PORTFOLIO

HCD Research CMS Medicare Part B Drug Average Sales Price (ASP)

Overview

Redesign of the Average Sales Price (ASP) application for CMS. The ASP application is used by drug manufacturers to submit federally required Medicare Part B Drug ASP data to the Centers for Medicare & Medicaid Services (CMS). The ASP system is part of the Fee-for-Service Data Collection System (FFSDCS) that CMS uses to house various Fee-for-Schedule modules.

Records for all drugs are maintained within the ASP system which also establishes a relationship between the manufacturers’ reported data and the billing codes Medicare providers use to calculate a weighted average sales price for each code. Billing codes are then used to determine payment limits of Part B drugs on certain Medicare claims.

My Role

Human-Centered Design Research

Research Methodologies Used

Discovery interviews with
- Internal CMS users
- CMS product owner
- Help desk technicians
Usability testing with drug manufacturers
Analyzing and synthesizing data obtained from all research to determine key findings using Dovetail
Sharing findings with the team via artifacts created in Figjam and Miro

Discovery Interviews with Internal CMS Users

I planned and facilitated discovery interviews with health insurance specialists and pharmacists on the CMS team who use the ASP system to learn why and how they currently use the application, pain points with the current process, and how it could be improved. I used the following moderation techniques during the interviews:

Concurrent Think Aloud (CTA)
- CTA encourages participants to keep a running stream of consciousness by thinking aloud during the session. CTA can help moderators understand participants’ thoughts as they interact with the prototype and complete. Observing an individual’s behaviors while thinking aloud gives us the closest representation of what their experience would be in real life.
Concurrent Probing (CP)
- CP requires that the moderator ask follow-up questions of participants as they work on tasks. When they say something interesting, or do something unique, or seem confused, etc., the moderators will ask follow-up questions to probe for additional information.

Discovery Findings

Interviews were recorded via Zoom and imported into Dovetail where I performed detailed analysis using Dovetail's affinity mapping feature. Findings were categorized by topics such as process, pain points, and recommendations for improvements. I used Figjam to create various findings artifacts which I presented to the CMS team and project manager.

Current Process & Tasks

The following Figjam board captures the current process and tasks which involves a lot of manual tasks. For example the Health Insurance Analyst manually checks if any reported new drugs are already in a spreadsheet she maintains (this spreadsheet includes over 40 columns and thousands of rows). If drugs are not in the spreadsheet, she manually compiles a list and sends it to pharmacists.

The pharmacists on the ASP team must then verify all new drug data manually using Google, searching drug manufacturer websites, or calling manufacturers directly because there isn't one compendia of all drug data. This process can take a several days.

Figjam diagram of current (as is) process and tasks

Pain Points

Various pain points about the current process were noted with a key theme being that its very manual.

The Health Insurance Specialist maintains data manually using spreadsheets that contain thousands of data rows and over 40 columns.
Pharmacists must verify all new drug data manually using Google or by searching drug manufacturer websites because there isn't one compendia of all drug data.
Pharmacists must manually calculate the ASP data.
Pharmacists must manually look for data that “seems off” or incorrect. If an incorrect item is missed, “everything from that point is false.”
The current system contains multiple names for some drug manufacturers which occurs when submitters use a different spelling or abbreviation.

Figjam diagram of pain points noted by CMS Users

Wish List of Improvements

During discovery interviews I asked the CMS users for ideas on how the current process could be improved. What would make the process easier, faster and more automated? Their ideas included:

Have the system flag new drugs not previously submitted
Create new records for drugs within the application
Send and track notifications to manufacturers via the application
Complete ASP calculations within the application
Build in functionality to upload drug data from industry compendia

Figjam diagram of wish list of improvements

Discovery with Help Desk Technicians

Additional discovery was completed by interviewing Help Desk Technicians who assist drug manufacturers who use the system and submit help desk tickets if they have issues they don't know how to resolve or errors.

The majority of tickets are submitted by new users
New users commonly don’t understand the process to follow to submit their drug data.
Many don’t bother to refer to the user guide to find answers.
Some users do refer to the user guide but are unable to find answers.
A common error is in the formatting of data in the Excel file they upload to the system.

These findings suggest that improvements can be made to make the overall process easier and more user friendly. Also, an audit of the user guide plus incorporating help at the point of need within the system would improve the user experience. Based on this finding, tool tips were included in the high fidelity prototype so it could be tested with users for feedback.

Usability Testing with Drug Manufacturers

Following discovery, findings and recommendations for improving the user experience in the new system was shared with the project manager and UX designer. Upon the completion of high fidelity prototypes, I planned and conducted two rounds of usability testing with drug manufacturers. A usability research plan was outlined the goals, process and tasks that participants would complete using the prototype. View Usability Study Research Plan.

Overall, usability testing was successful. Although minor issues were noted, the drug manufacturers who participated gave the new system favorable reviews. A full report was created in Dovetail and included links to video/audio quotes. The CMS team said hearing manufacturer feedback in their own words very helpful

The Business Process flow on the landing page that provides a high level overview of the process is helpful. They understood that the links in the flow would take them to the page to complete the step. Although one participant, an experienced user, commented that perhaps other content would be better since the links in the flow take users to the same locations as the navigation links.
Two of five participants said they would find the tool tips helpful, especially for new users.
Two of five participants suggested adding alerts or bulletins similar to the another CMS website.
Three of five participants were confused by adding National Drug Codes (NDCs) in some way, and wanted to go to Product Data to complete this task. This is new functionality in the redesigned system.
Another new feature is manufacturers will be able to add new NDCs in the system but must be approved by CMS. Participants were ok with the approval process but the system does not provide feedback on how long approvals will take.

Recommendations or considerations shared with the team:

Revisit the Business Process Flow on the landing page and its importance.
Consider adding more details on adding product data vs adding NDCs as this was not clear to even experienced users.
Add text to inform users of the approximate time needed for approval of new NDCs

View full report of findings created in Dovetail: Usability Testing Insights

Service Design

My last task as HCD Researcher for the ASP project was to map features and functionality by user type. This is not a usual activity for an HCD specialist but I knew this information was not documented so I decided to create simple service design diagrams using Miro. This information will be helpful reference going forward to anyone not familiar with the ASP system.