Brent James, M.D., M. Stat. Executive Director, Institute for Health Care Delivery Research Intermountain Healthcare Keynote Presentation. 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 pneumonia pulled all patients treated over a defined time period across all Intermountain inpatient facilities - typically 1 year identified and staged (relative to changes in expected utilization) - severity of presenting primary condition - all comorbidities on admission - every complication - measures of long term outcomes compared physicians with meaningful # of cases (low volume physicians included in parallel analysis, as a group)

1. iHT2
The Health IT Summit in Beverly Hills
Intercontinental Los Angeles Hotel, Beverly Hills, California
Wednesday, 7 November 2012 -- 11:25a - 12:10p
Health IT: The Critical Tool for
Managing Clinical Care
Brent C. James, M.D., M.Stat.
Executive Director, Institute for
Health Care Delivery Research
Intermountain Healthcare
Salt Lake City, Utah, USA

2. 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.

3. 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 pneumonia
pulled all patients treated over a defined time period
across all Intermountain inpatient facilities - typically 1 year
identified and staged (relative to changes in expected utilization)
- severity of presenting primary condition
- all comorbidities on admission
- every complication
- measures of long term outcomes
compared physicians with meaningful # of cases
(low volume physicians included in parallel analysis, as a group)

4. IHC TURP QUE Study
Median Surgery Minutes vs Median Grams Tissue
100 100
Grams tissue / Surgery minutes
80 80
60 60
40 40
20 20
0 0
M L K J P B C O N A I D H E G F
Attending Physician
Median surgical time Median grams tissue removed

5. IHC TURP QUE Study
Average Hospital Cost
2500 2500
2233
2140 2156
2000 1913 2000
1697 1662
1618 1598
1568 1552 1556
1500 1549 1543
1500 1500
Dollars
1269
1164
1000 1000
500 500
0 0
A B C D E F G H I J K L M N O P
Attending Physician

6. The opportunity (care falls short of its theoretic potential)
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. Unacceptablerates 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)

7. 50+% of all resource expenditures in
hospitals is
quality-associated waste:
recovering from preventable foul-ups
building unusable products
providing unnecessary treatments
simple inefficiency
Andersen, C. 1991
James BC et al., 2006

8. Total U.S. fiscal exposures
By 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.)
60,001.8
60
PV of Medicare Part D shortfall ($7,172.0 B)
50
PV of Medicare Part B shortfall ($17,165.0 B)
40
Trillion $
30
PV of Medicare Part A shortfall ($13,770.0 B)
20 Other explicit
PV of Social Security shortfall ($7,677.0 B) liabilities
($1,257.4 B)
10 Federal employee and veteran benefits ($5,283.7 B)
Federal debt securities ($7,582.7 B)
0
2009
Source: GAO. Financial Reports of the United States Government for the Years Ended September 30, 2009 and 2008.

9. The Fiscal Gap (unfunded federal obligations - 2009)
Unfunded obligations
Medicare
$38.1 trillion
National
Total Stimulus Defense TARP
National Debt $862 $714 $700
Social Security
$14.1 trillion billion billion billion
$7.7 trillion

10. Health care payments will be cut

11. We have found proven solutions
Dr. Alan Morris, LDS Hospital, 1991:
NIH-funded randomized controlled trial
assessing an "artifical lung" vs. standard ventilator management
for 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

12. 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)

13. Dr. Alan Morris, LDS Hospital, 1991
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 while
encouraging clinicians to vary based on individual patient needs;
then feeding back variation and patient outcome data in a
"learning system"

14. Sepsis bundle compliance
ER bundle ICU bundle All components
100 100
80 80
% compliance
60 60
40 40
20 20
0 0
n
n
n
n
p
p
p
l
l
l
08 ov
09 ov
10 ov
ay
ay
ay
ar
ar
ar
ar
Ju
Ju
Ju
Ja
Ja
Ja
Ja
Se
Se
Se
M
M
M
M
M
M
M
N
N
N
07
Month

15. Sepsis mortality - ER-ICU transfers
0.5 0.5
32 37 42 23 29 33 53 50 39 30 24 41 28 22 27 32 36 52 70 60 57 50 51 77 77 71 48 59 63 68 70 90 81 79 78 70 84
n= 28 44 45 42 34 41 45 38 47 31 34 40 35 27 28 24 44 39 51 65 47 52 61 43 73 65 69 52 46 68 63 94 75 69 81 82 74 91
0.4 0.4
Mortality rate
0.3 0.3
20.2%
0.2 0.2
0.1 0.1
8.0%
0 0
n
n
n
n
n
n
n
05 p
06 p
07 p
08 p
09 p
10 p
ay
ay
ay
ay
ay
ay
Ja
Ja
Ja
Ja
Ja
Ja
Ja
Se
Se
Se
Se
Se
Se
M
M
M
M
M
M
04
125+ fewer inpatient deaths per year Month

16. Lesson 1
We count our successes in lives ...

17. Lesson 2
Very often,
better care is cheaper care ...

18. Aligning financial incentives
Neonates > 33 weeks gestational age
who develop respiratory distress syndrome
Treat at birth hospital with nasal CPAP (prevents
alveolar collapse), oxygen, +/- surfactant
Transport to NICU declines from 78% to 18%.
Financial impact (NOI; ~110 patients per year; raw $):
Before After Net
Birth hospital 84,244 553,479 469,235
Transport (staff only) 22,199 - 27,222 - 49,421
Tertiary (NICU) hospital 958,467 209,829 -748,638
Delivery system total 1,064,910 736,086 -328,824
Integrated health plan 900,599 512,120 388,479
Medicaid 652,103 373,735 278,368
Other commerical payers 429,101 223,215 205,886
Payer total 1,981,803 1,109,070 872,733

19. 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

20. % Gross Domestic Product
15
25
0
5
10
20
19
60
148
19
65
19
70
357
19
75
19
80
1,106
19
85
19
90
2,281
19
95
Bending the cost curve
3,762
20
00
4,729
20
05
6,683
20
10
9,173
20
15
0
5
12,357
10
15
20
Total $ per US citizen (thousands)

21. Capitation makes a comeback
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").

22. Our answer:
A Shared Accountability Organization:
Physicians,
hospitals,
payers, and
patients
with aligned professional and financial incentives
to seek
the best medical result
at the lowest necessary cost

23. 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 undertreat
Payments targeted at break-even, most efficient cost of operations;
all upside $$ contained in shared savings
Hitting measured quality thresholds a prerequisite
to participate in shared savings
Involve employed and affiliated physician groups
via partner health plans

24. Process management is the key
higher quality drives lower costs
under capitation, all of the savings come
back to clinical process managers
more 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

25. Process management means health IT
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)

26. Better has no limit ...
an old Yiddish proverb