Unified, risk-based model of assessing healthcare performance expected to be more clinically relevant to physicians than current measures.
7 Big Data Solutions Try To Reshape Healthcare
(click image for larger view and for slideshow)
The group that developed many of the quality measures in the Meaningful Use standards for electronic health records (EHRs) is testing a new approach that pulls data from EHRs to help assess quality in a way that the organization believes to be more clinically relevant than current methods.
The National Committee for Quality Assurance (NCQA) has landed an 18-month grant from the Robert Wood Johnson Foundation (RWJF) to evaluate a single measure for assessing risk of cardiovascular disease called the "Global Cardiovascular Risk" score, or GCVR. Data collection will start this summer and NCQA expects to report its findings during summer 2014.
GCVR is based on a measure created by Archimedes, a Kaiser Permanente affiliate that applies mathematical principles to solve pressing healthcare issues. San Francisco-based Archimedes developed a metric, called the Global Outcomes (GO) Score, including the logic behind it and a tool for calculating it. The metric was described at length in a paper in the November 2012 issue of policy journal Health Affairs.
"This will be the first large-scale test" of the GO Score, Archimedes founder and retired chief medical officer Dr. David Eddy told InformationWeek Healthcare. Eddy said his goal in the NCQA test is to have the GO Score become a part of widely accepted clinical performance measures, which he hopes ultimately will lead to more accurate treatment choices. Archimedes will be providing technical support to NCQA.
"It is based on outcomes. It is not based on processes and targets," Eddy explained.
For this reason, Dr. Mary Barton, NCQA's VP for performance measurement, is optimistic that the GCVR can change the way healthcare professionals assess patient risk and allocate resources. "The evolution [of quality measures] over time has depended on the accumulation of evidence-based data," Barton said.
However, risk of any given disease is based on a complex set of variables. "These are all interventions you should be giving your patients, but some will give you more bang for your buck," Barton continued.
Existing measures of quality are "extremely binary," as well as "clumsy and one-size-fits-all," Barton said. Providers either meet or fall short of their goals for each patient's LDL ("bad") cholesterol goals. But if one patient is just five points above the LDL target and another exceeds the standard by 40 points, physicians might be tempted to put their efforts into reducing the cholesterol level of the patient just five above the threshold, even though the one with the higher number is a greater risk of heart disease, Barton suggested.
"Now I'm going to be measured on the benefit I bring to my patients" with the GCVR score, said Barton, who trained in internal medicine.
She described the GCVR score as a "vision" and a "dream" for performance measurement. "It's a big change in the primary care paradigm," Barton said. She called the RWJF grant a first step toward realizing that vision.
"The dream is, we can help more people and we can do much more good" by giving physicians the motivation to assess entire populations of patients and by helping organizations better allocate their resources, according to Barton. "It lets you look at the big picture."
Eddy likened existing standards to a compass without a map. "Our current performance measures are directing people north but not telling them how to get to the North Pole," he said.
The GO Score becomes the portion or percentage of the target providers hit with each patient. "The score is very meaningful," Eddy said. He believes it has three advantages: accuracy, simplicity and a basis in reality.
To Eddy, the score says, for example, that an organization is preventing half of the heart attacks that can be prevented, not simply that the doctors are prescribing beta blockers for, say, 63% of those at high risk of heart attack. It is a single score, not a series of yes/no questions. Plus, it does not have any "sharp cut points," as Eddy put it, arbitrary cutoffs that encourage physicians to treat everyone in the same way.
"Physicians are frustrated with current clinical measures because they are not clinically relevant," Eddy said. "This method leaves them free to choose how they treat their patients."
The proliferation of EHRs makes the test feasible now, though NCQA will be working only with organizations with enough EHR history to report clinical data electronically. This test will not have to rely on data from billing claims. Barton said she was unable to share names of testing partners because NCQA has not signed formal agreements yet.
Still, Barton and Eddy both envision measures like GCVR becoming part of Meaningful Use Stage 3.
The current Stage 1 just asks for "electronified versions of otherwise existing measures," according to Barton. Stages 2 and 3 will take fuller advantage of the "rich data" from EHRs, she said.
She hopes that the third stage, which will not begin before 2016, will count such things as how many patients are able to lower their blood pressure by 10% or 20% or some other metric likely to have a bearing on medical outcomes. "I'm very confident that it is possible with at least some of the current EHR systems," Barton said.
Regulatory requirements dominate, our research shows. The challenge is to innovate with technology, not just dot the i's and cross the t's. Also in the new, all-digital The Right Health IT Priorities? issue of InformationWeek Healthcare: Real change takes much more than technology. (Free registration required.)