Embracing Patient-Centered Reforms
Ahead of Payment Reform
Avrim R Eden, MD, MBA Medical Director of Quality
Summit Medical Group, NJ
Summit Health Management
No speaker disclosures
About Us
• For-profit, physician owned, managed and governed
• C-Suite with physician CEO & Board
• Summit Health Management provides management services to Summit Medical Group
• Growth of 70+ providers/year
About Us
• 390 physicians 70 specialties
• 100+ APNs & PAs 1,600 employees
• 210,000 patients 75,000 visits/mo
• 14,000 urgent care visits/year
• 11,000 ambulatory surgeries/year
Our Population Health Team
Robert W Brenner, MD, MMM Chief Medical Officer
Jamie L Reedy, MD, MPH Director, Population Health
Avrim R Eden, MD, MBA Director, Quality & Utilization
Allen M Khademi, MD Director, Transitions of Care
Our Vision
To provide patient-centered, outcomes-driven care at lower cost to the sick and well populations for whom we are accountable.
Our Strategy
• Prepare for change from FFS to FFV and full-risk contracts
• Voluntary front-end investment with no immediate ROI
• In network with all major payers
• P4P commercial payer & Medicare contracts at 52% in 2014
• 42 acre main campus
• 32 satellites, 65+ PODS
• 500K SF, another 250K in development
NYC
About New Jersey
• Most densely populated state
• Highly fragmented – 300+ physicians / 100,000 residents
– High hospital & PAC densities
– No fully integrated health systems
– No market-dominant groups
– 8 medical schools within 1hr drive
Our Patients Have Many Options
• 11 Hospitals (majority at 4 hosp.)
• 90+ PAC facilities (majority at 17)
• Specialists galore in NJ & NYC
• Not all our PCP patients use our specialists & vice versa
• Many patients winter in the South
Fragmentation of care is our biggest challenge!
ED
PAC
Home
Circles of Unaccountability
UCC
Hospital
Physician
ED
PAC
Home
UCC
Hospital
Physician
Our Interventions Have Saved $77 MM
Home
ED
PAC
Home
UCC
Hospital
Physician
Transitions of Care Strategy
Home
Summary of Outcomes
• Easier than expected: reduce hospital utilization & readmits
• More difficult than expected: HTN, DM, chronic disease
• As expected: preventive & immunizations
Medicare Dollar, 2012
$12,250
$1,500 $1,300 $650
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
$14,000
Inpatient /Part A
Outpatient /Part B
Pharmacy /Part D
Other Svcs /Part B
Use Cost To Deploy Scarce Resources
Cost ($/patient)
Hospital Utilization KSF
• Continuity-of-Care vs. Shift model
– lower patient/physician ratio
– minimize physician handoffs!
– early consultations
– seamless TOC to our PAC teams & care managers
• Our own hospitalists with 100% team- & value-based incentives
0
100
200
300
400
500
600
700
Medicare* 267
Commercial* 73
* 6-month rolling average at our main hospital
Medicare 320
Commercial 235
34% Reduction in Days/1000 Days/1000
0
100
200
300
400
500
600
700
* 6-month rolling average at our main hospital
Avoided Days
Avoided Days
Days/1000 $77 MM Saved in Avoided Days
0.80
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
LO
S R
ati
o
US average 1.0
* 6-month rolling average at our main hospital
Medicare* 0.85
Commercial* 0.96
LOS Ratios Are Below US Average
Hospitalist Chess Analogy
• Enemy are outliers & re-admissions
• Opening gambit by physician is KSF!
• Middlegame with PAs
• Endgame is seamless TOC home/PAC
Use Predictive Models
1. Predict outliers (LOS >2SD)
10% outliers = 30% hospital days 90% admits = 70% days
2. Identify High-Risk-For-Readmissions
3. Alert TOC team re HRR discharges
Predictive Model for Readmissions
• Day 2: identify high-risk-for-readmit
• Keep them an extra day!
Example: 100 patients
20 high-risk patients +20 days
80 patients ½ day early -40 days
NET DAYS -20 days
0%
5%
10%
15%
20%
25%J
an
-09
Ap
r-09
Ju
l-09
Oct-
09
Jan
-10
Ap
r-10
Ju
l-10
Oct-
10
Jan
-11
Ap
r-11
Ju
l-11
Oct-
11
Jan
-12
Ap
r-12
Ju
l-12
Oct-
12
Jan
-13
Ap
r-13
Ju
l-13
Oct-
13
Jan
-14
30
-da
y R
ea
dm
it R
ate
NJ Medicare 19.6%
Commercial* 3.5%
Medicare* 8.6%
NJ Commercial 6.4%
* 6-month rolling average at our main hospital
30-Day Readmissions Below Benchmark
25
100% Cmltv 68% 85% 44%
4%
40%
24%
17% 15%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Same Day Week 1 Week 2 Week 3 Week 4
30-Day Readmits, Time to Readmit N=298
Adjusted
Workload
N
Scaled
SOI
Scaled
LOS
Ratio
Scaled
Readmit
Rate
Scaled
Mortality
Index
Hospitalist Performance Index
+ + + +
25% 20% 20% 20% 15% Weight
61%
68%
75%
81%
82%
0% 20% 40% 60% 80% 100%
E
D
C
B
A
Hospitalists
Medicare Dollar, 2012
Geriatric PAC Pilot, N=311
Annualized Avoided Readmissions
Avoided readmits Avoided days Avoided cost
65 300 $800,000
13%
18%
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
SMG Non-SMG
Readmissions
Embracing Patient-Centered Reforms
Ahead of Payment Reform
Takeaways!
• Team-based continuity-of-care
• Inpatient registry for hosp & PAC
• Use predictive models, think chess!
• Seamless handoffs & TOC
• Unblinded reports!
• Avoid individual FFS for IP & PAC
Summary of Outcomes
• Easier than expected: reduce hospital utilization & readmits
• More difficult than expected: HTN, DM, chronic disease
• As expected: preventive & immunizations
4Q
/13
3Q
/13
2Q
/13
1Q
/13
4Q
/12
3Q
/12
2Q
/12
1Q
/12
4Q
/11
3Q
/11
2Q
/11
1Q
/11
4Q
/10
3Q
/10
2Q
/10
1Q
/10
4Q
/09
3Q
/09
2Q
/09
1Q
/09
4Q
/08
20
07
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Historical by Quarter
% I
n C
on
tro
l
US/NJ 90 pctl (72% )
Controlling High Blood Pressure (CBP), N=10,335
BP<140/90
BP control is flat for past 4 years!
130120110100908070605040
250
200
150
100
50
Diastolic BP
Sy
sto
lic
BP
100
160
Normal
Pre-HTN
Stage1
Stage2
Hypertension Patients, Blood Pressure, N=10,335 4Q13
• Two BP readings w 20 point change
Δ-Better and Δ-Worse Groups
7,700 970 890
BP in Control, N = 9,550 Plank 1
• Average 64% BP in control
• 80% of Δ-Better patients
• 27% of Δ-Worse patients
Averages can deceive you
Accurate BP Measurement, N=398,000 Plank 1
0%
10%
20%
30%
40%
50%
0 1 2 3 4 5 6 7 8 9
Pe
rce
nt
of
Re
ad
ing
s
BP Last Digit
Bias to zero even after 4
re-education programs!
Stronger Adherence For 90-Day Refills Plank 2
Takeaways!
• Re-engage physicians, staff, mgmt
• Adopt medication algorithm
• 2-week follow-up BP
• Patient engagement
• Transparent monthly reports!
33%
44%
44%
50%
55%
55%
57%
61%
65%
67%
78%
79%
82%
86%
87%
92%
93%
0% 20% 40% 60% 80% 100%
n=15
n=70
n=9
n=6
n=51
n=114
n=21
n=162
n=152
n=100
n=116
n=70
n=50
n=29
n=76
n=185
n=40
BP in Control <140/90 mm Hg
N=3,451 (18-85 yrs) Avg=69% Target=82%
Jan 2014: Unblinded report after 1st month
Process Outcomes
Outcomes that matter to patients!
• A1c done • A1c <8 • strokes/1000 diabetics • AMI/1000 • blindness/1000 • amputations/1000
• Spirometry • FEV1<70% • fewer & shorter admits for COPD
Use NNT to Guide Priorities
Number needed to treat: for every patient with BP in control we avoid 1 event (AMI and stroke) in the next 5 years!
Hypertension
Male Female TOTAL WELL
VISITS
Example Population 10,000 10,000 20,000 20,000
80% BP in control 8,000 8,000
NNT 18 38 1,000
AMI/Strokes avoided 444 210 654 20
Embracing Patient-Centered Reforms
Ahead of Payment Reform
Medicare Dollar, 2012
Use “What if” to Prioritize Resources
1. Value to patient? Outcomes that matter!
2. Value to healthcare? Quality/Cost
3. Economically sustainable?
4. Scalable?
Scarcest resource is physician forbearance for more QI initiatives!
Metrics Are Not People!
The greatest danger is a decline
in empathy for our patients, and
becoming a computer-centric
medical home.
“The doctor will see you now.”
Significance of Our Story
If we can successfully improve quality and value of care in our most unfavorable fragmented environment, that is good news for groups in similar situations.