Health Economics Spotlight

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

September 2025

In This Issue

1. Claim Processing Time for Community Care Data
2. Cyberseminar: Understanding VA Primary Care Work Load: Developing a new method using EMR timestamps
3. The Economics of Telehealth
4. Comparing PAID and GovTA for VA Time and Attendance, Human Resources, and Payroll Processing Data
5. Average Cost of VA Outpatient and Inpatient Care
6. Wage Index Updated for 2025
7. Identifying ED Visits in VA Data

Spotlight

Claim Processing Time for Community Care Data

VA community care datasets report claims level information on non-VA care that was paid for by VA. Because community care datasets are derived from claims, there may be a delay in the time between the provision of care and the payment for care. A new HERC technical report presenting the claim processing time for claims in the VHA Office of Integrated Veteran Care Consolidated Data Sets (IVC CDS) to better understand the length of time data users should wait after service date before including CDS data in their analysis.

The tech report summarizes the number of days between the service end date and the paid date for IVC CDS claims in FY2022. The authors focused on IVC CDS as it is the primary data source for VA community care beginning FY2019. Among all claims, the average time between the end of service and the paid date was 57 days. Dental claims had the quickest average processing time (49 days), followed by professional claims (55 days), and outpatient institutional claims (56 days). Inpatient institutional claims had the longest average processing time (159 days).

Figure 1. Percentage of Claims Paid by 30-Day Increments

The team found that 90% of claims were paid within 112 days of service. For most analyses, including a 4-month lag period is sufficient. However, research focused on inpatient data should consider including additional lag time before including data in their analyses.

Additional details are available in Technical Report 46: Including Lag Time for Community Care Data Analyses in VA Research.

Seminar

Understanding VA Primary Care Work Load: Developing a new method using EMR timestamps

HERC Econometrics Seminar

Wednesday, October 22 at 1pm ET

Register

 
Todd Wagner, PhD
Director, Health Economics Resource Center, VA Palo Alto Health Care System 
Professor, Department of Surgery, Stanford University

VA’s workforce of 8,000 primary care clinicians face competing demands—growing secure messaging and care coordination across VA and community providers, pressure to keep more care in primary care before consulting specialists, and rising productivity / documentation requirements—often pushing work into after-hours “pajama time.” To expand access and increase productivity, VA has implemented bookable-time targets and work relative value unit (wRVU) tracking, but the effects on workload, productivity, and outcomes remain unclear. Because metrics like wRVUs can distort incentives, micro-level time-use data are needed to understand provider productivity and efficiency. Using EMR timestamps, our team developed a VA-specific time-tracking system and describe the methods and resulting dataset.

All seminars are open to VA participants only until further notice. 

Publications

The Economics of Telehealth

A recent paper led by health economist Kritee Gujral synthesizes the evidence of telehealth’s impact on health care use, costs, effectiveness, and health equity using an economics lens. Below, we briefly summarize the paper, and the authors provide recommendations for future research based on their findings. 

VA Leading the Way in a New Era for Telehealth

Telehealth has long held substantial promise for improving health care access, but adoption had historically been limited by restrictive regulations. These regulations (Table 1) were loosened during the COVID-19 pandemic to enable widespread uptake.

VA has historically been a leader in telehealth, and this held true throughout the pandemic (Figure 1). In April 2020, 45% of VA health care visits occurred via phone and video telehealth, compared to 29% of Medicare,13% of private insurance, and 4%-18% of Medicaid claims, depending on the state.

Figure 1. Percent of Services Delivered via Telehealth by Insurance Type, 2019-2021

Definitions of telehealth varied by insurer type. See Gujral 2025 for details. 

Policies Governing Telehealth Care

Patient information sharing

Licensing

Provider reimbursement parity

Patient cost-sharing

Insurance coverage

Coverage of audio-only telehealth

Coverage of care types

Coverage in home vs. medical settings

Coverage in rural vs. urban areas

For details of policy changes, see Table 1 in Gujral 2025.

Supply- and Demand-Side Factors

The authors reviewed demand- and supply-side factors affecting the adoption of telehealth. They examined whether telehealth is a substitute or complement to in-person care. They also examined the impact of telehealth on downstream use of costlier emergency and inpatient care. Overall, they found that telehealth adoption was associated with modest increases in total outpatient care, but evidence was sparsely distributed across care types (e.g., primary or specialty), telehealth modalities (e.g., audio, video, in-person), and pre- and post-pandemic periods. Rigorous studies are needed on comparative effectiveness of modalities, downstream costlier emergency and inpatient care, and efficiency gains in patient/provider time costs for a full economic picture.

Based on this review, there are several possible directions for future VA research.

• VA research can inform national policies (Table 1) on coverage of audio-only telehealth, different care types, home vs. clinic-based telehealth, rural vs. urban areas, underserved populations and the digital divide broadly.
• Randomized Controlled Trials (RCTs) on economic questions consist primarily of pre-pandemic data and non-video modalities. VA can conduct RCTs using newer data to answer economic questions.
• Conduct discrete choice experiments to understand patient/provider preferences.
• Causal inference/quasi-experimental methods with large-scale VA observational data can address key confounders (e.g., patient selection into telehealth, baseline outcomes/trends).
• Future research should examine comparative effectiveness of telehealth modalities.
• As there are virtually no studies on implementation costs, production costs, and provider non-monetary incentives, these are important areas for research using rich VA data. A unique opportunity is to consider how telehealth may reduce production costs by taking advantage of lower labor costs in certain areas of the country.
• Although community care was not explicitly covered in this paper, VA research should investigate whether access to telehealth leads to reductions in community care use.

Resources

Comparing PAID and GovTA for VA Time and Attendance, Human Resources, and Payroll Processing Data

VA staff and workload data can be a key source of information for research on VA employees. PAID and GovTA are two domains in CDW-Raw that provide detailed person-level staff information. While PAID was the original source of staffing data, VA began transitioning to a new timecard system (VATAS) in 2014, and these data were available in a separate domain: GovTA. This transition was not instantaneous, and the date when GovTA contains a complete capture of workload data varies by site. Currently, documentation about which domain to use is lacking. In this tech report, the authors examined the overlap between data in PAID and GovTA to provide guidance on which data source to use by year.

They found that data are available in GovTA for many stations beginning in 2013, and by 2018, complete data are available in GovTA for most stations. Despite the transition to VATAS (GovTA), some data continued to flow into the PAID domain, and during the overlap period, neither system contained complete data capture. Therefore, for operations projects, the authors recommend using both PAID and GovTA during the overlap pay period and then relying on GovTA once it contains a complete capture of staffing and workload information. 

More information is available in Technical Report 47: Comparing PAID and GovTA for VA Time and Attendance, Human Resources, and Payroll Processing Data. This technical report is available on the VA intranet only.

Average Cost of VA Outpatient and Inpatient Care

For those looking to quickly estimate the cost of VA care, the HERC website includes the average cost of VA inpatient and outpatient care by year. All costs are calculated using HERC average cost data, and data are not adjusted for inflation. On the HERC Outpatient Average Cost Data page, visitors can find the average cost for 12 categories of outpatient care by year for FY2015-2024.

The HERC Inpatient Average Cost Data page includes the average daily cost of VA inpatient care for 11 categories of care. The page includes both local average cost estimates and national average cost estimates. Both pages also list the codes included in each category of care.

Wage Index Updated for 2025

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 2025, is saved as an excel file in the guidebook Medicare Wage Index for VA Facilities.

Data Q&A

Identifying ED Visits in VA Data

Q: How do I identify emergency department visits in VA data?

Emergency department (ED) visits are generally documented in outpatient data. Below are codes that HERC analysts commonly use to identify ED visits, including visits to urgent care (UC).

• VA-provided ED: stop code 130 (UC: 131)
• VA-paid community care
  • Fee (<FY2019): place of service code 23 (UC: 20); Fee Purpose Of Visit codes 22 and 23
  • IVC CDS (FY2019+): place of service code 23 (UC: 20); revenue code like 045* or 0981
• Cerner Millennium: combination of EncounterType and the CodeValue associated with the LocationNurseUnit
• Medicare claims: place of service code 23 (UC: 20); revenue code like 045* or 0981
• All data sources: procedure codes 99281-99285 and G0380-G0384 (UC: S9083)

Please review all codes before applying them in an analysis. Codes may vary by study.

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.