EHR Quality Measurement In Its Infancy, Study Says

11 Healthcare-Focused Business Intelligence Tools

Quality performance measures based on data in electronic health records are still in their infancy and have yet to tap many of the unique features of EHRs, according to a new study in the International Journal for Quality in Health Care. The study, which was re-published in Medscape, provides a conceptual framework for defining levels of electronic quality measures, or e-QMs.

The study proposes the following five-level typology for defining e-QMs:

--Translated e-QMs. Measures designed for use with paper records, such as whether patients with diabetes have received HbA1c tests. These measures can use claims data or information from chart abstraction, as well as EHRs.

--Health IT-assisted. Measures that could be derived from non-EHR data sources, such as blood pressure or body mass index information, but that require EHRs for reporting on 100% of a patient population.

--Health IT-enabled. Metrics that take advantage of an EHR's features, such as the percentage of abnormal test results read and acted upon by a clinician within 24 hours of receipt, or the percentage of relevant clinical alerts that are acted upon.

--Health IT system management. Measures of how providers use health IT systems, such as the percentage of all prescriptions ordered via electronic prescribing.

--E-iatrogenesis. Measures of patient harm caused at least in part by the health IT system, such as the percentage of patients for whom the wrong drug was ordered because of an error in an e-prescribing system, or the percentage of critical lab findings that did not lead to patient notification.

[ Read about how new technologies are revolutionizing healthcare IT. See How Mobile, Cloud Are Transforming Healthcare. ]

Although about half of U.S. doctors have some kind of EHR, less than a quarter of health care is "substantially documented" in electronic records, the study said. As a result, many physicians with EHRs are stuck at level 1 of e-QM capability.

Academic healthcare organizations and health IT pioneers such as Kaiser Permanente are capable of collecting level 2 quality data, said Jonathan P. Weiner, the lead author of the study and a Johns Hopkins professor of health policy and management, in an interview with InformationWeek Healthcare. "Sophisticated systems have moved into stage 2. They wouldn't think about gathering data on blood pressure or lab results on 100% of their sample with a paper chart audit, but it's not a big deal to do that with an EHR."

Even the most advanced organizations, however, have just begun collecting level 3 data, Weiner said. "I've been working with some of the leaders in health IT, and they're barely scratching the surface in terms of new ways to measure performance. They've nailed the current ways of doing it. An EHR allows one to get there more quickly, but I haven't seen many go beyond that."

Weiner noted that there are structural barriers to level 3 performance measures, including the complexities of workflow and diagnosis. Moreover, he said, physicians don't want to go through a "click-tree hell" in which they have to navigate multiple menus to perform simple actions. That's one reason why much of the data in EHRs is in the form of free text, rather than structured data.

In the American Hospital Association's comments on the Meaningful Use Stage 2 proposals, the AHA pointed out that many hospitals have had difficulty using their EHRs to collect quality data for two reasons touched on in the Johns Hopkins study. First, the association noted, the underlying measures themselves were developed for manual chart abstraction and had to be adapted to the EHR--in other words, they were "translated e-QMs." And second, much of the requisite data is locked up in dictated or written physician notes. In other words, the hospitals either haven't introduced electronic physician documentation, or the doctors aren't using it.

Weiner agreed with AHA's critique. But as EHRs become more familiar to physicians, he said, their documentation is bound to improve because it is the medico-legal basis of their records. Also, when EHR quality data begins to matter in reimbursement, physicians will want to document what they've done in a structured manner. Down the line, he added, natural language processing might make it easier for doctors to enter structured data.

Weiner said the measurement of errors caused by EHRs is very important. "We don't want to build in inefficiencies or safety problems or outright errors," he noted. But he also seconded the observation of Dean Sittig, an expert at the University of Texas Health Sciences Center in Houston: The benefits of EHRs in preventing medical errors far outweigh their potential to cause mistakes.

Get the new, all-digital Healthcare CIO 25 issue of InformationWeek Healthcare. It's our second annual honor roll of the health IT leaders driving healthcare's transformation. (Free registration required.)