Center for State Health Policy New Jersey Safety Net ACOs: New Findings on Opportunities for Better Care and Lower Costs PICO New Jersey Conference Promoting Good Care: Innovation and Organizing in Health Care Newark, NJ September 19, 2013 Joel C. Cantor Rutgers Center for State Health Policy
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Center for State Health Policy
New Jersey Safety Net ACOs: New Findings on Opportunities for Better Care
and Lower Costs
PICO New Jersey Conference Promoting Good Care: Innovation and Organizing in Health Care
Newark, NJ September 19, 2013
Joel C. Cantor
Rutgers Center for State Health Policy
Center for State Health Policy
2
Acknowledgements
Project support from the Nicholson Foundation
Our project team – Sujoy Chakravarty, Ph.D. – Jian Tong, M.S.
Others contributing – Ping Shi, NJ Department of Health, Center for Health Statistics – Derek DeLia, Daisuke Goto, Jose Nova, Oliver Lontok, Bram Poquette,
and Dorothy Gaboda, CSHP
Center for State Health Policy
3
CSHP Roles
• Advise the NJ Medicaid ACO Demonstration Program – Review gainsharing plans – Annual evaluation
• CMMI-PICO High-Utilization Team Model – Adaptation of the Camden approach in Allentown PA, Aurora CO, San
Diego CA, and Kansas City MO • Evaluation of RWJF super-utilizer projects • Deep dive into NJ avoidable hospital cost
– Sources of variation across 13 low-income communities – Behavioral health connection to avoidable use and cost – Avoidable hospital use among Medicare-Medicaid dual eligibles – Dental use of the emergency department – Opportunities to enhance data
Center for State Health Policy
4
13 Candidate ACO Regions Camden* Greater Newark** Trenton*** Asbury Park-Neptune Atlantic City-Pleasantville Elizabeth-Linden Jersey City-Bayonne New Brunswick-Franklin Paterson-Passaic-Clifton Perth Amboy-Hopelawn Plainfield, North Plainfield Union City-W. NY- Guttenberg-N. Bergen Vineland-Millville
*Camden zip codes (08102, 08103, 08104 & 08105)
**Newark zip codes (07102, 07103, 07104, 07105, 07106, 07107,07108, 07112, & 07114) East Orange zip codes (07017, 07018) Irvington zip code (07111) Orange zip code (07050)
***Trenton zip codes (08608, 08609, 08611, 08618, 08629 & 08638)
Source: Kathe Newman, Rutgers University
Center for State Health Policy
5
Data and Measures • New Jersey All-Payer Uniform Billing Hospital Discharge Data:
2008-2010 • Five measures of potentially avoidable hospital use among
adults living in the 13 regions – Avoidable inpatient admissions – Avoidable treat-and-release emergency department (ED) visits – Inpatient high use – ED treat-and-release high use – 30-day all-cause readmissions
• Potential cost savings estimated by comparing each community to the region among them with the best cost performance
Center for State Health Policy
6
Center for State Health Policy
3,754
3,207 3,098
2,858
2,587 2,549
2,268 2,262 2,215 2,185
1,839 1,830 1,658 1,727
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
Rates of Avoidable Inpatient Hospitalizations
Rate per 100,000 population
7
2.3 Fold Variation
Med
ian
Reg
ion
Center for State Health Policy
Rates of Avoidable Emergency Department Visits
Rate per 100,000 population 8
51,871
40,876
34,124
30,104
23,582 21,486 20,478 19,684 19,472 18,912 18,423
16,827 15,028 14,177
0
10,000
20,000
30,000
40,000
50,000
60,000
3.5 Fold Variation
Med
ian
Reg
ion
Center for State Health Policy
Rates of Inpatient High Use
Rate per 100 hospital users
9
5.16 5.03 4.84
4.62 4.58
4.00 3.96 3.93 3.91 3.90
3.28 3.15 3.12
4.27
0.0
1.0
2.0
3.0
4.0
5.0
6.0 1.7 Fold Variation
Med
ian
Reg
ion
Center for State Health Policy
Rates of Treat-and-Release ED High Use
Rate per 100 hospital users
10
16.81
11.96 11.45
9.04 8.08
6.48 6.28 6.27 6.20 6.02 5.87 5.85
3.59
4.96
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0 4.7 Fold Variation
Med
ian
Reg
ion
Center for State Health Policy
30-Day All-Cause Readmission Rates
11
Age-sex adjusted rate per 100 ‘index’ (initial) hospitalizations
16.41 15.35
14.82 14.54 14.19 14.18 13.94 13.75 12.55 12.49 12.47 12.39 12.12
12.69
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
1.4 Fold Variation
Med
ian
Reg
ion
Center for State Health Policy
Few Patients are Both Inpatient and ED High Users
High users per 100 hospital users with high inpatient use (IP), high treat-and-release ED use, or both high IP and ED use. High inpatient use is defined as 4 or more stays over 2008-2010. High ED use is 6 or more visits over 2008-2010. The worst performing regions for these three measures are Asbury Park, Camden and Atlantic City. The best performing regions for the first measure is New Brunswick, and for the remaining two is Union City.
12
5.2
3.1 4.2 4.3
16.8
3.6
7.7
5.0
1.8 0.5 1.0 0.8
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
Worst Regions Best Regions 13 ACO Regions All NJ
Inpatient (IP) ED IP and ED
Center for State Health Policy
Very Different Payer Mix of Inpatient and ED High Users
13
Medicare, 51.7% Medicare, 60.6%
Medicaid, 10.3%
Medicaid, 5.3%
Private, 20.9% Private, 22.9%
Charity Care, 10.9% Charity Care, 6.1%
Self Pay, 4.8% Self Pay, 3.4%
Medicare, 10.6% Medicare, 14.7%
Medicaid, 16.8% Medicaid, 13.5%
Private, 29.8% Private, 33.3%
Charity Care, 18.4% Charity Care, 14.5%
Self Pay, 21.5% Self Pay, 20.2%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
13 ACO Regions All NJ 13 ACO Regions All NJ
Inpatient High Users ED High Users
Center for State Health Policy
Demographics also vary for Inpatient and ED High Users 13 ACO Regions
14
White 28.1%
Black 42.7%
Hispanic 20.2%
Other 9.0%
18-39 13.9%
40-64 39.5%
65+ 46.5%
Male, 45.7% Female, 54.3%
0% 20% 40% 60% 80% 100%
Race/Ethnicity
Age Group
Gender
Inpatient High Users
White 13.9%
Black 53.0%
Hispanic 26.0%
Other 7.1%
18-39 60.0%
40-64 34.4%
65+ 5.6%
Male, 35.4%
Female, 64.6%
0% 20% 40% 60% 80% 100%
ED High Users
High ED users are more likely to be women, younger, and minority compared to high inpatient users
Center for State Health Policy
15
Most Common Principal Diagnoses Chronic conditions common among inpatient users and often vague symptoms in the ED
Inpatient High Users ED High Users
Heart failure Other symptoms involving abdomen
and pelvis
Septicemia Symptoms involving respiratory
system and other chest symptoms
Diabetes mellitus Other and unspecified disorders of
back
Other forms of chronic ischemic heart disease
Asthma
Symptoms involving respiratory system and other chest symptoms
General symptoms
Center for State Health Policy
High Users Commonly have Behavioral Health Co-Morbidities
“Mental health” diagnoses includes substance use diagnoses Percentages represent proportion of high use inpatient stays or ED visits
16
38.6%
22.5% 20.2%
15.4%
0%
10%
20%
30%
40%
50%
Inpatient High Users ED High Users
Mental health Substance use disorder
13 ACO Regions Combined
Center for State Health Policy
17
Regions with Highest Savings Potential
Greater Newark $35.7
Jersey City, $14.5 Trenton, $9.4
Other Regions $34.2
Greater Newark $23.9
Trenton, $10.7 Camden, $9.4
Other Regions $17.3
Greater Newark $119.3
Jersey City, $51.9
Trenton $27.7
Other Regions $85.4
Greater Newark $23.6
Trenton, $11.9 Camden, $10.5
Other Regions $24.2
Greater Newark $36.8
Jersey City, $17.6 Trenton, $7.8
Other Regions $31.6
$0
$50
$100
$150
$200
$250
$300
Avoidable Hospitalizations $93.8 million
Avoidable ED Visits $61.3 million
Inpatient High Use $284.3 million
ED High Use $70.1 million
Readmissions $93.8 million
Mill
ions
of 2
010
Dol
lars
Center for State Health Policy
18
Regions with Highest Savings Potential among Medicaid Patients
Greater Newark $33.5
Jersey City-Bayonne $11.5
Trenton, $6.2
Other Regions $19.2
Greater Newark, $4.9 Trenton, $3.0 Camden, $2.0
Other Regions, $4.4
Greater Newark, $7.5
Jersey City-Bayonne , $3.1 Trenton, $1.5
Other Regions, $3.2
$0
$10
$20
$30
$40
$50
$60
$70
Inpatient High Use $70.4 million ED High Use $14.3 million Readmissions $15.4 million
Mill
ions
of 2
010
Dol
lars
Center for State Health Policy
Study Implications
• Wide variation across the 13 communities suggests improvement is achievable
– The best performing communities do about as well as state average – But on average, ACO regions perform much worse than state average
• Substantial hospital savings if the 13 communities achieved the cost profile of the best performing area among them
– $284 million from reduced inpatient high user costs (2010 $) – $155 million from reduced avoidable inpatient and emergency department costs – $94 million from reduced readmission costs – $70 million from reduced emergency department high user costs
• Payer mix and demographics different for inpatient and ED users – Potential savings greatest from reducing avoidable inpatient use – Hospital financial incentives vary, payment reform vital
19
Center for State Health Policy
Safety net ACO work is challenging!
• High rates of behavioral health problems • Patient engagement • Identifying patients for whom better care can lead to lower cost • Achieving true collaboration among hospitals, primary care,
behavioral health, social services and the community • Bringing Medicaid managed care organizations to the table
20
Center for State Health Policy
Thank You
21
Complete findings available at www.cshp.rutgers.edu
Related Documents
