Randy L. Thomas
Healthcare organizations across the country have spent or are in the midst of spending hundreds of millions of dollars implementing electronic health records (EHRs) and other IT focused on improving quality and reducing cost-the "holy mantra" of IT investment.
Buoyed by numerous studies documenting potential savings, the insistent drumbeat of value-based purchasing, and good ol' intuition that IT can do for health care what it has done for other industries, hospital executives are looking for big returns. But what is success? And how do you measure it? Are we getting all the benefits we should be from these significant efforts to make health care "digital"?
Quality improvement and patient safety programs in hospitals are charged with helping the organization deliver safer, higher-quality care that results in improved outcomes. They are also responsible for ensuring compliance with the various quality reporting requirements-Centers for Medicare & Medicaid (CMS) Hospital Compare and The Joint Commission Core Measures are two examples-that can affect reimbursement and public perception. All of these efforts rely on data to measure base line metrics and plot progress over time.
Too frequently, the acquisition of the needed data relies on manual chart abstraction-done by scarce, expensive talent-that is both time-consuming and prone to inaccuracy. Although this is not a new revelation, what is particularly sad is to watch a highly skilled nurse call up a patient record in the EHR, visually scan the record for needed data, write those data on a worksheet-and then go to another system and enter the transcribed data into a reporting database!
To effectively drive internal quality improvement programs and comply with various regulatory reporting requirements, we need to automatically repurpose the data collected in EHRs and other systems to support these efforts. "Collect once, use many times" should be a guiding principle of data systems in health care. For quality improvement and patient safety programs to be the benefactors of the data collected in the course of patient care, they need analytic tools that take the collected data out of the various transaction systems and transform them into information that highlights not only where care could be improved, but also how. Regulatory reporting needs to be an automatic byproduct of the internal process of quality and safety improvement programs, not the focus of them.
To many, this is overstating the obvious. "We've just spent $100 million implementing an entire suite of clinical systems; of course we can automate our quality reporting!" cries the hospital CFO. Well, yes and no. There are two critical components that must be in place to fully leverage your EHRs for quality and safety reporting. One is directly in the control of your organization; the other you need to influence with your vendors and in Washington.
Fully Leveraging Your EHR
To be able to repurpose the data collected in your EHR, you need to implement your chosen system with that end in mind. For example, you may have the option to collect vital signs in free form text in a progress note (and the text note could be dictated and the transcribed document then "attached" to the patient record) or in structured fields specifically labeled for that purpose. The first option makes the data difficult to repurpose automatically, as a technology such as natural language processing will need to "read" and "interpret" the note to abstract the needed data and then map it to a common vocabulary. The second option supports easier (though not easy!) access for future reporting. Of course, physicians may be more amenable to dictation than data entry, but understanding the tradeoffs between data re-usability and time of data entry are important considerations in your EHR implementation. Knowing what data are needed to support current and anticipated future quality reporting can guide these decisions.
But making these decisions to implement your EHR in a way that more readily supports quality reporting is often not enough. You also need the cooperation of your EHR vendor to document where the data are physically stored in your database so a data appliance can extract the data for loading into an analytic data base that supports quality reporting. Even better would be if your EHR vendor automatically mapped and stored those data in industry-standard formats. These standards, some of which have existed for years and others of which are just emerging, are the second component you need to influence, though less directly than the first.
Since President Bush declared in July 2004 that the United States was embarking on a 10-year mission to deploy a ubiquitous, interoperable EHR, several national groups have worked diligently to define the scenarios and technology standards needed for interoperability. The American Health Information Community has defined a scenario to support the repurposing of data collected via an EHR for quality reporting (www.hhs.gov/healthit/ahic/quality). This scenario is supported by a set of data and transaction standards identified by the Health Information Technology Standards Board (www.hitsp.org).
Some of the required standards are already defined. The rest of the standards are in the process of being defined by various standards development organizations (SDOs). But none of this work will result in any benefit to your organization unless these standards are adopted by healthcare IT vendors and incorporated into their products. We all need to encourage the SDOs to rapidly conclude their work and to encourage the vendors to embrace the current and emerging standards into their products.
The benefits will be significant. If you can automate close to 100 percent of your data abstraction efforts to support quality reporting-for CMS, for The Joint Commission, for various value-based purchasing initiatives-you can realize a direct savings in labor. The rule of thumb is that one nurse abstractor is needed for each 100 beds. Further, the accuracy of your reporting will improve, since the data used for reporting are the same data used in the process of treating your patients.
Measuring the Benefits
Once you have automatically abstracted the data for quality reporting, they can also be used to measure and report on the various quality and patient safety initiatives under way in your organization. You can apply advanced analytics to your data to constantly be on the "look out" for new opportunities for improvement-mining your own data to identify where to focus your quality improvement expertise. Finally, you will be able to determine whether your IT investment is reaping the intended benefits of improving patient outcomes while simultaneously improving efficiency-saving money and lives.
Are your physicians following evidence-based standards of care? Does clinical decision support influence clinical decisions? What is the effect on patient outcomes? The data available to you in your analytic database can now guide your decision making on what, if any, course corrections are needed to ensure you are driving high performance.
Randy L. Thomas, FHIMSS, is a vice president, Premier Inc., Charlotte, N.C. (firstname.lastname@example.org).
Publication Date: Monday, December 01, 2008