Health Economics Spotlight

Updates on VA data, health economics research, and analytic methods

May 2025

In This Issue

1. Crosswalk VA and DoD health care services using HERC categories of care
2. Seminar: Empirical Bayes
3. HERC datasets updated with FY24 data
4. Facility-level average costs per day
5. Bridging the digital divide for older adults

Spotlight

Crosswalk VA and DoD health care services using HERC categories of care

A new HERC technical report crosswalks Department of Defense (DoD) and VA care location data using HERC categories of care. This crosswalk enables better tracking of clinical services across institutions, supporting seamless continuation of care. It also facilitates comparisons of service utilization and costs, which can ultimately improve decision-making and policy development for both DoD and VA health systems.

To understand health care cost and utilization data, researchers often group encounters into clinical categories of care. HERC creates categories of care for VA inpatient and outpatient data using bedsection codes (treating specialties in MCA data) and clinic stop codes respectively. DoD’s Defense Health Agency uses MEPRS codes for health care cost accounting, similar to VA’s bedsection and clinic stop codes. In Technical Report 45, the authors mapped MEPRS codes to HERC categories of care codes by carefully reviewing the code descriptions and applying clinical logic to determine equivalency or alignment across systems. The technical report includes tables with VA stop codes or bedsections and DoD MEPRS codes grouped by HERC categories of care.

VA and DoD have distinct missions and objectives, and it is crucial for data users to consider these differences before comparing systems. In addition, given the systems’ distinct missions, there are certain services unique to each system (e.g., medical readiness screenings in DoD and home-based primary care at VA). Before making cross system comparisons, data users should make sure to account for differences in the system goals, service types, and cost structures.

Technical Report 45: Mapping A Crosswalk of DoD and VA Inpatient and Outpatient Health Services Utilization Codes with HERC Patient Care Categories is available on the HERC website.

Seminar

Empirical Bayes

HERC Econometrics Seminar

Wednesday, May 28 at 2pm ET

Register

 
Dave Chan, MD, PhD
Health Economist, Health Economics Resource Center, and Staff Physician, VA Palo Alto Health Care System 
Professor of Economic Analysis and Policy, UC Berkeley

This seminar will provide an introduction to empirical Bayes. When we have a finite sample of observations and increasingly rich covariates, empirical Bayes provides a useful tool to form better policy-relevant predictions than those from standard regression methods. We will discuss the rationale behind empirical Bayes, connections with the machine learning literature, and illustrative examples from the applied economics literature.

Please note the date of this seminar has changed to May 28. All seminars are open to VA participants only until further notice. 

Resources

HERC datasets updated with FY24 data

Labor Costs
The VHA Labor Cost dataset can be used to determine the cost of staff time. For each type of personnel, the database gives the hourly labor costs from the MCA Office Account-Level Budgeter Cost Center (ALBCC) database and the Financial Management System (FMS). Data include the personnel type, total labor costs, workload, and hourly labor costs. The labor cost dataset is saved as an excel file in the Researcher’s Guide to Estimating VHA Labor Costs (VA intranet only).

Wage Index
Health care costs are more expensive in geographic areas that have higher wages (e.g., Boston or San Francisco). Therefore, researchers may need to adjust their cost analyses for these wage differences. The best-known method involves using the Medicare wage index. To assist VA researchers, HERC creates a wage index specific to VA facilities. The wage index, updated for FY24 is saved as an excel file in the guidebook Medicare Wage Index for VA Facilities.

Nosos
Nosos is the VA-specific risk score designed to predict costs. These annual risk scores reflect all the costs and care in a given year. These scores can be used for risk adjustment in analyses using VA data. Learn more about Nosos in HERC's Guide to the Nosos Risk Adjustment Score. Nosos, HERC Average Cost data, and the Discharge Dataset are available in VINCI on RB03 and the SAS grid.

HERC’s Discharge Dataset
The MCA Discharge (DISCH) NDE includes information on the entire span of an inpatient hospitalization. HERC’s Discharge dataset includes the information in the MCA DISCH NDE with additional fields containing cost and length of stay subtotals for each inpatient category of care (e.g., acute medicine, psychiatry, nursing home, etc.). Learn more in HERC's MCA Discharge Dataset with Subtotals for Inpatient Categories of Care.

HERC Average Cost Data
HERC Average Cost Data are estimates of encounter-level costs for VA care. Data users can find overviews on the Inpatient and Outpatient webpages, and more detailed information in the Average Cost data guidebooks. FY24 HERC Average Cost Data and Discharge Data are now available on the SAS grid. Data will be available on RB03 by early June. 

Data Q&A

Facility-level average costs per day

Q: How do I determine the average cost per day of an acute med-surg inpatient stay for a specific facility and VISN?

First it is important to know the objective of this question. If the purpose is to understand costs at a specific VA facility, analysts can use the MCA TRT NDE to pull this information. You’ll need to identify the treating specialties (i.e., bedsections) associated with a med-surg inpatient stay. One option to do so would be to review the HERC inpatient categories of care, which lists treating specialties for acute medicine and surgery. Data users can then identify the station and VISN within with TRT file, and for all the records for their station or VISN of interest, identify the costs associated with the med-surg treating specialties. TRT contains all the costs for each patient per month for each Treating Specialty, and these can be averaged to calculate costs per day.

Data users should keep in mind that they will be pulling local costs which reflect local wages. This means that wages in some regions (e.g., San Francisco or New York City) will be more expensive than other regions. Analysts evaluating costs across sites can use the Medicare Wage Index for VA facilities to adjust their cost analyses for these wage differentials.

Alternately, if the purpose of this question is to compare VA facilities to non-VA facilities, then a word of caution is warranted. Both MCA and HERC data represent what VA spent in a year. That includes the costs of services that are rarely or never provided in non-VA facilities. Also, any unit cost comparison assumes that when you send a patient to a non-VA facility, the non-VA provider will provide the exact same care as provided in VA (same CPT or DRG code). That is not true. Non-VA providers rely more heavily on specialty care, and provide more care over time, even if that care is not needed. See Chan DC, Card D, Taylor L. Is there a VA advantage? Evidence from dually eligible veterans. American Economic Review. 2023 Nov 1;113(11):3003-43.

Do you have a question about using VA cost data? HERC offers consultations to VA data users on a variety of health economics and data topics. Visit the HERC website to learn more about our consulting service.

Publication

Bridging the Digital Divide for Older Adults

Research by a team of researchers from HERC and Ci2i explores the importance of social support in bridging the digital divide for older adults who have received VA tablets. The team differentiated between structural social support (e.g., marital status or living with others) and tangible social support (e.g., reliable help with daily chores, traveling to doctors’ appointments, or preparing meals). They found that greater tangible social support was associated with 54% higher odds of having a video visit after tablet receipt, while structural social support was not significantly associated with having a video visit. These findings highlight the importance of social resources, especially more practical support with daily needs, in facilitating telehealth use among older adults. Read a summary of their work in the February 12 VA Research News Brief. The full paper is available in Medical Care.