Filling out a prescription. Ordering a lab test. Writing a referral. These are all integral elements of clinical practice that can lead to negative patient outcomes if the processes are flawed. Test results may not come back from the lab in a timely way or they may not be seen if they do come back. Patients may not follow through on important referrals or studies for a serious complaint. Pharmacies may mis-interpret an order.
Health Information technology can help with all of these, yet attendees at the Health IT conference in Boston, sponsored by Harvard's CRICO/RMF, heard about solutions that carry their own risks to be aware of and somehow managed.
The Chief Information Officer for CareGroup Health System, a multi-hospital network that includes Beth Israel Deaconess Medical Center in Boston, told the audience that they're living in a brave new world for information technology and medicine. Dr. John Halamka is a practicing emergency medicine physician and chair of the U.S. Healthcare Information Technology Standards Panel.
"Where we're going in 2008 and beyond: remote monitoring, the medical home, the idea that you can have care delivered if a patient is in the home; they've got home blood pressure cuffs; if you want to transmit data to the personal health record, electronic health record, do teleconsultation, or telemedicine, communication between doctor and patient—all of these remote monitoring use cases cover doctor-patient interaction wherever it might occur. If you're getting transferred from inpatient to outpatient or long-term care: how do we insure that your medical record follows you wherever you go including EKG's and all your images. So, by October of this year we'll have standardized all that stuff. So, you'll be able to have an imaging study at the Brigham and go across the street to Beth Israel Deaconess and not have to repeat the study because today we're spending roughly 15% of the healthcare dollar in Massachusetts on redundancy and waste because we're not having continuity of care talking from institution to institution."
Before we get there, along the way, the technology may pose some new challenges. Dr. Dan Rosenthal is Vice Chair and Professor of Radiology at Massachusetts General Hospital. The clinical and professional liability concerns associated with missing or overlooked test results and follow-up of referrals are obvious. Dr. Rosenthal described a situation that highlights potential risks in the area of test result management.
"Outside of the Radiology Department we may have multiple different systems that are used to provide requests towards us. And unfortunately individual prerogatives have been allowed to rule, at least to a greater extent than they should have when contrasted with institutional objectives for simplicity and uniformity, and as a result we get orders coming to us, traditional form by paper with the illegible requests, by fax which is a system that unfortunately we implemented and now can't eliminate. We get telephone orders, which are another system that is a legacy, and we have three different electronic ordering systems, as we were recently told."
That's just the information coming in with the help of technology. How do the reports coming out of radiology fare?
"Now as far as when the report is delivered I alluded to the hazards of incrementalism. We add a new system, we don't eliminate an old one. And as a result we produce our reports in the Radiology Information System, which then hands them off to a variety of other reporting functions. The LMR is certainly one of them, but we have a separate subcategory of that within MGH. Certain physicians prefer a system that was locally developed called On Call and whatever its merits double systems are a problem. We still have the traditional paper reports. We have faxed reports, again the legacy we can't eliminate and we have e-mail systems for delivering results as well."
Dr. Rosenthal suggested that clinical leaders should pursue standards and apply pressure on commercial vendors to resolve some of these issues.
Dr. Eric Poon was on hand to discuss the risks that persist in filling prescription medication orders, even with electronic order entry that communicates directly with the pharmacy. Dr. Poon is the Director of Clinical Informatics at Brigham and Women's Hospital in Boston.
"let me show you some examples of errors. So, thinking first about errors related to the drug product, the 325mg of Aspirin. There were a few examples of errors. Sometimes the drug name is not completely specified, you would think that this would be strange in the electronic world, but that does happen. Sometimes the strength is not specified so the pharmacist is confused. Sometimes the wrong strength is actually specified and I have an example of that in a second. And sometimes we realize that in retrospect the physician might have chosen the wrong product. Or sometimes they realize that they have chosen the wrong product; they print out the prescription and right before they send off the patient with it; they actually modify it with a pen, which in some ways is okay, but it illustrates that there may be opportunities for errors."
Dr. Poon suggested that institutions conduct more usability studies to help clinicians interact better with the system. He provided an example of a new capability provided by the technology that actually introduced new confusion.
"The patient's blood pressure was not well controlled and the provider decided at that point to increase the dose from 5-7.5 of Norvasc and as I said before 7.5mg is not an available product on the market and at this point the provider could have done two things that would have been right. The provider could have actually kept the product the same as 5mg and increased the number of tablets to be taken to 1.5—you can see the provider tried to do two things, but didn't get it right—or the provider could have actually specified the 7.5mg here under the dose field and the computer would have actually calculated it for the provider that this could be satisfied using 1.5 tablets 5mg tablets. So, you can see that while maybe this provider was busy, but you get the sense that perhaps there is a lot of confusion out there about what really constitutes a legitimate prescription."
Among the solutions everyone pointed to is careful and more extensive training. Dr. Poon said that case studies can be good teaching tools, and he offered an even simpler suggestion.
"We really need to figure out how to teach providers how to use the electronic medical record well in the context of the patient. For example, if the patient is telling me about his or her headache or back pain as I'm trying to renew the diabetes prescription maybe I should be empowered to say, ‘Okay, Mrs. Jones or Mr. Jones, I can make a mistake here renewing a prescription; let me pause for a second here and then we'll talk about your other things in a second."
