The CMS and the Office of the National Coordinator for Health Information Technology at HHS need to tighten up their oversight of the Medicare EHR incentive payment program, according to HHS' inspector general's office.
The watchdog office, headed by Inspector General Daniel Levinson, offered a couple of recommendations for the agencies in its report, "Early Assessment Finds That CMS Faces Obstacles in Overseeing the Medicare EHR Incentive Program" (PDF)
. The report is based on audits of EHR incentive payment attestations, reviews of internal CMS and ONC documents about the program and interviews with CMS personnel. The inspector general's office did not focus this time on the Medicaid portions of the program, although a previous report, issued in July 2011, did, focusing on 13 state-run Medicaid EHR incentive programs. The inspector general's office also is conducting "a series of audits of Medicare and Medicaid EHR incentive payments" to "verify the accuracy of professionals' and hospitals' self-reported meaningful-use
information, as well as eligibility and payment amounts. No time frame for those audits was included in the report.
The inspector general's review covered the early stages of the Medicare EHR incentive program, from when payments started flowing in May 2011 through December 2011. During that period, the program paid out about $1.7 billion to nearly 27,000 physicians and other eligible professionals and 668 hospitals, the report said.
The inspector general said that the CMS validates the presence of some required information and confirms some calculations provided by hospitals and providers. For example, "The validation checks that self-reported numerators and denominators calculate to required percentage thresholds and that all relevant yes/no measures were checked 'yes,' " according to the report. However, the report continued, the CMS "does not verify that numerators and denominators entered for percentage-based measures reflect the actual number of patients for a given measure or that professionals and hospitals possess certified EHR technology."
One "obstacle" the CMS faces in trying to get independent validation that what the providers are attesting to actually happened is that data from other sources—such as Medicare claims or private insurance data—is either incomplete for the task or unavailable.
The inspector general's office notes that although the CMS is not required to perform prepayment verification, "doing so would strengthen its oversight of the anticipated $6.6 billion in incentive payments" the program is expected to shell out over its lifetime, which runs through 2016.
Regarding post-payment oversight, the inspector general noted that, so far, the CMS "has not yet completed any post-payment audits." But the CMS has said it plans to use EHR-generated reports "to verify the accuracy of self-reported information where possible" and obtain supporting documents in instances where the reports don't cover the audit subject matter—and this is where the ONC comes in for criticism.
The ONC oversees the rule writing, and the testing and certification programs to determine whether EHR technology qualifies for use in the Medicare EHR incentive payment program.
The CMS "cannot use EHR reports to verify all self-reported meaningful-use information because ONC does not require certified EHR technology to be capable of producing reports for all meaningful-use measures," the inspector general's report said. The ONC requires an EHR to write reports on the 30 percentage-based measures but not the 19 yes/no measures users also are required to attest to in order to get paid.
"EHR reports also do not contain information necessary for CMS to verify all percentage-based measures," the inspector general's report said, specifically noting that denominators for many of those measures include data from both paper-based and EHR systems.
The inspector general's office recommended that the CMS beef up its prepayment assessment program, including by focusing on "high-risk" professionals and hospitals, asking them to "submit supporting documentation for prepayment review."
It also recommended that ONC "improve the certification process" to ensure that certification bodies "comprehensively test EHR reports for accuracy as part of the certification process" as well as not rely on "vendor-supplied data" during the testing phase.
The CMS, in an Oct. 9 letter from acting Administrator Marilyn Tavenner, said prepayment audits were not necessary at this time, but concurred with another inspector general's office recommendation to issue a guidance on proper provider documentation required for the program.
In a similar letter to the inspector general's office dated Sept. 25, ONC chief Dr. Farzad Mostashari concurred with the inspector general's office's recommendation of testing a "yes/no" reporting functionality. He said he would ask his two advisory committees, the Health IT Policy and Standards committees, to make recommendations "on the appropriate scope and feasibility of a certification criterion focused on 'yes/no' reports."
Mostashari also said the ONC has “already taken steps” to address a separate inspector general's recommendation that it improve its EHR testing and certification program. Specifically, the OIG recommended that the national coordinator supplant vendor-supplied data used in the initial rounds of its certification tests with a standard data set to be used by all vendors.
Last fall, GE warned customers of two of its EHR systems for ambulatory-care providers that errors had been found
in reports to support meaningful-use attestations. That incident was specifically mentioned in the OIG report, which added that the ONC's certification process "did not identify these potential inaccuracies because the vendor-supplied test data did not account for the manner in which some professionals use the products." Similar problems may exist with reports from other EHR products, the OIG report said, but it cited no other examples of report-writing failures.
In his letter, Mostashari said the updated 2014 edition testing and certification rules—which were released in February in conjunction with the CMS' Stage 2 meaningful-use rules
—contain "more rigorous testing requirements" that became effective Oct. 4, 2012. He said the ONC "will continue to migrate away from the exclusive use of vendor-supplied data."
In a telephone interview, Mostashari said the GE report-writing problem was "old news." Asked whether he was aware of any other incidents of EHR systems failing to produce accurate test reports, Mostashari said, "It's really a CMS question."