Information Overload
In 2002, Decentralized Hospital Computer Program (DHCP) was considered one of the best medical database management systems. However, the U.S. Department for Veterans Affairs (VA) knew it wasn't going to be able to handle increasing demand much longer.
The VA manages care, financial assistance and burial benefits for 70 million eligible participants and is facing debilitating database and patient care issues. Individual hospitals are responsible for upgrading DHCP, now called VistA, to meet their needs, but it is a patchwork system that hampers the transfer of data between hospitals.
“The root cause of the problem is the vast amount of potentially inconsistent data in the old system,” says Oliver F. Lehmann, PMP, vice president of professional development for PMI's Troubled Projects Specific Interest Group. “An aggravating factor is that the old database system is difficult to understand and, by design, susceptible to data corruption during transactions.”
In addition to information-sharing difficulties, the 27-year-old system is comprised of antiquated programming language and database elements.
Core Financial and Logistics System (CoreFLS), a $472 million project, was the first solution proposed to replace parts of VistA. The system was tested during a one-year pilot project starting in October 2003 at the Bay Pines Veterans Affairs Medical Center, St. Petersburg, Fla., USA. According to reports from the VA's Office of Inspector General (OIG), it was less than successful: Surgeries were delayed, radiology service was disabled, and risks to patients and financial problems increased.
When Bay Pines tried to return to the old system, management found massive problems in data quality and consistency. Reports by the OIG as well as a study prepared by Carnegie Mellon University found that reconciling data would be difficult. “These reports describe a situation of overall project management breakdown and insufficient responsibility taken to ensure correct procurement processes and a feasible implementation strategy. Documentation has been neglected and there has been no proactive risk management—which would have included sound reserves in people, equipment, time and money, plus mitigation strategies and fall-back scenarios,” Mr. Lehmann says.
The latest replacement system, HealtheVet, or HeV, is facing a similar fate. The $3.5 billion job bears unacceptably high risks and high potential for failure—failure which could lead to gridlock and compromised medical care for five million veterans, according to a new Carnegie Mellon University report and an internal VA paper obtained by the St. Petersburg Times. As of April 2005, the VA, Carnegie Mellon and the U.S. government can't agree on testing the system—the VA wants to proceed while the university and the government are not ready to take the next step.
“This effort will take a lot of time and require staff highly skilled in project management and software development,” Mr. Lehmann says. “The process of healing the VA needs to be done soon, because the amount of data is growing every day. U.S. taxpayers, hospital staffs and patients are paying a heavy price.”
If you know of a troubled project in which a project manager could have (or did) save the day, share your lessons learned. Contact [email protected].