The trusted source for
healthcare information and
Ten years after the passage of the Health Information Technology for Economic and Clinical Health (HITECH) Act, researchers are sounding the alarm about patient safety concerns related to the poor usability of electronic medical record (EMR) systems. They also are highlighting how poor usability leads to clinician burnout, which heightens the risk of medical errors. Proposals for change include a call for greater transparency so that EMR program flaws can be shared and addressed quickly.
This year marks the 10th anniversary of the Health Information Technology for Economic and Clinical Health (HITECH) Act, the legislation responsible for pushing healthcare providers to adopt electronic medical record (EMR) systems. However, rather than being a cause for celebration and accomplishment, the anniversary marks a tipping point for EMR users in their exasperation with systems that many find clunky and burdensome to use.
Emergency clinicians have been arguing for years that they spend more time sitting in front of a computer screen than interacting with patients. But now the literature base is adding new fuel to such arguments. Researchers have found that issues related to EMR usability are raising significant patient safety concerns that need to be addressed. Powerful forces are pushing for action.
In one recent study, led by researchers at MedStar Health’s National Center for Human Factors in Healthcare (NCHFH) in Washington, DC, investigators evaluated the use of EMRs from two of the largest EMR vendors, Epic and Cerner, at four ED sites. At each participating site, between 10 and 15 emergency medicine physicians performed six standardized clinical scenarios on their own systems. Researchers recorded the number of clicks required and the amount of time it took to complete each task.1
“What we found was actually pretty shocking to us,” explains Raj Ratwani, PhD, director of the NCHFH, an assistant professor of emergency medicine at Georgetown University School of Medicine, and a co-author of the study. “We found, number one, huge variability across the different sites. If you look at two Epic sites or two Cerner sites, you were getting very different results in terms of the number of clicks to perform a task, the amount of time it takes to perform a task, and even the error rate [associated with the task]. If you look across vendors, there was also huge variability,” he says.
These findings illustrate just how important implementation of an EMR system is, Ratwani notes. “It highlights how critical the customization and configuration of a product is from the provider side. It is highlighting what providers really need to do to improve their systems because there is so much variability there,” he says. “This suggests some providers are able to customize and configure [their EMR systems] more effectively — and others less so.”
Secondly, the study illustrates some of the safety issues that exist in products that have been certified by the federal government and are in use across the country, Ratwani notes. For example, one of the tasks researchers examined in the study involved ordering an X-ray of the left elbow, wrist, and forearm. “When the physicians do this, we see error rates ranging from 16% at one site all the way to as high as 36% at another site,” he says.
This error means that physicians could be ordering an X-ray on the wrong side or of the wrong body part, or they issued an incomplete order that only called for an X-ray on two of the three body parts, Ratwani observes.
“This has some pretty serious consequences. There could be unnecessary X-ray exposure, delays in care, and all kinds of other things,” he says.
Another task the researchers evaluated was a prednisone taper medication order, a direction to reduce the dosage of medication over time. In this case, the researchers observed error rates ranging from 16% to as high as 50% across the four sites.
“Those are very, very high error rates in systems that have been certified and are used to treat patients,” Ratwani stresses. “We were very alarmed by the error rates we were seeing and the amount of variability.” Using this study as a foundation, MedStar’s NCHFH has teamed up with the American Medical Association (AMA) in a campaign to push for changes that will address the known risks to patient safety as well as the contributing factors to clinician burnout that are related to poor EMR usability.
“We know that EMR usability is a major contributor to clinician burnout; that is harmful to our clinicians and it also indirectly has patient safety consequences. A clinician who is burned out is more likely to make a medical error,” Ratwani explains. However, he notes that suboptimal usability also directly affects patient safety. The NCHFH/AMA campaign is concerned with both issues.
The effort includes a new website, , that contains video demonstrations of the EMR usability challenges that the researchers uncovered in the study, and a call for all stakeholders to push for improvement in this area. There also is a letter directed to Congress, which urges broader action to address the EMR poor usability problem. (Editor’s Note: Read more at: .)
“We would like to have conversations with the Office of the National Coordinator for Health Information Technology [ONC] and with members of the House and Senate to help them understand some of the big challenges [related to EMR usability],” explains Ratwani, who authored the letter. “We would like to move toward creating a central reporting system so that any clinician in the country can report [EMR-related problems] to one central repository so that we can do some rapid analyses of those data, understand where the challenges are, and work to quickly address them on a national scale.”
In fact, the letter points out that Section 4002 of the 21st Century Cures Act, passed in 2016, calls on the ONC to create an EMR reporting system so that the office can collect and distribute information on core EMR features, including usability. As part of this task, the letter states that Congress should ensure that the ONC includes “objective measures of safety as part of the usability criteria.”
The letter to Congress also references another recent study of 9,000 health information technology (HIT) and medication safety events in three pediatric hospitals. Researchers found poor EMR usability contributed to one-third of the errors.2
Currently, so-called gag clauses, established by EMR vendors, have prevented the open discussion of EMR-related usability and safety issues on a more national scale, Ratwani observes.
“What may well be happening is you have one patient being harmed at a healthcare system or a provider in the Northeast that involves [an EMR usability issue] that could be completely addressable, but because of gag clauses, we are not able to share that issue,” he says. “Then, a patient could be harmed on the West Coast under a very similar context because we were not able to share [that visual information].”
In fact, providers have complained about these gag clauses in vendor contracts for years. Now, there is some movement toward progress in this area. The ONC has proposed a rule that would ban gag clauses. At press time, the ONC was receiving commentary on the proposed rule. Ratwani applauds this action, but reiterates that more steps are needed to address EMR usability problems effectively.
“Gag clause removal would allow us to report [EMR problems], but then we would need to have all those reports in one place so that we can actually analyze them and make rapid improvement,” he says. “If we want to solve the challenge, we need to have a central reporting system or another mechanism to bring all those reports together.”
Another provision in EMR vendor contracts that providers have long disparaged is the so-called “hold harmless” clause, which essentially frees vendors from any liability if patient harm occurs as a result of an EMR-related issue. Ratwani maintains that such clauses need to be abandoned in favor of a system in which accountability can be shared.
“If there is a health IT-related or an EMR-related safety issue that occurs, we need to look at that issue and determine who is accountable,” he says. For instance, it may be the provider, the EMR vendor, or some other stakeholder who is at fault, but there also may be some fault on the part of two or more participants in the equation.
“These issues have to be addressed with the lens of shared accountability, and there also has to be shared responsibility for making improvements to the system,” he says. “I think absolutely, [hold harmless] clauses need to be removed, and there has to be a better sharing of what is happening.”3
Interoperability, or the ability to exchange electronic health information between different healthcare settings that may use different EMR systems, has long been a goal of providers. Ratwani notes there has been some progress in this area. Further, he shares that the ONC is proposing new provisions that would improve the exchange of information. However, he states that usability needs to be a big part of the interoperability conversation going forward.
“You have to think about not just exchanging information ... but having the information presented to the clinician in a way that they can actually use it,” Ratwani shares. “It is not going to be helpful if we exchange all that information and it just gets dumped into the patient’s record.”
In that event, the physician would have to spend extensive amounts of time searching through all the information to find what is relevant. Some of this information may be contradictory to what is in the record already, Ratwani notes. “You can see how this could potentially introduce some very serious usability and safety challenges,” he says. “As we talk about interoperability and the exchange of information, we must also include discussion of the usability of that information. We think that is going to be a really important next step as we have this conversation about interoperability.”
Ratwani is hardly alone in his efforts to highlight safety issues relating to the use of EMRs. The ECRI Institute has cited the use of EMRs to communicate diagnoses and manage test results as its top patient safety concern for 2019. The designation is based on millions of reports the institute has received regarding errors, near misses, and safety concerns that have surfaced in healthcare facilities across the United States, explains Lorraine Possanza, DPM, JD, MBE, program director of ECRI Institute’s Partnership for Health IT Patient Safety. “ECRI’s PSO [patient safety organization] analysts look at all the information submitted through reports, they have weekly meetings to discuss these events, and through this [review] the analysts begin to see patterns occur,” she explains.
The ECRI Institute’s top concern for 2019 stems, in part, from the organization’s focus on “closing the loop,” providing the right information to the right people at the right time when a patient undergoes any type of diagnostic testing. “You begin with a patient who has to undergo the test, and you follow the course throughout the process from ordering the test to getting the test performed to getting the results completed to getting the results to come back to the provider,” Possanza observes.
Then, the provider needs to share the results and the plan with the patient, make any recommendations that stem from the results, and follow up with the patient. “In order to get a good diagnosis, you have got to make sure that all of those loops are closed. That is true whether it involves diagnostic testing such as labs or radiologic imaging or referrals and consultations,” Possanza says.
To carry out all these tasks, clinicians rely increasingly on technology and assume that the technology is going to handle certain tasks. However, technology is not always optimally leveraged as intended. “When you add any new technology to a system, you always have those hiccups that occur. There are things that maybe aren’t functioning the way you originally thought they were functioning,” Possanza says.
For instance, when reviewing all the safety reports from participating organizations, ECRI investigators discovered that it is important that information entered into an EMR is coded consistently in structured fields and coded very precisely. If this does not happen, the information may not be interpreted by other systems in a consistent manner, Possanza notes. “You may write an order for a lab test indicating that you want the patient to have a ‘CBC’ ... and someone else may write for that same test that they want their patient to have a ‘complete blood count,’” she says. “When that information comes back from the lab to the EMR, if one system called it a ‘complete blood count’ and the other called it a ‘CBC,’ the information may not file into the patient’s record correctly or it may not be recognized as the same exact test.”
Related issues may come into play, too. For example, Possanza notes that providers can become overburdened by the fact that they receive many alerts and notices coming into the record and communication files related to test results. “Are they actually seeing these alerts, or are their inboxes just too full?” she says. “Does the clinician read and react to each notice or conclude that he or she will react to the notice at another time and then forget to return back to it?”
Also, Possanza notes that patients are becoming increasingly involved with EMRs. Communicating and engaging with them appropriately is another issue that provider organizations need to address.
“One of the things that the Partnership for Health IT Patient Safety has recommended is to look at how we can better use patient portals,” she says. “We are working with specific organizations to actually put safe practice recommendations for test tracking to close the loop into focus.”
Possanza notes that her organization created a workgroup to study ways to improve HIT safety. Several EMR vendors/developers have been involved.
“They, too, are concerned with safety ... and I think sometimes they are very surprised to learn that their product didn’t function as they thought it was going to in the workflow or that a feature wasn’t used in the way they thought it was going to be used,” she says. Possanza notes that it typically takes one to three years for a vendor to develop, test, and release a feature and then for providers to use the feature widely.
Working with vendors is just one way to address HIT safety challenges. Another is focusing on the communication channels connected with other parts of the healthcare system, advises Patricia Stahura, RN, MSN, a senior analyst and consultant at the ECRI Institute.
“Sometimes, a patient is discharged from the ED before you have all the test results, which can create a dilemma for the folks in the ED because they have to track down the patient and they may not have the correct contact information,” she explains. “Then, the patient’s primary care provider often needs to be notified as well to confirm a diagnosis or perhaps go in a different direction.”
Such communications are complicated by the fact that providers in other healthcare settings or systems often use different EMRs. Still, even if they use the same platforms, Stahura notes that everyone in the healthcare continuum needs to create a backup to their electronic systems.
“We rely on EMRs, and we have very high hopes for that communication, but I always urge people to pick up the phone and to continue to have that human-to-human communication,” she says. “I think that is very important to make sure that the loop is closed, especially when you find out there are gaps in the system.”
Financial Disclosure: Physician Editor Robert Bitterman, MD, JD, FACEP, Nurse Planner Nicole Huff, MBA, MSN, RN, CEN, Author Dorothy Brooks, Editor Jonathan Springston, Executive Editor Shelly Morrow Mark, Accreditations Manager Amy M. Johnson, MSN, RN, CPN, and Editorial Group Manager Leslie Coplin report no consultant, stockholder, speaker’s bureau, research, or other financial relationships with companies having ties to this field of study.