Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... ·...

26
Findings from CCS Administrative Data Lee M. Sanders, MD, MPH Lisa J. Chamberlain, MD, MPH Stanford Center for Policy, Outcomes and Prevention (CPOP) CCS Redesign Stakeholder Advisory Board

Transcript of Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... ·...

Page 1: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

Findings from CCS Administrative Data

Lee M. Sanders, MD, MPH

Lisa J. Chamberlain, MD, MPH

Stanford Center for Policy, Outcomes and Prevention (CPOP)

CCS Redesign Stakeholder Advisory Board

Page 2: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

2

Analytic Guidance for CCS Program Reform

To use data to help protect the health of children with serious chronic illness.

1. To provide CCS and its stakeholders with data-driven analytic guidance to improve the quality and efficiency of care for children served by the CCS program.

2. To implement a coordinated strategy that bridges the gap between analytic activities and innovative care strategies in CCS subspecialty care centers.

Page 3: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

3

Essential Questions

How do we protect the health and well-being of a large population of children with serious chronic illness?

1. How do these children use health care services?

2. What is the quality (or appropriateness) of care received by this population?

3. What is the distribution of costs for that care?

Page 4: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

4

Analytic Design

Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

Total capture of all care episodesInpatient bed days

Outpatient visits (primary, subspecialty, non-MD)

ED visits

Home health and Durable Medical Equipment (DME)

Residential care

Pharmacy

Total capture of all CCS-related costs

Partial capture of non-CCS-related costs (FFS)

N = 323,922 children

Page 5: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

5

Stanford CPOPCCS Analytics Advisory Board

Out-Patient

Care Systems

CCS

Families

Foundations

Research

Policy

Hospitals

Advocacy

Advisory

Board

Bernardette Arrellano CCHA

Rich Cordova CHLA

Dr. Fran Kaufman CHLA

Dr. Bert Lubin CHORI

Richard Pan CA Assembly

Christy Bethell PhD OHSU

Neal Halfon PhD UCLA

Moira Inkelas PhD UCLA

Dylan Roby PhD UCLA

Meg Okumura, MD, UCSF

Dr. Robert Dimand State

Dr. Louis Girling Alameda

County

Dr. Mark Pian San Diego

Dr. Mary Doyle LA County

Maya Altman Health Plan

San Mateo

Dr. Melissa Aguirre Fresno

John Barry, OTR Shasta

County

Ted Lempert Children Now

Laurie Soman CRISSDr. Tom Klitzner UCLA Complex Care

Dr. David Bergman Stanford Complex Care

Chris Perrone CHCF

Dr. Ed Schor LPFCH

Teresa Jurado CCS, Health Plan San Mateo

Eileen Crumm PhD Family Voices

CCS Redesign Stakeholder Advisory Board

Page 6: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

6

CCS-enrolled Children: Social and Clinical Characteristics

%

Age – mean (SD) 7.3 (6.5) yearsSex - Female 43.0Race/EthnicityWhite 16.6Black 8.7Hispanic 56.4InsuranceMedicaid Managed care 47.6

Medicaid Fee for Service 19.6CHIP 7.5Mixed / Other 25.3Medical complexityComplex Chronic 51.4Non-complex Chronic 25.3Non-Chronic 23.3Diagnostic categoryNeurology 14.6Cardiology 12.6ENT / Hearing Loss 11.6Trauma / Injury 10.8Endocrine 6.8

CCS Redesign Stakeholder Advisory Board

> 2 organ systems, or progressive

* Simons TD, et al. Pediatric Medical Complexity Algorithm (PMCA), Pediatrics 2014

Page 7: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

7

Patterns of Care Visits per child per year

Children with > 1 visit

per year

Visits per child per year Mean (SD)

Outpatient Visits (MD) 94% 7.6 (8.6)

Outpatient Pharmacy Visits 87% 18.7 (28.7)

ED Visits 49% 1.6 (1.8)

Hospitalizations (Bed Days) 31% 14.8 (30.5)

Outpatient visits (Non-MD) 29% 10.2 (13.7)

Home health visits 16% 5.8 (13.3)

CCS Redesign Stakeholder Advisory Board

Page 8: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

8

Patterns of Care by Age

0

10

20

30

40

50

60

70

80

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Ageatfirstvisit

Medianencounterrateperyear

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

Numberofchildren

HospitalBedDays EmergencyDepartmentVisitsEarlyPeriodicScreening,Diagnosis&Treatment DentalVisitsHomeHealthVisits OutpatientclinicvisitsOtheroutpatientvisits PharmacyprescriptionsfilledNumberofchildren

CCS Redesign Stakeholder Advisory Board

Page 9: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

9

Patterns of Care by Medical Complexity

CCS Redesign Stakeholder Advisory Board

0

5

10

15

20

25

30

35

40

45

50

Complex Chronic Non-Chronic

Outpatient pharmacy fills

Other outpatient visits

Home health visits

ED visits

Outpatient physician visits

Bed days

Mean e

ncounte

r ra

te p

er

year

Page 10: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

10

Patterns of Care by Diagnostic Category

CCS Redesign Stakeholder Advisory Board

0

10

20

30

40

50

60

70

80

90

100

Neurolo

gy

Hemato

logy

Oncolo

gy

Cardio

logy

Otolary

ngology

Gastro

entero

logy

Pulmonary

Inju

ry

CCS eligible diagnosis

Perc

ent

of t

otal

exp

endi

ture

Inpatient Home health PharmacyOutpatient clinic Other outpatient DME% total expenditures % children

% o

f to

tal vis

its

Page 11: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

11

Outpatient: Inpatient Patternsby Diagnostic Category

CCS Redesign Stakeholder Advisory Board

CARD

ENT

ENDO

INJ

GASTRO

DEVT

HEME

NEONATAL

NEUR

ONC

OPTH

ORTH

URO

0 5 10 15 20

Mean bed day rate

6

8

10

12

Mea

n ou

tpat

ient

vis

it ra

te

Scatter plot by diagnosis group, Outpatient versus total LOS

Bubble sized based on number of children in each diagnosis group

Page 12: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

12

Patterns of Care Regional Variability

CCS Redesign Stakeholder Advisory Board

Page 13: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

13

Quality of Care:Potentially Preventable Hospitalizations

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

Bac

teria

l pne

umon

ia

Deh

ydra

tion

Asthm

a

Epile

psy

Kidne

y/ur

inar

y infe

ction

Sev

ere

ENT in

fect

ion

Cellulitis

Gas

troen

terit

is

Hyp

oglyce

mia

Mas

toiditis

Ane

mia

Nut

ritio

nal d

eficie

ncy

Imm

unizat

ion

prev

enta

ble

cond

ition

s

Tuber

culosis

Pelvic in

flam

mat

ory dise

ase

Diabe

tes

Stre

toco

ccal m

enin

gitis

Seizu

res

Failu

re to

thriv

e

Ambulatory Sensitive Condition

Nu

mb

er

of

Ch

ild

ren

Ho

sp

ita

lize

d

24.8% of all CCS

hospitalizations

CCS Redesign Stakeholder Advisory Board

Page 14: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

14

Quality of Care:No Care After Hospital Discharge

78.9

51.9

67.8

39

61.1

33.6

56.3

30.3

0 10 20 30 40 50 60 70 80 90

No MD visits within 7 days Post-hospitalization

No Outpatient Visit of any kind within 7 days Post-Hospitalization

No MD visits within 14 days Post-hospitalization

No Outpatient Visit of any kind within 14 days Post-Hospitalization

No MD visits within 21 days Post-Hospitalization

No Outpatient Visit of any kind within 21 days Post-Hospitalization

No MD visits within 28 days Post-Hospitalization

No Outpatient Visit of any kind within 28 days Post-Hospitalization

Percent of hospitalized CCS enrollees

(Overall Readmission Rate: 9.6%)

CCS Redesign Stakeholder Advisory Board

Page 15: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

15

Cost DistributionBy Child

50

2

40

26

5

15

4

32

1

25

0

10

20

30

40

50

60

70

80

90

100

Children Annual expenditures

Pe

rce

nt

Chart Title

CCS Redesign Stakeholder Advisory Board

Page 16: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

16

Cost Distribution By Medical Complexity

47%

27%

26%

Complex Chronic

Non-complexChronic

Non-Chronic

85%

6%

9%

Among All CCS Enrollees Among “High Cost” Children*

**Pediatric Medical Complexity Algorithm (PMCA), Mangione-Smith R. 2014

*Top 10% of annual expenses

Page 17: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

17

Cost DistributionBy Type of Care Chart Title

36%

21%

9%

17%

1%

14%

2%

0%

0%

Inpatient Home healthPharmacy Residential facilityOutpatient clinic Other outpatientDME DentalED

Hospital

Home Health

Outpatient (nonMD)

Residential Facility

Pharmacy

DME Emergency Care

Outpatient (MD)

Page 18: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

18

Cost Distribution, by Medical Complexity

Complex Chronic Non-complex/non-chronic

67.7

6.6

13.9

0.8

8.4

1.2

Inpatient Home health Pharmacy Outpatient clinic Other outpatient DME

67.7

6.6

13.9

0.8

8.4

1.2

Inpatient Home health Pharmacy Outpatient clinic Other outpatient DME

31.1

21.5

16.1

2.4

27.7

0.7

Inpatient Home health Pharmacy Outpatient clinic Other outpatient DME

CCS Redesign Stakeholder Advisory Board

Page 19: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

19

Cost by Diagnostic Category

13.8

10.7

5.93.9

7.12.52.62.93.3

31.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

Neurolo

gy/N

eurosu

rger

y

Hemat

ology

NICU

Cardiol

ogy/C

ardio

thora

cic Su

rgery

Oncolo

gy ENT

Gastro

entero

logy

Pulmonar

y

Gen Peds

/Behav

Devt

Exte

rnal/

Injur

y

CCS-Eligible Diagnosis

Pe

rce

nt

of

Tota

l Exp

en

dit

ure

sfo

r al

l

dia

gno

ses

0

200

400

600

800

1000

1200

1400

1600

1800

2000

Nu

mb

er

of

Ch

ildre

n

CCS Redesign Stakeholder Advisory Board

Page 20: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

20

Costs Distributionby Diagnostic Category

CCS Redesign Stakeholder Advisory Board

$0

$20,000,000

$40,000,000

$60,000,000

$80,000,000

$100,000,000

$120,000,000

$140,000,000

$160,000,000

Neu

rolo

gy

Hem

atol

ogy

EN

T

Onco

logy

Car

diolo

gy

Pulmon

ary

Gas

troe

ntero

logy

Inju

ry

Endoc

rinol

ogy

Tota

l ex

pen

dit

ure

s

Inpatient Outpatient pharmacy Residential facility Home health

Outpatient physician DME ED Other outpatient

Page 21: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

21

“High Cost Children”Over Time

CCS Redesign Stakeholder Advisory Board

Page 22: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

22

Cost Distributionby Hospital Type

CCS Redesign Stakeholder Advisory Board

OtherHospitals

1%

CountyHospitals

6%

UCSystem

Hospitals

13%

Non-Profit&For-

ProfitHospitals

25%

Free-Standing

Children's

Hospitals

55%

OtherHospitals

1%

CountyHospitals

8%

UCSystem

Hospitals

11%

Free-Standing

Children'sHospitals

37%

Non-Profit&

For-ProfitHospitals

43%

Infants (< 12 months)

Free-StandingChildren's

Hospitals

64%

Non-Profit&

For-Profit

Hospitals

18%

UCSystem

Hospitals

13% CountyHospitals

5%

OtherHospitals

<1%

Medically Complex Children

All Children

Page 23: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

23

Summary

• Distinct patterns of care use – particularly by age and medical complexity.

• Wide variability in care patterns, particularly before and after hospitalization.

• Costs are highly skewed, driven by inpatient and residential care, and persistent over time.

Page 24: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

24

Implications forCCS Program Reform• Care System Innovation

– Redesign Outpatient Systems to reduce Inpatient Care– Enhance Regionalized Subspecialty and Primary Care

• Population Health Management– Tier Care Coordination by Clinical Complexity– Build Regional Learning Collaboratives

• Public Policy and Payment Reform– Establish Risk Pools for Skewed Cost Distribution – Monitor and Evaluate Impact of Reforms

CCS Redesign Stakeholder Advisory Board

Page 25: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

25

CPOP Policy Briefs

CCS Redesign Stakeholder Advisory Board

https://cpopstanford.wordpress.com/our-work/state/

Page 26: Findings from CCS Administrative Datahealthpolicy.ucla.edu/Documents/Spotlight/Stanford... · Retrospective, population-based analysis of all paid claims for the CCS Program (2007-2012)

26

Thank You

CCS and DHCSCalifornia StakeholdersRobert Dimand, MDKatie Schlageter, AlamedaLouis Girling, MD, AlamedaMaya Altman, HPSMFiona Donald, MD; Anand Chabra, MDTeresa JuradoDavid Alexander, MDEd Schor, MDJuno DuenasEileen CrummLaurie SomanChristy Sandborg, MD, StanfordDavid Bergman, MD, StanfordJori Bogetz, MD; Doriel Pearson-NishiokaBert Lubin, MD, CHRCOTom Klitzner, MD, UCLAMoira Inkelas, PhD, UCLADylan Roby, PhD, UCLAMegie Okumura, MD, UCSFStanford University Center for

Policy, Outcomes and PreventionPaul Wise, MD, MPHJason Wang, MD, PhDVandana Sundaram, MPHEwen Wang, MDBen Goldstein, PhDMonica Eneriz-Wiemer, MDKeith van Haren, MDStafford Grady, MDSusan Fernandez, RN, PhDMyMy Buu, MDNathan Luna, MDRachel Bensen, MDStephanie Crossen, MDOlga Saynina, MSGene Lewitt, PhDMaureen Sheehan, RNRegan FoustSonja Swenson