No better proof of that is what Colene Byrne and Douglas Johnston said in an interview Tuesday. Byrne is a senior analyst at the Center for Information Technology Leadership. Johnston is executive director of the CITL, which is part of Boston-based Partners HealthCare System.
The CITL research team recently published an article about their findings in the policy journal Health Affairs. Byrne and Johnston said their team of researchers worked for an entire year piecing together what was at best only an estimate of the cost of just four parts of the VA's complex VistA system, which has more than 100 software modules for inpatient and outpatient care.
“We also originally planned to look at others, such as telehealth, but we narrowed it down to these four,” Byrne said.
“I think getting the data to do those four was challenge enough,” Johnston said.
The VA, it seems, didn't keep good track of how much it had spent on VistA, and didn't know whether it was getting a return on its IT investment, numbers its leaders now would like to know and be able to present to Congress. So the VA hired the CITL, which had some prior experience modeling IT expenditures and benefits, to come up with what Byrne is careful to say is not a return-on-investment estimate but rather an estimate of the financial value of the VA's IT investment.
Johnston said they selected the four VistA systems—bar code medication administration, picture archiving and communication, labs and the Computerized Patient Record System, or CPRS—because they had long been in use at the VA and were most likely to have had a significant effect on patient safety. The CITL determined the VA has spent $4.07 billion on the four VistA systems in the study while estimating the “gross value” of that investment at $7.16 billion.
There has been significant controversy since the CMS' December 2009 release of its proposed rule on the so-called “meaningful use” criteria for the EHR subsidy program under the American Recovery and Reinvestment Act, or stimulus law.
More than 1,000 public comments were submitted about the proposed rule and many called for the CMS to dial back the criteria from those in the proposed rule, which, for example, called for all hospitals to have CPOE installed and to attest that 10% of all physician orders come via CPOE. Only about 30% of hospitals have some form of CPOE up and running in at least one department, according to Don May, vice president for policy at the American Hospital Association.
At the VA, however, CITL researchers report close to 100% usage of key VistA components, including CPOE.
In contrast, the private-sector adoption rate in 2007, the end of the study period, was 16% for CPOE, 22% for bar code medication administration and 12% for outpatient EHR.
In addition, the patient-safety implications of health IT have once again been called into question with members of the IT community expressing concern about possible Food and Drug Administration regulation of IT systems as the FDA does now with medical devices.
The CITL researchers also looked at IT-related quality process and outcomes measures for 2004 through 2007 to see how the IT-intensive VA is doing compared with the private sector.
The researchers focused on diabetes care, because 25% of VA patients have diabetes.
According to their report, the VA did better on diabetic glucose compliance and control, cholesterol control and timely retinal exams compared with Medicare HMO patients, the private-sector benchmark.
“The VA averaged about 15 percentage points higher than the private sector on preventive care for patients with diabetes and 17 percentage points higher for patients with diabetes who have well-controlled cholesterol,” according to the CITL report. Those performance spreads were present consistently for each measure for each of the four years in the 2004-07 comparison period, a clean sweep for the VA.
The VA's ability to maintain that high quality of care, year after year, impressed Byrne.
“With the sustained use of the alerts, they've been able to sustain those measures, particularly with chronic conditions such as diabetes,” she said. “There is very high compliance with those alerts.”
That's not to say it is merely the VistA “machine” that produces the high level of quality care, according to Byrne. The alerts correspond to care guidelines, adherence to which is part of a care incentive program under which VA clinicians operate.
“The tool and the reminders can only facilitate the guidelines and the incentive system,” Byrne said. “In another study, we've had an opportunity to speak with a number of VA providers. They're so used to the CPRS, when they go to another system, it's just so frustrating.” Some practitioners said “they would never provide care again at another setting” that doesn't have an IT system like the VA's, Byrne said.
Byrne concedes the VA is a unique healthcare system and warned against making direct comparisons between the VA's results with VistA and what a private-sector organization might accomplish with its own IT system. The VA owns hospitals, employs its own clinical staff, has inpatient and outpatient records linked on the same IT system and provides covered veterans with insurance for life, so there is no concern that money spent on preventive medicine will be wasted if the veteran moves to another health insurance plan.
“They have scale that's hard to replicate in the private sector,” Byrne said. But the VA also had extremely strong leadership support to spend on healthcare IT and to push for provider adoption and buy in, a replicable factor anywhere.
As an example of leadership and organizational commitment to IT, Byrne said when the VA implemented bar code medication administration, it did it in four waves of six-month periods of training and rollout over two years.
In addition, demographically, “they also have an older population, so they have a higher potential for medical errors,” but also a higher potential for error reduction as well.”
The VA's structure makes it even more difficult to link the VA's IT investments as the “cause” of its high-quality care and producer of the $3.09 billion in “total net value” of the VA's investment in VistA components used in the study, according to Byrne.
“We've been very careful to say in the report, it's likely, it's very likely, that the VA achieved positive return on their investment,” she said. But those numbers are estimates of what it was possible to achieve, she said, “and not ROI.”