What are the steps to conduct economic analyses for implementation research?
There are four phases for conducting economic analyses for implementation research: (i) Study Design, (ii) Data Collection, (iii) Data Analysis, and (iv) Data Reporting.
(i) Study Design
HERC developed these Guidelines for Economic Analysis to consider which form of economic analysis is most appropriate for the implementation/QUERI study.
The most common forms of economic analysis for implementation/QUERI studies include:
- Cost-identification; micro-costing is often used when the cost of the intervention is not known.
- Cost-effectiveness analysis (CEA)
- Budget impact analysis
(ii) Data Collection
- Develop an Economic Analysis Plan.
- Estimate the costs of an implementation intervention and downstream budget impacts following the best practices described in these slides.
- Create a finder file or crosswalk file to link participants in a study to VA administrative and utilization files.
- Sometimes the study requires the collection of additional information to estimate the cost of the intervention. In these situations, create a tool used for collecting data for micro-costing.
Examples of micro-costing tools from other studies:
- Non-VA Outpatient and Inpatient Care Form
- Client Contact Form
(Note: This form was used in an implementation trial. See https://www.cdc.gov/pcd/issues/2007/jan/06_0058.htm).- Virtual Specialty Care QUERI Implementation Costing Spreadsheet
(Note: Please direct questions regarding this form to John.Fortney@va.gov and Suparna.Rajan@va.gov).- PROVE QUERI QI Coaching Time Tracking Form
(Note: Please direct questions regarding this form to Nicholas.Yankey@va.gov.
(iii) Data Analysis
Analyze the data and estimate costs for patients.
This Statistical Program, designed for a past project, provides guidance on how to estimate annualized costs for patients. Users can modify this program, as needed, to identify the appropriate time interval and cost subtotals. In this program, the time interval was set at 1 year. For many projects, shorter time intervals, such as monthly or quarterly intervals, may be appropriate.
The statistical program may or may not include the cost of the intervention.
- There are some studies that re-organize existing care. In these situations, the cost of the intervention is included in the administrative data and it can be estimated through analysis.
- In other studies, the existing administrative data do not include the cost of the intervention. In these studies, the cost of the intervention will need to be estimated separately and added to the costs observed in the administrative data.
- Some caution is warranted in the steps above to not double count costs.(iii) Data Analysis
(iv) Data Reporting
- Report the cost of the intervention for the intervention and comparison group (mean and measure of variance).
- Report the cost of the subsequent downstream health care costs for the intervention and comparison group (mean and measure of variance).
- Report the net effect, which is the cost of intervention plus cost of subsequent downstream effects for the intervention and comparison group (mean and measure of variance).
- Organizations have developed checklists to improve reporting of health economic results.
