Data FAQs - Determine Cost of X

First, determine how you will define the condition of interest (e.g., diagnosis codes), and create a comprehensive list of codes. When setting the cohort parameters, consider whether patients can enter and/or leave the cohort (e.g., if they have a condition that is curable) or if there is a time frame for the diagnosis.

Next, choose which cost data is the best fit for the study (Managerial Cost Accounting (MCA) data or Health Economics Resource Center (HERC) Average Cost data). For more information on the two data sources, review the question Which VA cost data source should I use: HERC or MCA?.

After choosing your cost data source, determine how you want to calculate annual cost, e.g., calendar year, fiscal year, a year from randomization date. Determine how you want to treat stays that continue beyond the study year. Define the boundaries of the data you want to include. Do you want to include inpatient, outpatient and pharmacy data? Do you want to include VA data only; VA data plus data for care paid for by VA but received in the community (i.e. community care data); or VA data, community care data, and non-VA data (e.g. Medicare)?

After DART approval, researchers will work with a VINCI data manager to pull data for the defined cohort. Merge the cohort file with the cost data. For more information on how to merge these data, see VINCI documentation on linking patient data in CDW, available on VINCI Central https://vincicentral.vinci.med.va.gov/SitePages/Home.aspx (VA intranet only). Consider whether you want to include all costs or only costs related to the diagnosis. If you want to limit to costs for a specific condition, consider partnering with a clinical expert and/or limiting to a condition with clearly defined associated costs.

First, determine how you will define the condition of interest (e.g., diagnosis codes), and create a comprehensive list of codes.

Next, determine whether you want to use Managerial Cost Accounting (MCA) data or Health Economics Resource Center (HERC) Average Cost data. For more information on the two data sources, review the question Which VA cost data source should I use: HERC or MCA?

After DART approval, researchers will work with a VINCI data manager to pull data for the defined cohort. Merge the cohort file with the cost data and pull costs. For more information on how to merge these data, see VINCI documentation on linking patient data in CDW, available on the VINCI intranet site https://vaww.vinci.med.va.gov/ (VA intranet only). Link your cohort file to any other data sources your team has decided to include (e.g., Community Care or Medicare).

As you design your analysis plan, consider which data you want to include. Do you want to include inpatient, outpatient, and pharmacy costs? Do you want to include VA data only; VA data plus data for care paid for by VA but received in the community (i.e., community care data); or VA data, community care data, and non-VA data (e.g., Medicare)? Recognize the unknown costs that are omitted from your analyses, e.g. Medicare Advantage, Medicaid, and private insurance. Consider whether cleaned datasets already exist for your cohort, e.g., Health and Retirement Survey (HRS) data or Medical Expenditure Panel Survey (MEPS) data.

for additional advice including sample SAS files, contact HERC at herc@va.gov.

For VA-approved research and operations projects, the Pharmacy Managerial Cost Accounting National Data Extract contains information on the cost of a medication itself (variable: act_cost) as well as the dispensing fee (variable: dispcost). Sum the actual total cost of the medication and the dispensing fee to calculate the total drug cost. For more information, see the guidebook “VIReC Research User Guide: Pharmacy Managerial Cost Accounting National Data Extract (PHA MCA NDE)” Available on the internal VA intranet only (https://vaww.virec.research.va.gov/RUGs/MCA-NDEs/RUG-MCA-PHA-NDE.pdf).

For those looking for cost information outside of a VA-approved project, or for cost information more generally, the VA National Acquisition Center provides a searchable Pharmaceutical Catalogue which includes the Federal Supply Schedule (FSS) price and Big 4 agency price (VA, Department of Defense, Public Health Service, and the U.S. Coast Guard) of specific drugs.

The actual amount VA Pharmacy Benefits Management (PBM) spends on a medical product may be lower than the FSS price or Big 4 price after rebates and discounts. The components of the Actual Total Cost (e.g., contracted price paid for the drug product) is confidential and may not be disclosed. If researchers need to report the pharmaceutical unit price, HERC recommends using the FSS price listed on the VA National Acquisition Center Pharmaceutical Catalog.

Additional information on the cost of pharmaceuticals for cost-effectiveness analysis is available in the HERC presentation, “Pharmaceutical Costs for Cost-Effectiveness Analysis,” and the webpage “Determining the Cost of Pharmaceuticals for a Cost-Effectiveness Analysis.”

To determine the cost of procedure X at VA, first identify the procedure codes of interest. CPT codes are used for outpatient data, while ICD-10 procedure codes are used for inpatient care. Note that some inpatient procedures are captured with CPT, but that is the discretion of the physician. VA employs professional coders to enter ICD-10 codes for inpatient care, so best to use these.

You can either use the MCA National Data Extracts (NDEs) or HERC Average Cost data to identify health care costs. Because neither data source contains sufficient clinical information such as procedures, you will first need to use the procedure codes to extract encounters of interest from the encounter files (i.e., MedSAS, CDW, OMOP). For example, if you were interested in cataract, you could extract all encounters where one of the CPT codes is 66984 (basic cataract) and 66982 (complex cataract). This would yield all cataracts for a given period of time and/or location. You would then merge these encounters to the HERC or MCA data. See HERC's Average Cost guidebooks or the MCA NDE Guidebook for details on merging datasets.

Outpatient procedures using MCA data: MCA OUT NDE contains outpatient costs per clinic stop per day. Once you’ve merged your encounter and cost datasets, you can use the Total Cost variable (ActTotCost or ocst_tot) in the MCA Outpatient (OUT) NDE to determine the cost of the encounter associated with your procedure of interest for all outpatient encounters.

Outpatient procedures using HERC Average Cost data: HERC outpatient average cost data contain a VA cost estimate that is based on hypothetical Medicare reimbursement per CPT code. HERC costs are reconciled so that the estimates tally to actual national VA expenditures for that type of care. HERC Outpatient average cost estimates include the HERC Value (PAYMHERC), the National VA average cost (COSTN), and the local VA average cost (COSTL) for each encounter.

Inpatient procedures using MCA Data: Both the Discharge (DISCH) and Treating Specialty (TRT) NDEs contain information about the cost of inpatient stays. One advantage to using the TRT NDE is that a new record is created for each change in treating specialty during an encounter. This can be used to limit costs to those related to the procedure of interest and remove unrelated costs (e.g. the cost of long-term care). Use the Placeholder Total Cost variable (TotCost or tcst_tot) to determine costs for the inpatient stay.

Inpatient procedures using HERC Average Cost Data: HERC Inpatient Average Cost data contain the national average cost of a hospital stay given its Diagnosis Related Group (DRG), overall length of stay, and days in intensive care. Files include local cost (costl) and national cost (costn) estimates. The average cost method assigns the same cost to all inpatient stays with the same demographic and discharge information. Therefore, stays with identical characteristics will have the same cost. if you are interested in comparing the cost of two procedures, these data may not be appropriate.

The costs described here are the costs associated with the encounter for the procedure of interest. Depending on your study, you may want to consider whether to include additional costs (e.g. follow up visits, travel time, medication costs) or break down costs to the department level, using the Intermediate Product Department (IPD) Inpatient and Outpatient NDEs.

For more information on MCA NDEs, visit HERC’s MCA NDEs webpage. Information on HERC Inpatient and Outpatient Average Cost data is available in the Average Cost guidebooks.

First, determine which VA cost data source better fits your study’s needs: Managerial Cost Accounting (MCA) data or Health Economics Resource Center (HERC) Average Cost data. Both data sources include clinic stop. For more information on the two data sources, review the question Which VA cost data source should I use: HERC or MCA?

Are you interested in costs for a specific cohort? If so, merge the cohort file to the cost data (HERC or MCA). For more information on how to merge these data, see VINCI documentation on linking patient data in CDW, available on the VINCI intranet site https://vaww.vinci.med.va.gov/ (VA intranet only).

Consider whether you are looking at the cost for a given time frame or for select VA Medical Centers. If so, limit the data accordingly. Consider also whether to include all costs (fixed and/or variable). You can then limit to the specific clinic stop to calculate the average cost of treating a patient.