Today, I testified before the U.S. House of Representatives, House Committee on Veterans Affairs, Subcommittee on Technology Modernization. Click here to download a pdf of the written testimony.
Chairman Rosendale, Ranking Member Cherfilus-McCormick, and members of the Subcommittee:
Thank you for the opportunity to testify this morning. My name is Reynold Schweickhardt. I am a non-resident senior fellow with the Foundation for American Innovation, a nonprofit organization with a mission to develop technology, talent, and ideas that support a better, freer, and more abundant future.
My testimony is based on my experience working as the Senior Advisor for Technology at the General Services Administration, and as the Chief Information Officer, Chief Technology Office, and acting Chief of Staff at the Government Publishing Office, as well as my tenure as a senior staffer with the U.S. House of Representatives’ Office of the Chief Administrative Officer and with the Committee on House Administration.
The mission of the Department of Veterans Affairs—“to fulfill President Lincoln’s promise to care for those who have served in our nation’s military and for their families, caregivers, and survivors”—is one of the most important and solemn responsibilities of the federal government.
In the twenty-first century, fulfilling the VA’s mission is inextricably linked to the department’s ability to manage and modernize major information technology systems, including those required to deliver health and other benefits to the nation’s roughly 16 million veterans. Unfortunately, for too long, the VA has struggled with information technology management, modernization, and cybersecurity. The Government Accountability Office (GAO) identified “managing risks and improving VA health care” (including IT challenges) as a high-risk area in 2015. In 2019, GAO added “VA acquisitions management” to its high-risk list.
The costs of the VA Department’s longstanding challenges in its management of health care benefits, information technology, and major acquisitions programs are unacceptable. Too many of our nation’s patriotic veterans have faced challenges accessing owed benefits and timely health care service. For American taxpayers, VA’s IT acquisitions management programs have become prominent examples of the federal government’s track record of IT modernization projects running billions of dollars over budget and years behind schedule.
Smaller and more agile strategies can help the VA modernize with a consistent series of short-term improvements. The primary challenges to implementing this strategy at a scale to make a significant difference are not technical. They are cultural and managerial.
As the 118th Congress comes to a close, and as we prepare for the 119th and the start of a new administration, I urge Congress as well as the current and future leaders of the Department of Veterans Affairs to consider the following recommendations to improve VA information technology modernization and management:
First,The next VA Secretary should establish several agency-wide modernization goals around standardization of processes and effective IT implementations that can be used consistently. These priorities should become a shared component of every executive and senior executive service member’s performance plan to focus the entire department on a unified vision.
Second, the next administration should review all ongoing VA IT acquisitions management and modernization programs to identify which are working, which are failing, and which need to be prioritized. The next Secretary and their staff should ask the following questions: Where are there areas of high-performing IT delivering timely progress, and where are there areas which are underperforming? What are the differences, and how do we work to spread high-performing approaches to other areas? The answers to these questions can be used to update the VA Department’s IT Strategic plan.
Third, Create smaller and more manageable projects in the future and move towards more agile methods.
While VA has begun to use agile methods, it must confront the challenge of too many number-one priorities. The VA has many stakeholders with important missions. These include congressional committees, OMB and the President’s Management Agenda, agency leadership, Veterans Service Organizations, VA staff, GAO, and the VA OIG. The tendency is to show each stakeholder that IT values their requirements and to show progress on as many initiatives as possible. The problem, of course, is that the effort is so diluted that many initiatives make little progress.
Even if the CIO wanted to focus on more rapidly completing a lesser number of priorities, agency leadership would need to resolve the priorities of the various stakeholders.
New initiatives such as supply chain modernization, which will consolidate 63 existing systems, need to have an incremental strategy that can deliver value over shorter periods of time.
Starting more manageable projects is not the same as agile development.
The management philosophy of agile project management is to focus resources on completing and delivering benefits to stakeholders in a rapid sequence of sprints of one to four weeks. Projects are selected because of improvement to customer experience, priority of the functionality, reduction of complexity, shrinking technical debt, and building momentum to increase the capacity of the agile approach. Cultural challenges will need to be addressed before agile development plays a major role in VA IT modernization.
Fourth, the next administration should automate one VA. The VA leadership should identify the full scope of projects such as EHRM by defining the future state for both business processes and standards of care as a gold standard. Each center should document their unique requirements and where they diverge from the standard so that the scope is fully understood.
The goal of process convergence is to enable “build once, use many,” not “build once and modify many, support many.” One hundred thirty separate instances of VistA demonstrates an unmanageable divergence. As EHRM is installed at additional centers, it should be one instance for one VA. Standards and a common future state would provide the following benefits:
- Disagreements about the future state can be resolved prior to future implementations.
- Any changes in legacy systems that are noncompliant can be made conforming.
- Partners, such as the remaining 10 percent of community hospitals that do not interface with EHRM, have a longer lead time to implement changes. It should not be VA’s responsibility to integrate with individual institutions that are not using a standard interface.
Most importantly, each center can be assessed for implementation compliance. There is a truism in IT project management that 90 percent complete means 50 percent cost because the hardest work is often put off until the end. Having visibility into all of the required changes at each center will improve the accuracy of estimates and timelines.
Fifth, the next administration should increase the focus on standards, shared services, and commercial alternatives to reduce the complexity and scope of VA procurements. Standard definitions, scope, and interfaces allow for more modular designs that support an agile development philosophy.
The VA should work with the Department of Health and Human Services and other federal agencies and the private sector to continue to standardize healthcare interoperability. A significant benefit would be to simplify the integration of modular capabilities into existing systems (see below).
The selection of a shared service provider that meets federal requirements for financial and acquisition business requirements for the FAIMS implementation is commendable. If the VA is investing in heavily customizing the system, it is not really using a shared service, because it is turning it into different software which, after customization, requires its own maintenance and lifecycle costs. Good customization involves using the vendor-provided options and tables to meet VA requirements. Bad customization is changing the base software.
Any changes which require a separate software instance for shared-service or commercial software packages should be carefully examined, and the presumption should be to change business processes instead. If unique software is required, it should be designed as a plug-in invoked by the software.
Could the VA acquire and use commercial healthcare components? While the VA supports over nine million veterans, there are examples of integrated healthcare systems of similar size using commercial software. The VA should evaluate opportunities for a shorter project to license existing capabilities. Kaiser Permanente, with over 12 million members, is using capabilities like these at scale:
- “A program called Advance Alert Monitor uses AI and helps prevent emergencies in the hospital before they happen. Every hour, the program automatically analyzes hospital patients’ electronic health data. If the program identifies a patient at risk of serious decline, it sends an alert to a specialized virtual quality nursing team. The nursing team reviews the data to determine what level of on-site intervention is needed.”
- “A Desktop Medicine Program: This program uses natural language processing (NLP) algorithms to categorize patient messages and direct them to appropriate respondents, reducing the workload for physicians.”
Note that these examples are not intended to suggest these should be VA’s priorities going forward.
Sixth, Extend the tenure of the agency CIO. The Subcommittee has expressed concern about the average tenure of CIOs, including that the average tenure is more than 10 months and has extended to two years.[9] The current CIO was confirmed by the Senate in December 2021, and will serve just over three years in that position. Three years is as good as it gets for the service of a qualified private sector CIO in one term. Time is spent in the vetting and confirmation process, including addressing conflicts of interest. Consistent long-term leadership would reduce churn and create an environment to focus on completing long-running and often delayed projects.
There are advantages to either a political or career CIO. The next administration should consider which approach is a better fit for VA. Regardless of that decision, creating an environment for someone to make a longer commitment is necessary. One option could be a term appointment for five years, subject to renewal.
- In conclusion, the next administration can improve the care and support that America’s veterans receive by taking a step back and addressing the root causes of the existing delays and cost overruns in modernization projects. These key steps are:
- Creating a unified VA through agency-wide secretary priorities incorporated in every executive’s performance plan.
- Implement a mindset of build once, use many.
- Define a single set of business and patient care standards.
- Improve agency-wide IT planning by fully evaluating the complexity of agency-wide implementation during requirements definition.
- Reduce EHRM cost overruns and schedule slippage by assessing each unimplemented center against the business and patient care standards to uncover and document the actual scope of work to successfully implement EHRM.
- Break up requirements into smaller projects with shorter implementation. Adopt a philosophy of “minimal viable product” to do the least work that makes a difference for stakeholders. Implement it and repeat until that area is no longer a priority for further modernization.
Increase the tenure of the CIO to provide consistent leadership to implement these changes in a lasting manner.
Thank you for the opportunity to testify. I look forward to your questions.
