Health IT: The Critical Tool for Managing Clinical Care

Brent C. James, M.D., M.Stat.Executive Director, Institute for Health Care Delivery ResearchIntermountain HealthcareSalt Lake City, Utah, USA

iHT2

The Health IT Summit in Beverly HillsIntercontinental Los Angeles Hotel, Beverly Hills, California

Wednesday, 7 November 2012 -- 11:25a - 12:10p

Disclosures

Neither I, Brent C. James, nor any family members, have any relevant financial relationships to be discussed, directly or indirectly, referred to or illustrated with or without recognition within the presentation.

I have no financial relationships beyond my employment at Intermountain Healthcare.

Quality, Utilization, & Efficiency (QUE)Six clinical areas studied over 2 years:- transurethral prostatectomy (TURP)- open cholecystectomy- total hip arthroplasty- coronary artery bypass graft surgery (CABG)- permanent pacemaker implantation- community-acquired pneumoniapulled all patients treated over a defined time period

across all Intermountain inpatient facilities - typically 1 yearidentified and staged (relative to changes in expected utilization)- severity of presenting primary condition- all comorbidities on admission- every complication- measures of long term outcomescompared physicians with meaningful # of cases

(low volume physicians included in parallel analysis, as a group)

IHC TURP QUE StudyMedian Surgery Minutes vs Median Grams Tissue

M L K J P B C O N A I D H E G F0

20

40

60

80

100

0

20

40

60

80

100

Attending Physician

Median surgical time Median grams tissue removed

Gra

ms

tissu

e / S

urge

ry m

inut

es

IHC TURP QUE Study

1500 1549 1568 16181543

1697

1913

22332140 2156

1598

12691164

1552 15561662

A B C D E F G H I J K L M N O PAttending Physician

0

500

1000

1500

2000

2500

Dol

lars

0

500

1000

1500

2000

2500

Average Hospital Cost

1. Well-documented, massive, variation in practices (beyond the level where it is even remotely possible that all patients are receiving good care)

2. High rates of inappropriate care

3. Unacceptable rates of preventable care- associated patient injury and death

4. A striking inability to "do what we know works"

5. Huge amounts of waste leading to spiraling prices that limit access (46.6 million uninsured Americans)

The opportunity (care falls short of its theoretic potential)

50+% of all resource expenditures in hospitals is

quality-associated waste:recovering from preventable foul-upsbuilding unusable productsproviding unnecessary treatmentssimple inefficiency

Andersen, C. 1991James BC et al., 2006

60,001.8

2009

0

10

20

30

40

50

60

Trill

ion

$

Total U.S. fiscal exposuresBy layering on future obligations, the total net prevent value (PV) of debt rises

to over $60 trillion -- about $195,000 for every man, woman and child in the U.S. More than two-thirds of the shortfall arises from health care delivery.)

Source: GAO. Financial Reports of the United States Government for the Years Ended September 30, 2009 and 2008.

Federal employee and veteran benefits ($5,283.7 B)

Federal debt securities ($7,582.7 B)

PV of Social Security shortfall ($7,677.0 B)

PV of Medicare Part A shortfall ($13,770.0 B)

PV of Medicare Part B shortfall ($17,165.0 B)

PV of Medicare Part D shortfall ($7,172.0 B)

Other explicit liabilities

($1,257.4 B)

The Fiscal Gap (unfunded federal obligations - 2009)

Social Security$7.7 trillion

Medicare$38.1 trillion

TotalNational Debt

$14.1 trillion

Stimulus$862

billion

NationalDefense$714

billion

TARP$700

billion

Unfunded obligations

Health care payments will be cut

NIH-funded randomized controlled trialassessing an "artifical lung" vs. standard ventilator managementfor acute respiratory distress syndrome (ARDS)

discovered large variations in ventilator settings across and within expert pulmonologists

created a protocol for ventilator settings in the control arm of the trial

Implemented the protocol using Lean principles (Womack et al., 1990 - The Machine That Changed the World)- built into clinical workflows - automatic unless modified- clinicians encouraged to vary based on patient need- variances and patient outcomes fed back in a learning loop

Dr. Alan Morris, LDS Hospital, 1991:We have found proven solutions

Challenges building guidelines

Lack of evidence for best practice- Level 1, 2, or 3 evidence available only about 15-20% of the time

Expert consensus is unreliable- experts can't accurately estimate rates using subjective recall

(produce guesses that range from 0 to 100%, with no discernable pattern of response)- what you get depends on whom you invite (specialty level, individual level)

Guidelines don't guide practice- systems that rely on human memory execute correctly ~50% of the time (McGlynn: 55% for adults, 46% for children)

Results:survival (for ECMO entry criteria patients) improved from 9.5% to 44%costs fell by ~25% (from $160k to $120k)physician time fell by ~50%

we generalized the concept: Shared Baseline protocols ("bundles") to standardize care whileencouraging clinicians to vary based on individual patient needs;then feeding back variation and patient outcome data in a "learning system"

Dr. Alan Morris, LDS Hospital, 1991

07 Ja

nMarMay Ju

lSep Nov

08 Ja

nMar May Ju

lSep Nov

09 Ja

nMarMay Ju

lSep Nov

10 Ja

nMar

Month

0

20

40

60

80

100

% c

ompl

ianc

e

0

20

40

60

80

100

ER bundle ICU bundle All components

Sepsis bundle compliance

04 Ja

nMay Sep

05 Ja

nMay Sep

06 Ja

nMay Sep

07 Ja

nMay Sep

08 Ja

nMay Sep

09 Ja

nMay Sep

10 Ja

n

Month

0

0.1

0.2

0.3

0.4

0.5

Mor

talit

y ra

te

0

0.1

0.2

0.3

0.4

0.5

Sepsis mortality - ER-ICU transfers

20.2%

8.0%

125+ fewer inpatient deaths per year

2832

4437

4542

4223

3429

4133

4553

3850

4739

3130

3424

4041

3528

2722

2827

2432

4436

3952

5170

6560

4757

5250

6151

4377

7377

6571

6948

5259

4663

6868

6370

9490

7581

6979

8178

8270

7484

91n=

We count our successes in lives ...

Lesson 1

Very often,

better care is cheaper care ...

Lesson 2

Aligning financial incentivesNeonates > 33 weeks gestational age

who develop respiratory distress syndromeTreat at birth hospital with nasal CPAP (prevents

alveolar collapse), oxygen, +/- surfactantTransport to NICU declines from 78% to 18%.Financial impact (NOI; ~110 patients per year; raw $):

Birth hospitalTransport (staff only)

Tertiary (NICU) hospitalDelivery system total

Integrated health planMedicaid

Other commerical payersPayer total

Before 84,24422,199

958,4671,064,910

900,599652,103

429,1011,981,803

After 553,479

- 27,222 209,829736,086

512,120373,735

223,2151,109,070

Net 469,235

- 49,421 -748,638-328,824

388,479278,368

205,886872,733

Current payment mechanisms

Actively incent overutilization: do more, get paid more - even when there is no health benefit

I am paid to harm my patients (paid more for complications)

Actively disincents innovation that reduces costs through better quality (a key success factor for the rest of the U.S. economy)

Very strong, deep, wide evidence showing exactly this effect throughout U.S. healthcare

Bending the cost curve

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

2015

0

5

10

15

20

25

% G

ross

Dom

estic

Pro

duct

0

5

10

15

20

Tota

l $ p

er U

S ci

tizen

(tho

usan

ds)

2,281

4,729

3,762

6,683

9,173

12,357

148 357 1,106

1. ACOs, AMHs, bundled payment, shared savings, pay for value: sophisticated forms of capitation- provider at (financial) risk ... but with far better data systems for

(1) quality measurement and (2) risk adjustment

2. Represent "managed care at the bedside"- ask clinical teams at the bedside to manage the care, not distant

and disengaged insurance companies

3. More than 80% of cost saving opportunities live on the clinical side; 70+% of clinical improvement activities reduce costs by freeing up care delivery capacity (technically, "fixed cost leverage").

Capitation makes a comeback

Our answer:

A Shared Accountability Organization:

Physicians,hospitals,payers, and

patients

with aligned professional and financial incentivesto seek

the best medical resultat the lowest necessary cost

Some key elements:Pay first dollar, not last dollar

(defined contribution, not defined benefit; reference payment)

Whoever makes the consumption decision bears the (appropriate) financial consequences (patients and physicians have skin in the game)

No incentive to risk-select patients (community-rated premiums, but risk-adjusted capitation payments)

Levers: No incentives to overtreat or undertreatPayments targeted at break-even, most efficient cost of operations;

all upside $$ contained in shared savingsHitting measured quality thresholds a prerequisite

to participate in shared savingsInvolve employed and affiliated physician groups

via partner health plans

Process management is the keyhigher quality drives lower costsunder capitation, all of the savings come

back to clinical process managersmore than half of all cost savings will

take the form of unused capacity (fixed costs:empty hospital beds, empty clinic patient appointments, and reduced procedure, imaging, and testing rates)

balanced by increasing demand(Baby Boom; obesity; community growth; technological advances; may still require some capacity management / reduction)

major financial model shift, from revenue enhancement to cost control

key difference: it takes a team

1. Identify a high-priority clinical process (key process analysis)

2. Build an evidence-based best practice protocol(always imperfect: poor evidence, unreliable consensus)

3. Blend it into clinical workflow (= clinical decision support; don't rely on human memory; make "best care" the lowest energy state, default choice that happens automatically unless someone must modify)

4. Embed data systems to track (1) protocol variations and (2) short and long term patient results (intermediate and final clinical, cost, and satisfaction outcomes)

5. Feed those data back (variations, outcomes) in a learning loop - constantly update and improve the protocol- provide true transparency to front-line clinicians- generate formal knowledge (peer-reviewed publications)

Process management means health IT

Better has no limit ...

an old Yiddish proverb