One of the biggest barriers to meeting the Stage 1 meaningful use criteria for government electronic health record incentives has been the collection of data for clinical quality measures (CQMs). And, to judge by the statements of organizations representing health care providers and EHR vendors, electronic quality reporting and the software rewriting it requires may be even more difficult in Stage 2.
In its comments on the proposed rule for meaningful use Stage 2, the American Hospital Association told CMS that hospitals had encountered "significant difficulty" in using EHRs to do quality reporting in Stage 1. Citing "inaccurate e-specifications" for the electronic measures and "unworkable, but certified, vendor products," AHA asked CMS not to add any additional measures in Stage 2, but to help providers and vendors "get it right" on quality reporting.
The College of Health Information Management Executives, which represents chief information officers, made similar comments, saying that the CQM data has so far been "inaccurate and largely incomparable across different providers."
The EHR Association, representing the major vendors, also opposes an increase in the number of CQMs. In its comments, the association expressed concern about the lack of e-specifications in the proposed Stage 2 rule, the absence of provisions for field-testing and the amount of time allotted for vendors to rewrite their software so that it can be used to generate CQM data.
CMS is reserving comment until it releases its final rule, which is due this summer. The industry is concerned not only because of the meaningful use program, but also because the quality reporting mechanism will be critical to new reimbursement methods spawned by health care reform.
"A lot of what will happen in the coming years is contingent on these quality measures," Jeffery Smith, assistant director of advocacy programs for CHIME, said. "Performance-based payments across the board will be based on them. Look at the accountable care organization program, for instance. It includes 30-odd CQMs, and that's going to make or break a hospital payday. So it's important to get this right, but we're trending in the right direction."
When CMS issued its Stage 1 final rule in July 2010, the regulation included "e-specifications" that vendors had to follow in reprogramming their EHRs so that providers could collect the requisite quality data. But these e-specs had not been tested and contained many inaccuracies, Ginny Meadows -- executive director of the program office for McKesson Provider Technologies -- said on behalf of EHRA.
She said CMS in its Stage 2 proposal incorporated many CQMs that have not been endorsed by the National Quality Forum or by the new Measure Applications Partnership that the department created to advise it on quality measures. "MAP is supposed to be looking at the feasibility issues and whether these are measures that can be truly implemented electronically," she said.
Then, to make matters worse, CMS did not include the e-specifications in its Notice of Proposed Rulemaking for Stage 2. "We feel the specs should be released for comment, and that vendors, providers and other stakeholders should be given a chance to look at these specs and give feedback on them," said Meadows. "But CMS has been clear that they're not going to do that and that they won't be included in the final rule. We don't know when we're going to get them."
Both EHRA and its parent organization, the Healthcare Information and Management Systems Society, have emphasized the need to field test the e-specifications to make sure that they will work in the settings for which they were designed. But CMS has not said whether it will require field tests before putting the e-specs into EHRs.
Timeframe Too Short
In the view of EHRA, vendors should be given at least 18 months to translate the e-specs into workable software and train health care providers on it. But, even with the Stage 2 start date having been pushed back a year -- to Oct. 1, 2013, for hospitals and to Jan. 1, 2014, for eligible professionals -- the organization said that vendors will have barely more than a year to integrate the new CQMs into their EHRs. That calculation assumes that CMS will release its final rule by the end of the summer, but it could be issued as late as October, according to Meadows.
"This timeframe is unrealistic, and there is risk of unintended consequences through the need to truncate normal testing, validation and training processes," EHRA said in its comments to CMS.
Another danger is that many vendors will not program all of the CQMs into their EHRs by the start of Stage 2. That's what happened in Stage 1, Smith noted. Vendors gave hospitals and EPs a list of the CQMs that they could supply. If the number of CQMs exceeded the requisite number, then providers had a choice of what they could report on. That happened because of the short timeframe and because vendors did not want to program CQMs that they did not think most of their customers would use, he said.
To avoid another situation where vendors limit the CQM data their EHRs can generate, CHIME wants the Office of the National Coordinator of Health IT to require that EHRs provide the ability to collect data for all CQMs as part of base certification. But Meadows argued that this would be impossible because "a lot of the data that feeds quality measures come from other systems that aren't even certified, like lab results. So it's hard to segregate it to the base EHR."
Insufficient Data Sources
In Stage 1, AHA pointed out in its comments, much of the quality data that hospitals need are not in a structured format that makes it easy to collect. In many cases, these data reside in written or free-text physician notes, AHA said. As a result, many hospitals have had to develop custom screens, provide extra training and, in some cases, have chart abstractors fill in the necessary data elements.
This is a human-factor problem that cannot easily be fixed with current technology. "The majority of information in EHRs is largely inaccessible," said Jonathan Weiner, a professor of health policy and management at Johns Hopkins Bloomberg School of Public Health, who has studied electronic quality measures. "Doctors don't like these workflows and these templates, so most of it is in the free text."
Dispute Over Measures
In general, providers and vendors agree that the number of CQMs should be as low as possible, and they don't want CMS to increase the number of hospital measures from 15 to 24, as it has proposed. But they disagree over the details.
For example, CMS has asked for comments on two options for quality reporting by eligible professionals. In option 1a, EPs would have to report on 12 CQMs, but would be able to select them from a fairly long list, including measures that apply to many specialties. In option 1b, EPs would have to report on 11 core measures and could choose one additional CQM from a menu.
CHIME favors option 1a because it provides more flexibility to physicians, Smith said. EHRA supports option 1b because having everyone use the same measures would increase data comparability and help physicians integrate the measures into their workflow, Meadows said. Although she did not say so, having to rewrite software for a smaller number of CQMs would also greatly reduce vendor costs.
Clearly, CMS will have to confront some serious issues related to CQMs in its final rule. What the agency decides in this area will affect the future of quality reporting in government health care programs -- and, likely, in private ones as well -- for many years to come.