I arranged to spend last Thursday afternoon with Peter Yu, at his El Camino Real office of the Palo Alto Medical Foundation (PAMF). PAMF has largely implemented EPIC as their EHR, and as Peter and I had an ASCO-related conference call scheduled, I was going to take the call with him and observe his EHR in action. The call was canceled, but he had three patients whom he had set up to see. The first patient was a young woman with newly operated small “triple negative” breast cancer, without axillary nodal involvement. We had a long discussion about prognosis and treatment, the current unavailability of a PARP inhibitor clinical trial, and the benefits and risks of participating in the available cooperative group trial, involving standard therapy with or without various bevacizumab schedules. Using the exam room computer terminal, Peter was able to use AdjuvantOnline! to illustrate the benefits of systemic therapy to the patient and her family in real time.
Another patient, a fellow with long standing myeloma now on lenalidamide, Peter saw emergently for cough and dehydration. After evaluating him, Peter showed me their CPOE implementaion, which he used to order a chest X-ray, CBC and biochem panel. He was also warned about drug interaction between the levofloxacin he was prescribing, and the warfarin that the patient was taking, and he produced a report for the patient which he handed to him, instructing him on how to adjust his warfarin.
Each exam room has a computer terminal mounted on the wall, the keyboard is set to be used in the standing position, and the monitor swivels so it can be viewed from most parts of the room (see photo). There is a proximity recognition system, so when Peter’s RFID-enabled name badge (which he wears around his neck) is in the vicinity, he is able to use the finger-print recognition system to access the system. (This likely saves minutes each day, avoiding repeated password sign-ins.)
For chemotherapy, they have used Intellidose for the last four years—pre-dating availability of EPIC’s Beacon Module, which Peter says is coming—and for documentation, they use a combination of templates and macros for routine patient care, and voice-recognition software from Dictaphone for the longer narratives. Though it sounds klugey, Peter was able to navigate through the screens and programs with alacrity, and his three visits that day would have, in aggregate, met the current iteration of “meaningful use.” (including e-prescibing, warning of medicine interaction, handing the patient a visit summary, and CPOE.)
I was also able to meet with Charlotte Mitchell, RN, who is the project manager of their EPIC implementation. She described a rolling implementation, with their metric being one analyst or clinical support person per fifty physicians. During their rolling implementation, Peter told me that the ratio seemed more like one-to-one, as there was a support person at or near the elbow of each doc, or at least available in the hallway outside the room. Charlotte also told me that they adjusted each physician’s clinical load during the implementation for a period of two weeks, and cut their patient numbers by fifty percent. When docs and nurses went for training, their clinic coverage also had to be assured. After one month, she thought that productivity was generally back to where it should be. She noted that some of her docs could close a visit (meaning complete their note and their bill) before walking out of the exam room.
Later, on Friday night at social function, I ran into a PAMF orthopedic surgeon, who is also using the EPIC system. A 1983 high school graduate, and thus putatively more computer savvy than my generation, he allowed how the availability of medical records was better with their EHR, and when he went to see one of their patients in the hospital, if they have a good internest, the H&P had been done. However, his distinct and strongly held opinion that the EHR slowed him down in the office: whereas in the paper world, his medical assistant would line up the X-rays on the view box, now he had to wait for his PACS to load the images after he signed on to the computer terminal; formerly he would have documented his findings on one page of paper, wherehas now he had to point and click, dictate, and type his way through a note; in the paper world he would have checked boxes on a form to schedule the return visit and submit charges, now he had to point and click his way through the form. He did tell me that he never filled out an X-ray requisition—he had a point that his time was too valuable—and he was happy to delegate that to his medical assistant. He did have some fair points—the hardware and network did not operate at “think speed”—usually defined as ½ second—for his x-ray viewing, the documentation software probably wasn’t the most user-friendly for an orthopedist likely suited by temperament and training to expressing himself in drawing and other non-grammatical written modes.