sets and measures

Using and innovating best

practices

Knowledge management for population health

#HASummit14

Analytic Insights

AQuestions &

Answers

#HASummit14 55

Session Feedback Survey

55

1. On a scale of 1-5, how satisfied were you overall with the Penny Wheeler, MD / Allina session?

2. What feedback or suggestions do you have for Penny Wheeler, MD / Allina session?

3. On a scale of 1-5, how satisfied were you overall with the Charles Macias / TCH session?

4. What feedback or suggestions do you have for the Charles Macias / TCH session?

The Imperative of Linking Clinical and Financial Data to Improve Outcomes - HAS Session 17

Healthcare

Transcript of The Imperative of Linking Clinical and Financial Data to Improve Outcomes - HAS Session 17