Metrics & Scoring Committee April 22, 2016 Consent agenda - - PowerPoint PPT Presentation

Metrics & Scoring Committee

April 22, 2016

Consent agenda

*Approve February minutes

Agenda overview

• Updates
• Public testimony
• Measures by race, ethnicity, etc
• Health Equity Index development
• Health Share Presentation
• Committee discussion

• Committee applications open

through May 13th

• CY 2015 data in – CCOs will

start final validation May 1st

• Stakeholder survey in the

field through May 13th

Public testimony

MEASURES BY RACE / ETHNICITY / GENDER / AGE / GEOGRAPHY

Milena Malone

Health Equity Index

Sarah Bartelmann Kristen Rohde

History

May 2015: Committee received public testimony from Dr. Dannenhoffer suggesting a novel “meta-measure” that would:

– Measure the reduction of health disparities – Use already collected data and existing measures with large denominators – Incentivize CCOs to attain the same performance for the historically disadvantaged populations as they do for the overall CCO population

June 2015: Committee expressed interest in using this type of measure as a challenge pool measure in the future, requested additional information and measurement development from staff.

8

Today’s goal

• Share Health Equity Index development to date.
• Request Committee input on 3 high-level questions to

guide additional development.

• Avoid getting too far in the

weeds!

9

Step 1: convened workgroup

Health Analytics Office of Equity & Inclusion Program Design & Evaluation Services Transformation Center OHSU Center for Health Systems Effectiveness Oregon Health Care Quality Corp. FamilyCare | Health Share | PacificSource CareOregon

10

Step 2: reviewed what is known

Initial concept for the health equity index was based on key documents:

• National Quality Forum’s National Voluntary Consensus Standards

for Ambulatory Care – Measuring Healthcare Disparities (2008).

• Institute of Medicine. Access to Health Care in America: A Model for

Monitoring Access (1993).

• Institute of Medicine. Unequal Treatment: What Healthcare

Providers Need to Know about Racial and Ethnic Disparities in Health Care (2002).

• Health Affairs. Analysis Raises Questions on Whether Pay-For-

Performance In Medicaid Can Efficiently Reduce Racial And Ethnic Disparities (2011).

11

Step 3: developed a framework

12

Facets

Seeking Care Access to Provider Quality of Care Differential Treatment based on Needs Self‐Reported Health Status

Measures

Measure 1 Measure 2 Measure 3 Measure 4 Measure 5 Measure 6 Measure 7 Measure 8 Measure 9

Variable 1 Variable 2 Variable 3 Variable 4 Variable 5

Each measure in the composite could be stratified in a variety of ways, including, but not limited to:

• Race / ethnicity
• Language
• Gender
• SPMI
• Disability
• Geography
• Etc…

Composite will likely start with race/ethnicity at minimum, then expand to include other variables.

Step 4: identified parameters

MUST

• address the Medicaid population
• use available data
• be statistically sound
• be applicable to other populations (gender, disability, etc)
• have some way to tie performance to quality pool $

IDEALLY

• based on current CCO incentive / state performance metrics
• generate meaningful and actionable results
• be understandable
• allow for tracking over time

13

Many options with different implications

14

Question #1

Is the intent of the Index to:

• Reduce variation in performance across groups within a

CCO…

• Improve performance of all groups towards the

benchmark or target?

15

SBIRT: “variation”

16

SBIRT: “benchmark”

17

Considerations

VARIATION

• Measuring variation between

groups may not be the best measure of equity. All groups could do equally well / poorly.

• CCOs could improve their

“variation” score if performance worsens for some populations.

• May be hard to measure

trends over time. BENCHMARK/TARGET

• May be easier to interpret /

understand than variation.

• Does not measure variation

between groups.

• CCOs are already incentivized

for meeting the benchmark or improvement target through individual measures (but may leave some groups behind).

18

Question #2

Is the intent of the Index to:

• Reduce measurement burden by using existing CCO

incentive measures, regardless of their appropriateness for measuring disparities…

• Use measures that are more sensitive to identifying

disparities, even if they are not currently in use?

19

Question #2 (cont)

Report identified set of 76 measures that are “disparities- sensitive,” based on

• Prevalence of conditions
• Gaps in quality of care
• Community impact
• Communication challenges
• Clinical discretion
• Social determinants of health

20

http://www.qualityforum.org/Publications/2012/11/Healthcare_Disparities_and_Cultural _Competency_Consensus_Standards__Disparities‐Sensitive_Measure_Assessment.aspx

Question #3

Just because we can make an Index methodology work, should we?

Consider: what will motivate CCOs to engage in disparities reduction work? Incentives may motivate CCOs but measure needs to be “moveable” to see progress. Will a composite measure motivate behavior change?

21

Other considerations for Index

• Can we expect CCOs to move an aggregate

score within 12 months?

• Does single aggregate score have meaning?
• More statistical analysis may hinder

understanding.

• May not be meaningful to communities being

measured.

• May not reflect real clinical implications / known

risk factors.

22

Alternate suggestion #1

Instead of the Index, select specific measure(s) of disparities (using list of known disparities-sensitive measures, or other method) and adopt into measure set (or challenge pool).

23

Alternate suggestion #2

Instead of the Index, utilize the new core / menu measure set concept to require one (or more?) of the CCO menu measures be related to disparities. Allow CCOs flexibility to identify disparities within their own population and prioritize the measures that make the most sense for them (based on local community need, data, etc).

24

Next steps

Depending on Committee discussion today – workgroup can finalize recommendation (either Index methodology, or develop one of the alternate suggestions) and bring back to future meeting.

• Variation across v. improvement toward benchmark?
• Use existing metrics or use disparities-sensitive metrics?
• Just because we can build an Index, should we?

25

Health Share of Oregon Metrics and Equity Work

Quick Caveats:

All CCOs lack complete data about Race, Ethnicity and Language Our data and numbers may look slightly different from that reported by OHA (different sources) Timeframes of these data may be inconsistent

Health Share’s metrics and equity work

Metrics are increasingly familiar and well‐understood gauges of system health Use as a critical framework for identifying potential areas of inequity Develop a framework for assessing and making meaning from these performance differences (Disparities Data Analysis and Reporting Team)

Equity can be assessed in a number of ways

As it relates to

• Race & Ethnicity
• Language
• Gender
• Age
• Eligibility Category
• Geography
• Special Populations (foster kids, SPMI, refugees)
• Others…

Developmental Screening – CCO Level

55.2% 30% 35% 40% 45% 50% 55% 60% 65% 70% Oct'14 Dec'14 Feb'15 Apr'15 Jun'15 Aug'15

Performance by CCO

Developmental Screening—by Race

0% 10% 20% 30% 40% 50% 60% 70% Oct'14 Dec'14 Feb'15 Apr'15 Jun'15 Aug'15

Performance by Race

African American American Indian or Alaska Native Asian Caucasian Unknown 2014 Benchmark Target

0% 10% 20% 30% 40% 50% 60% 70% Oct'14 Nov'14 Dec'14 Jan'15 Feb'15 Mar'15 Apr'15 May'15 Jun'15 Jul'15 Aug'15 Sep'15

Performance by Language

Sorted by Current Rolling Performance % Descending

Spanish Undetermined English Other Vietnamese Chinese Russian Arabic Somali Burmese 2014 Benchmark Target

Developmental Screening—By Language

Developmental Screening—Making Meaning

Metric improvement approach  Gap reports, incentives to providers or families, tests of change Meaning‐making approach  Deeper dive on population:

• Of those not screened, many were engaged in primary

care Is the screening tool translated into other languages? What is the perception of refugees of screening tools? How to best engage these populations? Would improvement in Dev Screening address the root of the disparity? Or the system intended to help?

Developmental Screening—System Mapping

Foster Care Assessment Metric

Metric improvement approach get lists for assessments physical, mental and dental health care (we do that) Meaning making approach how well does the system respond to the actual health needs of the foster care population? How can we use the measures to monitor system performance? Interview providers, analyze the system, reveal the chronic health conditions of the population

Moving from difference to improvement

Interview PCP clinics and systems:

• Do you have programs or additional services for

foster children?

• Do you identify children in foster care?
• What differences do you see in chronic conditions for

your kids in foster care? Very few systems were doing anything differently with their foster kids population (but eager to help)

Show difference in overall prevalence of some conditions

Show differences among assigned membership (make the issue local)

Find approaches that address these differences

Advanced Primary Care (APC) Model for Foster Kids Beyond three assessments, shouldn’t these kids get special attention for:

• Developmental Screening
• AWCV
• Dental Sealants
• SBIRT
• ECU
• ED Utilization

Clinic level dashboards: multiple metrics for special populations

Reflections based on our experience

Should we use the metrics to point us toward potential system improvements? Or as ends in themselves?

Questions?

Next Meeting: April 22nd at 9AM - noon

Agenda

• Measures by race/ethnicity and language
• Health Share presentation on disparities work
• Information from OEI on language access
• Continued discussion of future program structure (TBD)