What Ails the NPFIT?
The last of the contractsthe deal for BT to build the N3 networkwas signed in February 2004. "The focus of the national program has now moved to the challenge of ensuring the timely implementation of high-quality IT services to help deliver a patient-centered NHS," Granger wrote in a 2004 article put out by the NHS. "Once in place, patients will benefit from a modern, IT-enabled NHS, every time they come into contact with it. The electronic revolution will help deliver coordinated convenient and integrated care, placing the patient at the heart of the NHS."
As Granger defined it, the NHS was building "a single electronic health-care record for every individual in England; a comprehensive, lifelong history of patients' health and care information, regardless of where and when and by whom they were treated."
Additionally, the NHS would provide health-care professionals with immediate access to summaries of care encounters and clinical events held in a national data repository, and support the NHS in collecting and analyzing information and monitoring health trends to make the best use of clinical and other resources, Granger said.
Before this utopian, cradle-to-grave vision of a centralized, monolithic national health-care system could become anywhere near a reality, however, a succession of daunting obstacles had to be overcome.
For one thing, there's the sheer size of the country's health-care system. Between 2002 and 2003, NHS served 52 million people; dealt with 325 million consultations in primary care, 13 million outpatient consultations and 4 million emergency admissions; and issued 617 million prescriptions.
Granger and the CfH had also inherited what Brennan terms "a mixed bag of incompatible computer systems, islands of technology that may work well in isolation but which cannot communicate with other systems. This wasn't a greenfield opportunity."
All of the systems had to be replaced with systems that could interact directly with the National Spine records system, but not before the data from the old computers was transferred from the old systems to the new, Spine-compliant systems.
Considering that the NHS alone, Brennan estimates, had 20,000 computers, that was a sizable task.
Then there was the little matter of managing the contracts and the LSPs effectively. Some critics of the program such as Brampton have charged that the CfH has dealt with perceived vendor deficiencies largely through Granger's threats to punish poor performance, and that it hasn't been flexible enough in dealing with unexpected problems confronted by the contractors.
A heavy hand on the whip, in other words, and little in the way of a carrot. At one point, Granger likened the NHS project to a sled and the LSPs to huskies. "When one of the dogs goes lame and begins to slow the others down, they are shot," he said, according to The London Times. "They are then chopped up and fed to the other dogs. The survivors work harder, not only because they've had a meal, but also because they have seen what will happen should they themselves go lame."
"He's not a diplomat," says John White, iSoft's director of corporate communications, of Granger, "but you need to be tough to manage something like this."
Spelling out the challenges confronting the NHS, Edwards wrote in an April 2005 Gartner report: "The larger the project, the greater the risk of managing vendors inadequately. Putting the prime contractors at risk for not delivering value is a sound idea. The challenge is one of balance. The stringent nature of the CfH contracts, which allow for payment only upon completion, reduces the scope for flexibility. It also increases the danger that when problems arise, the CfH and the prime contractors will become absorbed in arguments over contractual details, rather than concentrating on overall goals."
"The attitude was that the LSPs were responsible for solving all problems," Brampton says. "But you need to manage the contracts and the vendors, especially on something of this magnitude."
The CfH responds that this wasn't the case. "There is a mix of central, supplier and trust project management resources deployed in support of implementations," the agency stated to Baseline in an e-mail.
Another challenge spelled out by Edwards in April 2005 is the need for the CfH to remain focused on clinical adoption and change management. "Ensure that clinicians are adequately consulted and involved from the earliest stages of the IT program," Edwards wrote. "Allocate sufficient funds for change management and training. Identify and develop clinician champions in different geographic and functional areas of your organization. Work with them to determine what effects the program will have on the way clinicians practice medicine."
Edwards noted at the time that in a then-recent survey of clinicians in England, only 5% stated that the CfH had adequately consulted with them. "Evidence indicates that CfH has made insufficient progress in getting the wider community of clinicians involved and in motivating them to adopt the new applications," he noted.
This didn't happen at the program's outset. "Clinicians couldn't be brought in early on," Brennan explains, because of the secrecy surrounding vendor negotiations.
One consequence of excluding front-line health-care professionals from the early phases of the program, says the consultant, is that it fell largely to the IT vendors and the bureaucrats to create the system: "To a large extent, the result is a black box NHS is trying to sell to physicians who were not engaged from the outset." That black box is flawed on a number of counts, says the consultant. Among them, he claims, is that the program is too focused on administrative needs and not enough on clinicians' concerns.
Another challenge with the vendors was that in the early stages of the project, the LSPs, according to Brennan, focused on the easier applications and establishing links from the existing general practitioner systems to the National Spine. That should have been a relative cakewalk. A confidential report commissioned by NPfIT that was leaked to the U.K. newsletter E-Health Insider in April 2005, however, showed that the job of transferring 10 years or more of data from existing practice-based systems into the Spine-compliant systems that were being provided by the LSPs was far more complicated than had originally been anticipated, requiring clinical and computer expertise that often wasn't readily on tap. Typically, it took up to six months and cost around $9,000 per practice to enact the transfer, the report stated.
The overriding problem, however, was the software that was being developed. Both iSoft and IDX had to write some of the software for the CfH from scratch and to specifications established by the CfH. As an example, White, iSoft's corporate communications director, notes that one of the requirements for the clinical applications was that they had to have a communications interface that would enable them to transmit data to the National Spine.
iSoft was writing a new core application set called Lorenzo at its development center in India. Similarly, the IDX system, Carecast, was being written from the ground up in Seattle in conjunction with a team from Microsoft.
The difficulty was that the programmers, systems developers and architects involved didn't comprehend some of the terminology used by the British health system and, more important, how the system actually operated, the CfH conceded. The solution: In August 2005, the CfH announced it was looking for at least 100 clinicians to spend several weeks in India and Seattle working hand-in-hand with developers to anglicize the new software and "make sure their product is fit for purpose."