36C10G24R0038 Attachment E - 2023 BY22 Model Documentation and Analysis Report.pdf

VA Enrollee Health Care Projection Model Documentation Report Milliman, Inc.

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

DEPARTMENT OF VETERANS AFFAIRS

VETERANS HEALTH ADMINISTRATION

CHIEF STRATEGY OFFICE

2023 VA ENROLLEE HEALTH CARE

PROJECTION MODEL

- BASE YEAR 2022 -

2023 EHCPM DOCUMENTATION & ANALYSIS

DECEMBER 21, 2023

36C10G24R0038

ATTACHMENT E: VA ENROLLEE HEALTH CARE PROJECTION MODEL DOCUMENTATION REPORT

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

TABLE OF CONTENTS December 21, 2023

A Executive Summary A1 Executive Summary

B Enrollee Health Care Projection Model (EHCPM) – Overview & Tutorials B1 Overview of the Enrollee Health Care Projection Model B2 EHCPM Tutorial – Enrollment Projections B3 EHCPM Tutorial – Private Sector-Based Projections B4 EHCPM Tutorial – VA Experience-Based Projections B5 2023 EHCPM (BY22) Updates B6 Overview of Modeling for Community Care Services B7 Overview of BY22 Modeling Approach

C Enrollment Projections & Analyses C1 Enrollment Projection Methodology C2 Creation of the Master Enrollment File C3 Veteran Population Proxy Development C4 Enrollment Rates C5 County Consolidation Methodology C6 Priority and Morbidity Class Transition C7 Geographic Migration C8 Mortality Rates C9 Patient Projection Analysis C10 Enrollment and VetPop Proxy County Allocations

D Private Sector Based Service Projections D1 Private Sector-Based Service Benchmarks D2 Covered Benefits Package and Adjustments D3 Copay Levels and Adjustments D4 Geographic Area Adjustments D5 Age and Gender Adjustments D6 Inpatient Efficiency Analysis D7 Inpatient Hospital Under - and Over-Coding Analysis

E VA Experience Based Service Projections E1 Modeling Methodology for VA Experience-Based Service Projections E2 Special Modeling Adjustments for VA Experience-Based Service Projections

F Section Intentionally Left Blank

G VA Morbidity & Reliance Adjustments G1 Overview of Reliance and Morbidity Modeling G2 Morbidity Risk Weight Development

Milliman Health Practice Seattle

36C10G24R0038

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

G3 Enrollee Morbidity Analyses – Ages 65 and Over G4 Enrollee Morbidity Analyses – Under Age 65 G5 Enrollee Reliance Analyses – Ages 65 and Over G6 Enrollee Reliance Analyses – Under Age 65 G7 Morbidity Birth Year Cohorts G8 Morbidity and Reliance Drift Adjustments G9 Enrollment Duration Utilization Adjustment G10 Condition-Based Morbidity Adjustments G11 Enrollee Reliance Generational Shift

H VA Workload Data Manipulations H1 VA Inpatient Workload Data H2 VA Inpatient Encounter Workload Data H3 VA Ambulatory Workload Data H4 VA Prosthetics Workload Data H5 VA Pharmacy Workload Data H6 VA Dental Workload Data H7 FY 2022 Baseline Database Construction

I VA Unit Costs I1 VA Base Year Unit Cost Development I2 VA Inpatient Base Year Unit Cost Development I3 VA Ambulatory Base Year Unit Cost Development I4 VA Pharmacy Base Year Unit Cost Development I5 VA Unit Cost Age Band & Gender Adjustments

J EHCPM Calibration J1 Actual-to-Expected Utilization Adjustment J2 Unit Cost Reconciliation Methodology

K Policies, Populations, Events, and Initiatives K1 Overview: Policies, Population Events, and Initiatives K2 Mental Health Policy Assumptions K3 Regional and Facility Reviews and Adjustments K4 Trend Rate Assumptions K5 Enrollment Policy Assumptions K6 Women's Health Assumptions K7 Impact of Economic Forecasts K8 Inpatient Efficiency Assumptions K9 Outpatient Efficiency Assumptions K10 Pharmacy Efficiency Assumptions K11 Pipeline Drug Impact Policy Assumptions K12 Homeless Program Adjustments K13 Long-Term Services and Supports Policy Assumptions K14 Civilian Wage Policy Assumptions K15 Patient Aligned Care Team (PACT) Policy Assumptions K16 Telehealth Policy Assumptions K17 Community Care Adjustments K18 Hepatitis C Drug Impact Policy Assumptions K19 Dialysis and Related Services Assumptions

36C10G24R0038

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

K20 Pharmacy Tiered Copayment Assumptions K21 COVID-19 Pandemic Modeling K22 Post-9/11 Era Combat Veteran Enrollee Assumptions K23 PACT Act

L Special Projections to Support Capital, Strategic, and Workforce Planning L1 Projections to Support Planning and Special Analyses L2 EHCPM Global Relative Value Units L3 Health Systems Planning Category Projections L4 Ambulatory Health Systems Planning Category Development L5 Parent Facility Allocation Methodology

M Data Sources M1 2023 EHCPM (BY22) Data Sources

N Projections Outside of EHCPM N1 CHAMPVA Cost Projection N2 PCAFC Projection Model N3 Newborn Care Projections N4 Fertility Services Projections N5 Non-Enrolled Registered Patient and Other Than Honorable Projections N6 MHRRTP 3D Model N7 Readjustment Counseling Service Projection Model (RCSPM)

Appendix Glossary

36C10G24R0038

Section A1 (BY22) – Page 1 Executive Summary

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Section A1 (BY22)

The Department of Veterans Affairs (VA) has developed four actuarial models to support the VA health care budget and strategic and capital planning as well as assess the impact of potential policies and changes in a dynamic health care environment. The VA Enrollee Health Care

Projection Model (EHCPM), first developed in 1998, projects Veteran demand for VA health care services. Actuarial methodology for projecting institutional and non-institutional long-term care services was developed and integrated into the 2012 EHCPM (BY11). The remaining non-modeled services are foreign medical program, state Veterans homes, spina bifida, and non-

Veteran medical care.

The EHCPM’s methodology, supporting analyses, and assumptions are discussed in Sections B through M. The Civilian Health and Medical Program Veterans Administration (CHAMPVA)

Model, which was new in 2010, projects CHAMPVA beneficiaries and the claim costs associated with their health care. (The CHAMPVA Model is documented in Section N1.) The

Program of Comprehensive Assistance for Family Caregivers (PCAFC) Projection Model, first developed in 2016, provides enrollment and program cost projections for the stipend payment, CHAMPVA, mental health, and respite care benefits that make up the majority of total PCAFC costs. (The PCAFC model is documented in Section N2.) The Readjustment Counseling

Service (RCS) Projection Model, developed in 2021, projects future readjustment counseling utilization by month and year to assess how RCS utilization may develop over time, how RCS resources could be reallocated geographically, and estimate the budget required to meet the totality of projected demand for the RCS eligible population. (The RCS model is documented in

Section N7.)

OVERVIEW OF THE ENROLLEE HEALTH CARE PROJECTION MODEL

As an assumption-based demand projection model, the EHCPM projects future utilization and expenditures by making explicit assumptions about how specific utilization and expenditure patterns may differ from current patterns under various scenarios. The projections are supported by extensive research and analyses of the Veteran enrollee population and the drivers of demand for VA health care. VA program, field, and research staff provide expertise on program strategies and initiatives, the unique needs of the enrollee population, and the VA

36C10G24R0038

Section A1 (BY22) – Page 2

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

health care system. One of the EHCPM’s strongest features is that it can be modified easily, by changing any of its multiple assumptions, to reflect the VA health care system as it evolves.

The EHCPM projects enrollment, utilization, and expenditures for the enrolled Veteran population for 148 categories of health care services 21 years into the future. First, the EHCPM determines how many Veterans will be enrolled in VA in each projection year and their age, priority, and geographic location. Next, the EHCPM projects the total health care services needed by those enrollees and then estimates the portion of that care those enrollees will demand from VA. Finally, total health care expenditures are developed by multiplying the expected VA utilization by VA unit costs.

The EHCPM produces enrollment, utilization, and expenditure projections for Post-9/11 Era

Combat Veterans using a force-deployment scenario originally developed by the Congressional

Budget Office and subsequently adjusted for the Administration’s force withdrawal policy.

The total enrollee demand for VA health care projected by the EHCPM can be reported separately for care expected to be provided in VA facilities and care expected to be purchased in the community. However, the proportion of total care that will be provided in VA facilities and purchased in the community can vary significantly depending upon eligibility criteria, operational guidelines, and resource availability.

The EHCPM is updated and enhanced each year, and Section B5 of this report describes enhancements to the 2023 EHCPM (BY22). The BY22 reference indicates that FY 2022 is the

EHCPM base year. Section B7 of this report describes the modeling approach that was used in the 2022 and 2021 EHCPM to account for the effects of the COVID-19 pandemic on FY 2020 and FY 2021 workload The 2023 EHCPM (BY22) largely returns to the typical modeling process used prior to the COVID-19 pandemic.

OVERVIEW OF THE CHAMPVA PROJECTION MODEL

CHAMPVA provides medical coverage to the spouse or widow(er) and to the children of a

Veteran, also referred to as a sponsor, who is rated permanently and totally disabled from a service-connected disability, was rated permanently and totally disabled from a service-connected condition at the time of death, died of service-connected disability, or died on active

36C10G24R0038

Section A1 (BY22) – Page 3

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

duty and the dependents are not otherwise eligible for Department of Defense TRICARE benefits.

The CHAMPVA Model was developed using multiple data sources provided by the Veterans

Health Administration (VHA). Data from fiscal years (FY) 2012 to 2022, publicly available research, and input from a development team (including individuals from VHA and VHA’s

CHAMPVA program) were used to develop model assumptions.

The CHAMPVA Model consists of two major components: the enrollment model and the claim cost model. The enrollment model projects the number of beneficiaries and sponsors enrolled in

CHAMPVA, whereas the claim cost model projects the CHAMPVA program’s total costs.

The enrollment model projects the number of CHAMPVA beneficiaries. Sponsors are projected for each fiscal year and then the beneficiaries of those sponsors are projected. Within a given year, sponsors are projected by age, gender, degree of service-connected disability, living status (whether the sponsor is living or deceased), and the sponsor’s enrollment lag (the number of years a sponsor delays enrolling a beneficiary), whereas beneficiaries are projected by age, beneficiary type, and gender (if the beneficiary is a spouse). Beneficiaries eligible for

CHAMPVA as a primary caregiver enrolled in the Program of Comprehensive Assistance for

Family Caregivers (PCAFC) (i.e., are not also an eligible spouse or child of a sponsor) are not modeled in the CHAMPVA projection model. Such caregivers are projected as part of the

PCAFC Model.

The claim cost model is driven by several factors, including enrollment counts produced from the enrollment model, assumed annual claim cost trends, age/gender factors, and actual

CHAMPVA paid claims data. The projected number of beneficiaries from the enrollment model then is linked to the claim cost model and aggregated into total expenditures. (A more in-depth look at the CHAMPVA Model can be found in Section N1.)

OVERVIEW OF THE PCAFC PROJECTION MODEL

The PCAFC and the Program of General Caregiver Support Services are collectively referred to as VA’s Caregiver Support Program. The PCAFC began in May 2011 to compensate the caregivers of Veterans who both separated from active duty after September 11, 2001, and met the requirements for PCAFC. Starting in October 2020, the PCAFC expanded to include eligible

36C10G24R0038

Section A1 (BY22) – Page 4

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Veterans who were injured in active duty prior to May 7th, 1975. As of October 1, 2022, the

PCAFC is available to the remaining cohort of Veterans injured between May 8, 1975, and

September 10, 2001. For eligible Veterans, their primary caregivers are eligible for a range of benefits; the monthly stipend payment, CHAMPVA benefits, mental health benefits, and respite care benefits are modeled.

In addition, the model results include projections for Veteran sponsor counts, caregiver counts, and stipend payment counts by fiscal year for each of the four eras of Veteran service. Cost projections produced by this model are not projections for the total costing of the PCAFC or the overarching Caregiver Support Program’s funding needs. (A more in-depth look at the PCAFC

Projection Model can be found in Section N2.)

OVERVIEW OF THE RCS PROJECTION MODEL

The Office of Readjustment Counseling Services (RSC) oversees the Vet Centers which are community-based counseling centers that provide a wide range of social and psychological services, including professional counseling to eligible Veterans, service members, including

National Guard and Reserve components, and their families. The model was developed in 2021 after the passage of the Commander John Scott Hannon Veterans Mental Health Care

Improvement Act of 2019 (Public Law 116-171); this law established that RCS report resources required to meet unmet need and predictions and trends in demand for care. The model was developed to assist with the required reporting from the new law and with resource planning.

The model projects future readjustment counseling utilization using two different approaches: an experience-based utilization projection and an anticipated demand-based projection. The model projects Vet Center utilization for 10 years in the future by a number of dimensions: service location, type of service, mode of service and demographic details. (Additional details of the

RCS Projection Model can be found in Section N7.)

MODEL LIMITATIONS

This communication and any associated analysis or data were prepared to provide assistance to the Department of Veterans Affairs. These deliverables are solely for the use and benefit of the Department of Veterans Affairs and may not be appropriate for other uses.

36C10G24R0038

Section A1 (BY22) – Page 5

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Milliman recognizes that materials it delivers to the Department of Veterans Affairs may be public records subject to disclosure to third parties, however, neither the Department of

Veterans Affairs nor Milliman intend to benefit and assume no duty or liability to any third parties who receive Milliman's work and may include disclaimer language on Milliman’s work product so stating. To the extent that Milliman's work is not subject to disclosure under applicable public records laws, the Department of Veterans Affairs agrees that it shall not disclose Milliman's work product to third parties without Milliman's prior written consent; provided, however, that the

Department of Veterans Affairs may distribute Milliman's work to (i) its professional service providers who are subject to a duty of confidentiality and who agree to not use Milliman's work product for any purpose other than to provide services to the Department of Veterans Affairs, or

(ii) any applicable regulatory or governmental agency, as required. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

This analysis relies in part on data and other listings provided by various personnel at the

Department of Veterans Affairs. That data has been reviewed for reasonableness and compared to past data submissions and other information, when possible. The information has not been audited by Milliman for accuracy. If the data or other listings are inaccurate or incomplete, this analysis may also be inaccurate or incomplete.

Models used in the preparation of our analysis were applied consistently with their intended use.

We have reviewed the models, including their inputs, calculations, and outputs for consistency, reasonableness, and appropriateness to the intended purpose and in compliance with generally accepted actuarial practice and relevant actuarial standards of practice (ASOP). The models, including all input, calculations, and output may not be appropriate for any other purpose.

Where we relied on models developed by others, we have made a reasonable effort to understand the intended purpose, general operation, dependencies and sensitivities of those models. We relied on input, review, and validation by other experts in the development of our models.

The results contained in these reports are projections based on modeling assumptions and historical data. Actual results will differ for many reasons. For example, it is impossible to determine how world events will unfold. Those events that impact the economy and the use of the nation's military may have a profound impact on enrollment and expenditure projections. In addition, many of the modeling variables are assumed to be constant over time, which may not

36C10G24R0038

Section A1 (BY22) – Page 6

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

match actual events. This analysis has not attempted to present results for all possible or even all likely outcomes. Emerging experience should be continually monitored and adjustments are made, as appropriate.

Section B1 (BY22) – Page 1 Overview of the Enrollee Health Care Projection Model

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Section B1 (BY22)

This section provides an overview of the methodology used to develop the enrollment, utilization, and expenditure projections in the VA Enrollee Health Care Projection Model

(EHCPM). It also provides an overview of the modeled health care services.

MODELING OVERVIEW

VA developed the EHCPM to assess Veteran demand for VA health care services over a

21-year planning period. The EHCPM projects the number of Veterans expected to be enrolled during each year as well as their priority, age, gender, special conflict status, and geographic location. The EHCPM then projects the total health care needs of the enrolled Veteran population and determines the portion of care that they are expected to receive from VA versus their other health care options.

VA uses the 21-year projections from the EHCPM to support VA health care budget development and capital and strategic planning. The EHCPM also projects the impact of policy scenarios and legislative initiatives on enrollment, utilization, and expenditures. VA can look at different views of the future because the EHCPM is built at a very detailed level and accounts separately for all major drivers of utilization and cost, based on changes in:

• Veteran enrollment;

• Enrollee morbidity and mortality;

• Enrollee reliance on VA versus other health care providers;

• Health care access policies;

• Broader health care policy environment; and

• VA health care system.

VA is at risk for providing health care to the enrolled Veteran population; therefore, the utilization and expenditure projections are developed for the enrolled Veteran population, not just those enrolled Veterans who are currently VA patients. Many policies and legislative proposals also have the potential to change enrollee health care demand. The EHCPM can easily provide alternative views of the future under different assumptions about enrollee VA health care

36C10G24R0038

Section B1 (BY22) – Page 2

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

demand because the projections are based on enrollees. Patient projections also are produced, and the methodology is discussed in Section C9, Patient Projection Analysis.

Utilization and expenditure projections are developed based on where enrollees live to support population-based, long-term planning efforts. The results based on enrollee residence then are allocated to VA facilities, largely following current practice patterns, but also including adjustments to account for expected future changes by facility. The modeled changes by facility include program changes at existing facilities, as well as the impact of new or closing VA facilities. This methodology is discussed in Section L5, Parent Facility Allocation Methodology.

Exhibit B1–1, VA EHCPM Methodology Flow Chart, provides a graphic description of the

EHCPM methodology.

EHCPM DETAIL

The EHCPM projects Veteran enrollment, utilization, and expenditures by geographic and demographic level. This section outlines these projection levels by sector/submarket/market/VISN, age group, gender, priority, enrollee type, special conflict status, Maintaining Internal Systems and Strengthening Integrated Outside Networks (MISSION) eligibility status, and service location (VA direct or community care)

Sector/Submarket/Market

Geographically, projections are developed at the submarket level which are an aggregation of sectors. Sectors are either large counties or groupings of smaller counties and are designed to ensure a large enough Veteran population in a given geographic area to produce credible projections. The 3,000-plus counties in the United States are mapped into 582 sectors. Sectors are then aggregated to 148 submarkets, which are contiguous counties made up of one or more sectors.

The submarket-level projections then are aggregated to 97 markets and the market-level projections are aggregated to the 18 Veterans Integrated Service Networks (VISN). Market and submarket boundaries were identified by VISNs to aid strategic and capital planning.

Section B1 (BY22) – Page 3

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Age Groups and Gender

Projections are developed for five-year age bands: under age 25, ages 25 to 29, ages 30 to 34, …, ages 85 to 89, and ages 90 and over. Projections also include gender detail. This detail level ensures that the EHCPM can capture the impact of aging and gender on enrollment and health care services utilization.

Priority

Projections are developed for Priorities 1a, 1b, 2, 3, 4, 5, 6, 7, and 8. (Section C2 Creation of the Master Enrollment File discusses the distinction between Priorities 1a and 1b.)

Enrollee Type

Projections are developed separately for enrollees who used VA health care before the

Veterans’ Health Care Eligibility Reform Act of 1996 (enrollees pre) and enrollees who enrolled after eligibility reform (enrollees post) because their morbidity and reliance profiles are significantly different. Enrollees pre have higher morbidity and greater reliance on VA health care and use more of VA’s special program services than enrollees post after accounting for differences in demographics such as age and priority.

Special Conflict Status

Special conflict status enables VA to model separately for enrollees of specific engagements to account for their unique morbidity and reliance on VA health care services. As of the 2021

EHCPM (BY20), the only special conflict tracked in the EHCPM is Post-9/11 Era Combat

Veteran (Post-9/11 CV) status. Post-9/11 CV status includes those that served in Operation

Enduring Freedom (2001–2014), Operation Iraqi Freedom (2003–2011), Operation New Dawn

(2010–2011), and other subsequent, related, and subordinate combat operations including

Operation Inherent Resolve (2014–present).

MISSION Eligible Enrollees

The EHCPM was expanded in BY18 to include specific projections for the MISSION geographic access eligible enrollees, those enrollees who are eligible for community care based on their

36C10G24R0038

Section B1 (BY22) – Page 4

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

proximity to primary (30-minute drive time) and specialty care (60-minute drive time), and those grandfathered under the Choice Act. In BY19, the EHCPM expanded the Choice Act grandfathered status to specify whether they were strictly grandfathered based on living in one of the 5 states with the lowest population density or recently using VA health care services, or loosely grandfathered based on living in an area where enrollees were predominantly eligible under the Choice Act. The latter category was identified as likely to be treated as eligible under the best medical interest provision of MISSION and on the convention that VHA tends to use its discretion and flexibility to maintain eligibility to groups that previously had it. The criteria used was to identify enrollees residing in Sectors where at least 45% of enrollees were Choice Act eligible under the distance criteria (40-mile) by June 2019.

This enables the EHCPM to account for the unique reliance behavior for the MISSION geographic access eligible enrollees under the MISSION Act. There are seven statuses available for differentiating utilization assumptions in the projection: Grandfathered Strict - 5

State, Grandfathered Strict - User, Grandfathered Loose, Drive Time Eligible Primary and

Specialty, Drive Time Eligible Primary, Drive Time Eligible Specialty, and Not Drive Time

Eligible.

Community Care

The EHCPM was expanded in BY15 to include projections with separate detail for VA direct and community care. This enables the EHCPM to account for several modeling factors specific to

VA direct and community care separately. (Section B6, Overview of Modeling for Community

Care Services discusses the differences in modeling.)

ENROLLMENT PROJECTION METHODOLOGY OVERVIEW

The EHCPM produces several types of enrollment projections:

• Unique Enrollment: Enrollees alive during any part of the year. This is the highest enrollment number for any given year. Enrollment projections are generally reported using unique enrollment.

• Average Enrollment: The amount of time during the year that a Veteran is enrolled

(measured in life years). For example, a Veteran who enrolls three months into the

36C10G24R0038

Section B1 (BY22) – Page 5

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

fiscal year would count for nine months, or 0.75 life years. The average enrollment projections are used in developing the utilization projections.

• Beginning of Fiscal Year: Enrollees who are alive at the beginning of the fiscal year.

• End of Fiscal Year: Enrollees who are alive at the end of the fiscal year.

• Suspended Enrollment: The number of Priority 8 Veterans who would have enrolled if enrollment had not been suspended in 2003.

The enrollment projections begin with VetPop data, 31-year projections of the Veteran population produced by the VA Nation Center for Veterans Analysis and Statistics (NCVAS) office. At this time, VetPop does not provide Veteran projections by priority, so VetPop data are combined with other data sources to create VetPop proxy data, which provide Veteran projections by priority.

Historical enrollment data are analyzed to develop enrollment rates by priority, age band, geographic area, and special conflict status. The enrollment rates then are applied to the enrollment pool, which is VetPop minus the enrolled Veteran population, to determine projected enrollees for any given year.

Mortality rates specific to age, gender, and priority then are applied to the enrollee population, and the enrollment and potential enrollee pool are aged one year at the end of each fiscal year to arrive at the projections for the beginning of the next fiscal year. The process of applying enrollment and mortality rates then repeats for the duration of the enrollment projections.

The EHCPM also accounts for geographic migration and enrollees who transition between enrollment priorities. (The development of the enrollment projections is discussed in detail in

Section B2, EHCPM Tutorial⎯Enrollment Projections.)

UTILIZATION AND EXPENDITURE PROJECTION METHODOLOGY OVERVIEW

VA provides many health care services comparable to those provided in the private sector and some health care services unique to VA. The utilization and expenditure projections for these services are developed using two approaches with the same basic methodology, but with several key differences. Exhibit B1–2 identifies the modeling approach used for each modeled health care service.

Section B1 (BY22) – Page 6

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Private Sector-Based Methodology⎯Many VA services are comparable to private sector services, for example, inpatient medical and surgical care, ambulatory care, and prescription drugs. (Section B3, EHCPM Tutorial⎯Private Sector-Based Projections discusses this approach in detail.)

• Projections for these services are developed using national utilization rates reflecting the health care provided in the United States in the commercial and Medicare markets. These utilization rates then are adjusted for the demographics of the enrollee population and characteristics of the VA health care delivery system.

• The large volume of available private sector commercial and Medicare data enables the EHCPM to project health care service utilization at a very detailed level. The data’s volume and breadth support a robust assessment of the impact of age, gender, copayments, and health care benefit structure on health care utilization.

VA Experience-Based Methodology⎯Special program ambulatory and non-acute bed services, ambulatory mental health services, prosthetics type services, dental services, over-the-counter pharmacy, pharmacy supplies, medication therapy management, compensation and pension exams, nutritional counseling, recreational therapy, and home telehealth and long-term services and supports fall under this methodology. (Section B4, EHCPM Tutorial⎯VA Experience-Based Projections discusses this approach in detail.)

• Projections for these services are developed using national utilization rates developed from VA workload data. The national VA utilization rates also are adjusted for differences in age/gender, morbidity, and geographic area based on a VA workload data analysis.

• The small volume of VA workload data for many of these services limits the detail level at which credible projections can be developed.

The EHCPM Tutorials use examples that walk the reader step-by-step through the methodology for developing utilization and expenditure projections. Many sections of the 2023 EHCPM

(BY22) report refer to the EHCPM Tutorial examples. As an example, Section J1, which discusses the development of Actual-to-Expected factors, contains this reference:

Section B1 (BY22) – Page 7

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

OVERVIEW OF MODELED HEALTH CARE SERVICES

The EHCPM projects health care services in two service definition structures: health care service categories (HSC) and health systems planning categories (HSPC). Exhibit B1–3 presents the modeled HSC and HSPC services and the utilization projection basis for each.

This section discusses the differences between the HSCs and HSPCs as well as the ways VA workload aligns with both structures. (The workload data manipulations performed to assign

VA’s workload experience into HSCs and HSPCs are discussed throughout Sections H and L.)

Many HSC and HSPC services are identical and reflect aggregations of detailed VA workload:

dental services, pharmacy, some mental health and homeless outpatient programs, inpatient mental health, special VA program inpatient care, LTSS home- and community-based services

(LTSS HCBS), and some LTSS facility-based services (LTSS FBS). (See Exhibits B1–4 through

B1–19 for details.)

The projected HSC and HSPC services are identical for acute inpatient services. However, VA acute inpatient workload is recorded by treating specialty, or location of care, and the acute inpatient HSCs and HSPCs are defined by diagnoses. Modeling acute inpatient services using the Diagnosis-Related Group (DRG)-based HSCs allows projections to reflect how the type of care provided in the acute inpatient environment is expected to change over time. (See Exhibits

B1–11 and B1–12 for details of the HSC categorization.)

LTSS HSCs were included in the EHCPM starting in BY10. (See Exhibit B1–10 for the LTSS

HCBS mapping and Exhibit B1–9 for the LTSS FBS mapping.) LTSS HCBS is identified and mapped by clinic stop, fee purpose of visit (FPOV) code, HCPCS code, and care category.

EHCPM Tutorial Exhibit B3–1: Office Visit Projection Example

• The methodology described above generates an Actual-to-Expected adjustment factor of 0.924 for office visits in the example.

• In the office visit example, 0.924 is entered into the utilization formula on line “t.”

Section B1 (BY22) – Page 8

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

LTSS FBS HSCs and HSPCs are defined by treating specialty; some are further defined by treating facility, diagnosis code, and data type.

The most significant categorization and counting differences between HSCs and HSPCs occur in ambulatory services, excluding the VA special outpatient mental health programs discussed above. VA records each ambulatory visit in one or more clinic stops that represent the location where care was provided, such as a primary care clinic stop. VA also records the Healthcare

Common Procedure Coding System (HCPCS) workload associated with each clinic stop.

The ambulatory HSCs represent groupings of HCPCS services. HCPCSs are used to report medical procedures and services. Modeling ambulatory services using the HCPCS-based HSCs allows the projections to reflect how the type of services, tests, and procedures provided in a clinic stop are expected to change over time. The HCPCS workload recorded in each clinic stop is assigned to HSCs based on the nature of the service provided.

To support strategic and capital planning, the ambulatory primary care, specialty care and mental health HSC projections (defined by HCPCSs) are allocated to HSPC projections (defined by provider specialty). (Exhibits B1–4 and B1–14 detail the transition from VA workload to HSCs to HSPCs. The ambulatory HSPC allocation process is discussed in Section L4.)

Non-modeled services include foreign medical program, spina bifida, and non-Veteran medical care.

Inpatient Workload Allocation Exhibit

Exhibit B1–21 illustrates how FY 2022 inpatient treating specialty days were assigned to HSCs based on DRGs. This exhibit lists FY 2022 inpatient bed days by treating specialty down the rows. The columns demonstrate the ultimate number of bed days assigned to each HSC. For example, of the total 21,840 bed days provided under treating specialty 2 (cardiology), 17,607 bed days were assigned to medical-related DRGs, 4,146 bed days were assigned to surgical-related DRGs, and so on. (The mapping of DRGs to inpatient HSCs is included in Exhibit B1–

20.)

Section B1 (BY22) – Page 9

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

Ambulatory Workload Allocation Exhibit

Exhibit B1–22 illustrates how the FY 2022 HCPCS workload was assigned to HSCs based on the nature of the service provided. This exhibit lists FY 2022 clinic stops down the rows. The columns demonstrate the ultimate number of CPT-based procedures from the clinic stop assigned to each HSC. For example, clinic stop 203 (Audiology) had a total of 1,897,852 stops, and 3,168,706 CPT-based procedures were recorded within those clinic stop visits. These CPT-based procedures then were assigned to HSCs, including 51,750 office visits (Office

Visits/Urgent Care/Physical Exams–Non-Mental Health in the exhibit), 1,624,345 Hearing Aid

Services, and so on. (The mapping of HCPCSs to ambulatory HSCs is included in Exhibit B1–

20.)

MODEL LIMITATIONS

This communication and any associated analysis or data were prepared to provide assistance to the Department of Veterans Affairs. These deliverables are solely for the use and benefit of the Department of Veterans Affairs and may not be appropriate for other uses.

Milliman recognizes that materials it delivers to the Department of Veterans Affairs may be public records subject to disclosure to third parties, however, neither the Department of

Veterans Affairs nor Milliman intend to benefit and assume no duty or liability to any third parties who receive Milliman's work and may include disclaimer language on Milliman’s work product so stating. To the extent that Milliman's work is not subject to disclosure under applicable public records laws, the Department of Veterans Affairs agrees that it shall not disclose Milliman's work product to third parties without Milliman's prior written consent; provided, however, that the

Department of Veterans Affairs may distribute Milliman's work to (i) its professional service providers who are subject to a duty of confidentiality and who agree to not use Milliman's work product for any purpose other than to provide services to the Department of Veterans Affairs, or

(ii) any applicable regulatory or governmental agency, as required. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

This analysis relies in part on data and other listings provided by various personnel at the

Department of Veterans Affairs. That data has been reviewed for reasonableness and compared to past data submissions and other information, when possible. The information has

36C10G24R0038

Section B1 (BY22) – Page 10

This report and associated databases were prepared solely to provide assistance to the Department of Veterans Affairs. Neither the Department of Veterans Affairs nor Milliman assume any duty or liability to other parties who receive this work. Milliman recommends any recipient be aided by its own actuary or other qualified professional when reviewing the Milliman work product.

not been audited by Milliman for accuracy. If the data or other listings are inaccurate or incomplete, this analysis may also be inaccurate or incomplete.

Models used in the preparation of our analysis were applied consistently with their intended use.

We have reviewed the models, including their inputs, calculations, and outputs for consistency, reasonableness, and appropriateness to the intended purpose and in compliance with generally accepted actuarial practice and relevant actuarial standards of practice (ASOP). The models, including all input, calculations, and output may not be appropriate for any other purpose.

Where we relied on models developed by others, we have made a reasonable effort to understand the intended purpose, general operation, dependencies and sensitivities of those models. We relied on input, review, and validation by other experts in the development of our models.

The results contained in these reports are projections based on modeling assumptions and historical data. Actual results will differ for many reasons. For example, it is impossible to determine how world events will unfold. Those events that impact the economy and the use of the nation's military may have a profound impact on enrollment and expenditure projections. In addition, many of the modeling variables are assumed to be constant over time, which may not match actual events. This analysis has not attempted to present results for all possible or even all likely outcomes. Emerging experience should be continually monitored and adjustments are made, as appropriate.

36C10G24R0038

Exhibit B1–1 VA EHCPM Methodology Flow Chart

A. Enrollment Projections

Note: In addition to the data inputs indicated, EHCPM assumptions may be adjusted to reflect the expected impact of VA policies and initiatives.

VA

VetPop

Post-9/11 Era Combat Veteran Deployment and

Veteran Projection

VA

Enrollment &

User Data

Veteran Enrollee Mortality

Assumptions

Develop Historical and Projected Enrollment Pools

Develop Enrollment Rates

Apply Enrollment Rates & Anticipated Enrollment Policy to Projected Pools

Apply Mortality and Age of Enrollees for Each Projection Year

Enrollment Projections to B-2 to Cto B-1

A.

Projected

Enrollment

B.

Projected Utilization

Rates

C.

VA Direct Care or

Community Care Trended

Projected Health Care

Expenditures

Apply Geographic Migrations and Priority Transitions

Exhibit B1-1 Flow Chart - Page 1

36C10G24R0038

VA EHCPM Methodology Flow Chart

B-1. Private Sector-Based Utilization Projections

CDPS*

CDPS

CDPS

HEI

HEI**

HEI

Note: In addition to the data inputs indicated, EHCPM assumptions may be adjusted to reflect the expected impact of VA policies and initiatives.

*CDPS: Chronic Illness and Disability Payment System (a risk-adjustment grouper) **HEI: Hospital Efficiency Index (developed by Milliman)

Adjust Private Sector Utilization Rates & Unit Costs for MBP &

Adjust Utilization & Unit Costs for Enrollee Age & Gender Mix

Assess VA Relative Morbidity & Adjust Utilization

Assess VA DoCM and Adjust Utilization & Unit Costs

Assess Enrollee Reliance on VHA & Adjust Utilization

Assess Residual Model Differences & Adjust Utilization

Trend Utilization & Unit Costs

VA Medical Benefit

Package (MBP), Copay Levels

VA

Workload

VA Enrollee Age & Gender

VA Patient Diagnosis

Data

VA Inpatient Workload with Diagnoses & Procedures

A.

Private Sector Acute

Utilization & Unit Cost Average by Geographic Area

Commercial

Medicare Diagnosis & Claims Data

VA/Medicare Data Match

Inpatient

LOS

Benchmarks

Community Loosely

Managed Levels

VA

Enrollee Surveys

A.

Projected

Enrollment

B.

Projected Utilization

Rates

C.

VA Direct Care or Community Care Trended

Unit Costs

Projected Health Care Expenditures

Enrollment Projections

Projected Utilization Rates

Medicare Allowable & Community Billed

Charges by Submarket to C to C

Unit Cost Process Unit Cost Process

Exhibit B1-1 Flow Chart - Page 2

36C10G24R0038

VA EHCPM Methodology Flow Chart

B-2. VA Experience-Based Utilization Projections

Note: In addition to the data inputs indicated, EHCPM assumptions may be adjusted to reflect the expected impact of VA policies and initiatives.

VA Experience-Based Projection Model

Enrollment Projections

A.

VA

Workload

VA

Enrollment

Data

VA Experience-Based Services & LTSS Utilization

Rates

A.

Projected

Enrollment

B.

Projected Utilization

Rates

C.

VA Direct Care or

Community Care Trended

Projected Health Care

Expenditures

Exhibit B1-1 Flow Chart - Page 3

36C10G24R0038

Exhibit B1-1 VA EHCPM Methodology Flow Chart

C. Unit Costs By Submarket

Note: In addition to the data inputs indicated, EHCPM assumptions may be adjusted to reflect the expected impact of VA policies and initiatives.

Develop Unit Costs by Facility

Develop Submarket, Age, and Gender specific unit costs using HCG based Medicare Allowable relativities for Private Sector-

Calculate Unit Cost Reconciliation Adjustments

Projected VA Unit Costs by Submarket

A.

Projected

Enrollment

B.

Projected Utilization

Rates

C.

VA Direct Care or

Community Care Trended

Projected Health Care

Expenditures

VA Unit Cost Data (MCA)

VA

Workload

VA Base Year Budget

Obligations

A. B-1

B-2

CMS RBRVS RVUs for Physicians and Milliman Outpatient

Hospital RVUs

Exhibit B1-1 Flow Chart - Page 4

36C10G24R0038

Modeled Service Private Sector-

Based

VA

Experience-

Based

AMB PC & SC - DIAGNOSTICS

Cardiovascular X Colonoscopy X Dermatology Services and Diagnostic Exams X Hearing/Speech Exams X Miscellaneous Medical Services and Diagnostics X Non-Invasive Vascular Studies X Ophthalmology Services and Diagnostic Exams X Pathology X Pulmonology Svcs, Diag Exams, Ventilator Mgmt X Radiology - CT X Radiology - General X Radiology - Mammography Diagnostic X Radiology - Mammography Screening X Radiology - MRI X Radiology - Nuclear Medicine X Radiology - PET & PET/CT X Radiology - Radiation Oncology X Radiology - Ultrasound X Vision Exams X

AMB PC & SC - E&M SVCS

Case Management - Rehabilitation Therapists X Compensation & Pension Exams X Medication Therapy Management X Office Visits/Urgent Care/Physical Exams - Non MH X Telephone E&M Services - Primary and Specialty Care X

AMB PC & SC - PROF SVCS AND PROCS

Ambulance X Ambulatory Surgery - ASC Setting X Ambulatory Surgery - Office Setting X Ambulatory Surgery - Outpatient Setting X Emergency Room Visits X Eye Glasses…