Hawaii Pacific Health: PACS Takes the EMR for a Drive

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
hawaiiIn the island state of Hawaii, there is a four-hospital system, based in Honolulu, called Hawaii Pacific Health (HPH), with outposts and imaging technology deployed throughout the Hawaiian archipelago. This system is served by four different radiology practices, reading approximately 300,000 studies annually. About three years ago, a systemwide upgrade gave radiologists improved reading from all locations with unprecedented access to everything from physicians’ notes and laboratory reports to referrers’ phone numbers and emails with the click of a tab on their desktops. Effectively, from the radiologist’s perspective, PACS is now driving the electronic medical record (EMR). It is unclear whether this access is speeding up or slowing down radiologists, according to Robert Lipman, MD, a member of the 10-radiologist group that covers Straub Clinic & Hospital (Honolulu) and seven satellite sites on the island of Oahu. There is, however, no doubt that desktop context sharing has provided radiologists with ready access to information that has resulted in greater clarity in reporting and improved patient care. “It makes us do a better job,” Lipman says. Consider this: A radiologist reviewing a CT exam of the abdomen encounters a slight lucency in the femur. Is it or is it not a bone lesion? At HPH, the radiologist hits a button labeled problem list on the EMR (which the PACS has opened to the correct patient’s record) and discovers that the patient has lung cancer. “It would have been nice if the referring physician had put that in the history sent to me, but now that I know it, that bone lucency has significance,” Lipman says. How It Works The enriched reading environment for radiologists at HPH didn’t happen overnight, but instead, occurred in stages as the IT infrastructure (and the system itself) evolved from individual information systems to an integrated whole. Each of the four hospitals in the system and the HPH women’s health center has its own PACS server (Synapse®, FUJIFILM Medical Systems USA, Stamford, Connecticut), and they all share images via Fujifilm’s CommonView® platform. The EMR (Epic Systems, Verona, Wisconsin) is housed on a single server at a data center, and each individual PACS is now interfaced with the Epic EMR. Robert Lipman Prior to the tight integration, HPH had the EMR, but there was no integration with PACS, so the radiologists had to open a separate window, enter an eight-digit patient identifier, and (if they managed to get to the right place) they could search the patient record for the information that they needed. Robert Lipman“Before that, it was worse,” Lipman notes. “We had a RIS, we had a laboratory system, and we had a clinic-notes system. For each different thing we wanted, we would have to log onto that application, and we had to put in a patient identifier. Now, Synapse takes care of that part by identifying the patient to the entire EMR, which includes all of the pieces that used to be individual.” In the current configuration, Epic is always open on the desktop, but PACS drives the workflow. When a radiologist logs onto Synapse and opens a case, Epic synchronizes with PACS and goes directly to that patient; by hitting a tab, the radiologist has access to Epic Radiant (the RIS portion of the Epic EMR) or any number of resources, including clinical history, laboratory results, and phone numbers and email addresses for the patient and his or her physicians. The Desktop The radiologists at Straub Clinic & Hospital read on five-monitor workstations: Four black-and-white monitors display images and one color monitor displays the EMR and what Fujifilm calls its Power Jacket. “The Power Jacket tells you about prior imaging,” Lipman explains. “It tells you that the person had an ultrasound of the liver—and if you are reading a CT of the liver, you might want to look at an ultrasound of the liver.” The EMR, open in the leftmost monitor, includes a view called Snapshot, which Lipman uses to get information on medications and allergies, phone numbers, and Chart Review—the bulk of the record. Chart Review features 13 tabs, including notes, imaging, procedures, laboratory results, medications, cardiology, and another seven categories of useful information. Different specialists can choose to access information in different ways. For instance, rather than retrieving laboratory results through Chart Review, Lipman prefers to use a different view called Results Review, which displays laboratory values and other useful information, in a spreadsheet format. Another radiologist-friendly aspect of the integration is the ability to use the messaging function of Epic, called InBasket: “When you are in a patient’s record, you can hit a button and it will open a message field for that patient,” Lipman explains. “You don’t have to open an email application and type in the patient’s name and accession number. Those things automatically appear; the name, the number, and the study you are reviewing. You just type in the person you want to send it to and what you want to tell him or her.” Patient-care Benefits Before the most recent stage in the integration of PACS and the EMR, radiologists had to work to get to the answers to questions that occurred to them while reviewing a study. “It’s tempting not to look at the clinical information because it is such a hassle. The integration makes the information easily available and, therefore, tempting to look at,” Lipman says. In fact, Lipman estimates that for a queue of 40 to 50 CT studies, he will seek further information (primarily laboratory values and clinical notes) from the EMR on 20% to 30% of those studies. “That is personal, too,” he acknowledges, adding that each radiologist uses their own clinical judgment to determine when to access additional information that may be contained in the EMR. Returning to the example of the study with lucency on the femur, Lipman explains that knowing that a patient has cancer would cause him to seek out the referring physician and tell him or her about the suspected bone lesion. “It tells me when I need to call a clinician and whom to call,” Lipman says. “Sometimes, the physician ordering the test really isn’t the one who knows what’s going on in the area I’m interested in; her or she may be ordering something related to the liver, but I am looking at a problem with the kidneys, and there is a record that a urologist is involved who might be able to help me more than the person who ordered the study. There is a lot of information like that.” Theoretically, the integration should improve radiologists’ accuracy and ultimately improve patient care. Lipman says, “As far as sensitivity and specificity are concerned, it should help both: The more you know, the more you know where to look, and the more you know, the more you should know what the likely diagnosis and the unlikely diagnosis are.” With their records collected in a central place, patients also benefit from a feedback loop that is generated among caregivers, and radiology is an important part of that circle. “Once in a while, I will say, ‘This finding is common in patients with diabetes, but I find no record of that in the EMR.’ That is to catch their attention: It may be because the EMR is not up to date or the patient doesn’t have the condition, but that is up to the clinician to decide,” Lipman says. Workflow Enhancements The integration also has encouraged the radiologists to adopt a paperless workflow by performing study protocols directly in the EMR. Radiologists using this workflow are consulting the EMR for laboratory values for each CT exam, significantly increasing the rate at which radiologists access the EMR in their case work. Each shift makes protocols for the studies scheduled for that shift, and a text message alerts radiologists on duty when there is an addition that needs a protocol. This access to the patient record while preparing protocols gives radiologists a level of certainty that, for instance, a creatinine reading is the most current available—perhaps more current than when the order was written. A high creatinine level would automatically pop up as an alert while the radiologist writes the protocol for the technologist. The PACS-driven workflow at HPH suits radiologists perfectly, Lipman believes, keeping them in the application custom-designed for the specialty’s workflow, but offering access to information as needed. Ideally, he adds, the integration would be bidirectional, so that if he were in the EMR/RIS signing reports and wanted to see an image, the EMR/RIS would tell the PACS to open the image, but that capability is not available yet. This could turn out to be the next evolutionary step in the development of HPH’s enriched reading environment. Cheryl Proval is editor of Radinformatics.com.