Is your health IT a danger to patients?
Is your health IT a danger to patients?
IOM report doubts you're doing all you should
The Institute of Medicine (IOM) released a report in December that was critical of efforts made thus far to ensure medical errors associated with the use of health IT are minimized. It calls for greater oversight by the public and private sectors, and suggests that the Department of Health and Human Services (HHS) publish a plan within a year on minimizing risks associated with health IT, along with annual updates of progress made.
While billions are being invested in health IT by both public and private entities, the report notes that there is no solid evidence that using technology will improve quality of care and patient safety, and there is concern that some technologies may actually lead to harm being done to patients. The report notes that there has been little published about the risks associated with health IT, in part because some vendors don't like for safety-related information to be freely disclosed.
Knowledge is power, though, and vendors and users alike need to report any deaths, injuries, or other events related to health IT, the report continues, adding that Congress should establish an independent federal organization to investigate these adverse events.
Automated systems not inherently safer
While of interest and important, the report doesn't include software for medical devices, says Christina Thielst, MHA, FACHE, an independent health administration consultant based in Santa Barbara, CA, with more than three decades of experience in hospital administration, risk management, and patient safety.
"Bedside and other medical devices are increasingly embedded with software that captures patient health information and data and also interfaces with EHRs and other health IT," she says. "This means that the worlds of biomed and IT will have to merge and interface, as well. If these two hospital departments are not working together to ensure safety, there could be problems." On a positive note, the FDA has begun to look more closely at the topic, which could be a reason it was left out of the report.
Thielst says that the notion that just because we are moving from a paper to a computerized system doesn't mean it is inherently safer. "You are changing from humans dealing with paper to humans dealing with machines. There are still humans involved and we make mistakes whether we write something down or enter it into a computer or forget it altogether," she says. It's true that some programs make it a lot harder to leave some important field incomplete, but the accuracy of the data input into the field is just as important.
It seems logical that something printed will be easier to read and thus less prone to errors than something scribbled in a physician's notoriously illegible hand. "But there isn't really any solid evidence that an electronic record is safer," says Yosef D. Dlugacz, PhD, of the Krasnoff Quality Management Institute at North Shore-Long Island Jewish Health System in New Hyde Park, NY.
Having mechanisms that automate entry, libraries that provide information about dosages and side effects while you are inputting information, and automatic calculations for dosage conversions all potentially help patients. But the problem is that humans are fallible, as Thielst says.
Another issue is that because errors don't always lead to adverse events, the electronic record could just as easily become a database for errors you don't look for. The assumption is that it's better than a handwritten record, he says, so why look for a problem there?
People have been waxing poetic about the benefits of computerized physician order entry (CPOE) for years. "One attractive thing about it is that there is software that checks for errors and problems," says Thielst. "But humans create the templates and order sets. It is also fairly rigid and doesn't accommodate even appropriate deviations."
For instance, a physician may want to use a dose of medication or a different medication based on information that is relatively new and the CPOE system hasn't been updated. "If it's not in the system the way the physician orders it, he can't override it. That can end up causing delays in treatment."
Providers also don't have adequate access to support. Because many of these systems are new — and they change rapidly — a 24/7 support system is imperative, Thielst says. "If a physician is working on a weekend night and can't remember his password, he can try so many times that he's locked out of the system," she says. "You don't want to keep that care from the patient because a doctor doesn't know what to do next." Some physicians are so flummoxed by new technology that they have another physician or a nurse take care of it for them. "Then you are adding another human into the mix and a new layer of risk."
A newer risk that Thielst says she is hearing more about are malicious data breaches and attacks on systems. "This is an issue in the future," she says. "How will you protect against that?" Authentication of users and of the data will be a big issue going forward.
Health IT is all about data-driven reporting and decisions. "Right now, there is a shift in paradigms. Physicians have gone from a system that emphasized knowledge, experience, and expertise to a system where they trust data," says Thielst, "But what about their instincts? What should they do when their gut disagrees with the data?"
Another issue is trust in the people making the technology. For some companies, the money being invested in health IT is a giant draw into an industry they don't have experience with. Thielst warns: "You have to make sure that what you buy is built by people who understand your needs. Do they have the expertise to make a good, safe, quality product?"
And if there are adverse events, Thielst isn't sure that another entity to which you should report it is a good idea — or even necessary. "We already report patient deaths, injuries, and other events to state and federal organizations," she says. More important would be noting that technology was part of the adverse event. Creating a way to track that is something that can happen much sooner than a new body can be created.
What to do now?
Dlugacz says he'd love to see every high risk environment — oncology, chemotherapy, dialysis, intensive care and the emergency department — have a Pharm.D. "If there is someone who can give ongoing assistance in delivering different drugs, who can oversee the smart pumps and other technology that is designed to prevent errors, I think we can get better information on near misses and where errors occur." People only report things that happen if something bad results, he says, but errors happen all the time, at every hospital. Knowing where they occur is important to preventing them, and no machine will be able to catch near misses as well as a trained professional.
Dlugacz notes that most experts agree that when you do something a lot — such as giving medication in an acute care setting — you can expect an error rate of about 2%. But no hospital will show that kind of medication error rate. "We just aren't in tune to looking for all of our errors, and they think that technology can't make a mistake." He adds that he would like to see an organization that admitted to a 20% error rate because its leadership and clinical staff would know they had something to work on.
He'd like to see an audit of all medications — who gave them, the dosage, why, and how that compares to what is appropriate. The Pharm.D. can then pinpoint where the potential for harm existed, even if none occurred, says Dlugacz. See how that matches up with the drug reactions you have documented — are the things that happened in the documented cases similar to things that the Pharm.D. found in the audit that didn't cause harm? If so, you may be able to pinpoint a problem that no electronic health record could.
Hospitals should also audit the medications received by a random selection of patients who had their dosages computed by technology, not a person, Dlugacz says; then you should look to see if the calculations were correct.
One idea that Dlugacz thinks would be great to steal from surgeons is the time out. "ICUs have a lot of technology and are highly complex. You have very smart people working to take care of the sickest patients. But how many of them stop to think before they give a medication? People are busy in a hospital, but shouldn't they stop and check, like they do with surgery patients?"
Thielst says you should pull together a work-group with the specific purpose of assessing the risk to the organization from health IT. That group, made up of key stakeholders, should prioritize what needs to be addressed first. "Some things will be issues like integrating electronic patient safety into your existing patient safety program," she says. "How will you do that? How often will you look at it?"
Second, she recommends that physicians and other providers have a strong training program that they are required to attend and complete — no leaving in the middle because of a call — and that there is back up support always available for them.
Third, she says you should look at workflow issues and make sure that templates and order sets that go with any electronic process are clean and fit with the way care is delivered in your organization. "Assuming you are going to purchase a program, plug it in and go is asking for problems. Make sure it works for your specific needs."
Finally, Thielst says everyone should read the Joint Commission alert from Dec. 11, 2008, on the topic. More information is also found at her website and associated blog at http://thielst.typepad.com.
For more information on this story contact:
- Christina Thielst, MHA, FACHE, independent health administration consultant, Santa Barbara, CA. Telephone: (805) 845-2450. Email: [email protected].
- Yosef D. Dlugacz, Ph.D., Krasnoff Quality Management Institute, North Shore-Long Island Jewish Health System, Hyde Park, NY. Email: [email protected].
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.