The explosive growth of Facebook and MySpace illustrates the market for electronic tools to enhance communication and collaboration. Could there possibly be another workplace more in need of social networking tools than the modern hospital?
If you are not familiar with Facebook, find yourself a teenager and take a look over his shoulder while he is using it (mine are available for rent if you get desperate; the best time to catch them is when they should be doing homework). In one thrilling, chaotic electronic e-universe, the site allows users to exchange instant messages with dozens of friends, to post pictures and videos, and to link to virtually everything on the Web – all at the same time. John McCain would be flabbergasted.
Now, take a look at today’s medical record, and ask yourself whether – if we could start fresh – this is the tool you would have constructed if your goal were to allow a diverse group of providers to collaborate while caring for terribly sick patients. In particular, would you have members of each tribe – docs, nurses, physical therapists, nutritionists – writing notes in their own style, using various totems and ritualistic phrasings, in files separated by colored dividers that might as well be electrified fences?
Of course not. In fact, today’s medical record virtually guarantees the silo-ization of care. Few physicians ever read nurses’ notes, even though all of us depend on the nurses to be our eyes and ears. And the situation iteratively worsens every day. Why would a nurse, realizing that no doctor ever reads her notes, even try to write them to be useful to physicians? And visa versa, obviously. Over the years, this divergence has been codified into ritual, calcified by templates, and hard wired through regulations whose original rationale no one can remember.
Even within the tribes, it’s not much better. If you are a physician, have you tried to decipher an ophthalmologist’s note recently? It might as well be carved in Sanskrit.
The result is that the patient’s chart – which cries out to be a collaborative tool – is precisely the opposite. Since the transmission of information is so crucial to the conduct of patient care, a maddening workaround emerged: the paging system. If you’re a physician, think how many of your daily pages (or, if you are très modern, cell phone calls) could have been handled via collaborative written interactions, if such a forum existed in the chart. If you’re a nurse, think how much of your time you’ve wasted paging doctors, since you lacked a method of posting a comment or question that would be reliably seen and addressed.
The need to bring social networking-, Web 2.0-type thinking into the workplace is not unique to healthcare. For example, the CIA and FBI have built a Facebook-like system, called “A-Space,” and will launch it later this month. “It’s every bit Facebook and YouTube for spies, but it’s much, much more,” said Michael Wertheimer, assistant deputy director of national intelligence for analysis. (He could have given us more detail, but then he’d have to kill us.) Our spooks will be mining the conversations for patterns that would be unlikely to emerge (read: 9/11) absent this kind of tool.
You’d think that medicine’s conversion from paper to electronic records would solve many of these problems, but – to date – all it has done is create new-fangled electronic silos. In most EMRs, including the GE system we’re using at UCSF, the notes are really just electronic incarnations of what previously lived on dead trees – no more likely to facilitate collaboration than the paper records they replace. The EMR is still progress, mind you (at least people can read my writing), but it doesn’t address the fundamental problem.
Let me describe an interesting natural experiment that vividly demonstrated to me the need for a collaborative social networking tool within the medical record. When the housestaff duty hours limits first hit in 2003, we began focusing on how to improve resident sign-outs, which would soon nearly double as we shifted to a world of Day Floats, Night Floats, Weekend Floats, Root Beer Floats… The result was a homegrown program called Synopsis – a stand-alone module that lived on a couple of computers in our housestaff lounge (I’m hoping we’ve passed the HIPAA statue of limitations, since the system wasn’t even close to compliant), on which residents could keep a running update of their patient’s condition, including meds, problems lists, and – most importantly – to do lists and “if/then” statements.
Even in its early, primitive state, Synopsis was a blessing, but a few problems arose quickly. First, about a week after we launched Synopsis at UCSF Medical Center, I got a call from a chief resident at San Francisco General Hospital, our sister (but unaffiliated) hospital through which our residents rotate. “Who is doing the IT support for Synopsis?” she asked. It turned out, we had just worked out which of our IT folks would be supporting the system – at UCSF Medical Center, where the system had been implemented. But why was the chief resident calling from SFGH, I asked. “Oh,” she said, “ one of the residents really liked the program, loaded it on a CD, and brought it to SFGH yesterday when he came for his clinic.”
Does anybody know the antonym for “firewall”?
The SFGH migration demonstrated that Synopsis was addressing a crucial need. But a second problem emerged that is more germane to today’s topic. Within a few weeks of the launch of Synopsis, we began getting calls from the nurses at UCSF, begging for better access to the program (at that time, it lived only on our departmental computers; it has since been embedded in the official hospital EMR). At first I wondered why – after all, this was a program build for the housestaff, largely by the housestaff, to fill a gap: preventing fumbled resident-to-resident sign-outs.
Well, that’s precisely what the nurses loved about it. Turns out that the nurses, literally starving for information from the docs about what was actually happening to their patients, took to Synopsis like a moth to flame – seeing it as the only place in the chart where one could get the real scoop about the patient and the plans. So they literally begged to have access to Synopsis, and then to leave notes to the residents on it. Ditto the social workers and case managers. It seems that these non-MDs didn’t like having to page the overworked physicians for every little thing, any more than the residents liked receiving their daily barrage of pages.
That’s when I finally got it: how great would it be if, through the medical record, I could interact with multiple specialists who have seen my patient – in real time, just like my kids are interacting with far-flung friends on Facebook. And if nurses could leave me a note which I could answer online without having to respond to a page. And if the daily plan for a patient – developed collaboratively – could be shared among all the caregivers, with notes appended when a patient’s clinical ship seemed to be blowing off course.
Development of a Facebook-like medical record would not be trivial. Russ Cucina, my hospitalist colleague and window into the IT world, tells me that some of the big EMR vendors have begun working on it, and I wish them Godspeed. I recognize that even if such a module existed, the need for physicians and nurses’ notes, in their classic stilted style, would not evaporate, largely because of regulatory and billing requirements. It will be crucial to look hard at what is truly required and what is simply ritual and custom, and to jettison the latter while we try to modernize the former. In the meantime, an EMR with Facebook-like functions will need to thoughtfully import other parts of the medical record to minimize redundant work.
There are other challenges. For example, I wonder how we’ll ensure that a more free-form portion of the medical record doesn’t become trivial or gossipy. Why worry? When we first brought Synopsis into the EMR, housestaff often didn’t realize that it was just as “official” as their progress notes, leading to problems ranging from the widespread use of unapproved abbreviations to comments like “try to avoid Mrs. Jones’ daughter – she’s a real pain.” Obviously, we’ll need to both educate the providers and implement some electronic tools to ensure that the site’s use is professional and passes regulatory muster. As one example, our EMR now automatically finds verboten abbreviations (“qd”, “MSO4”) and prompts you to replace them. Though the system is pretty slick, it doesn’t yet flag words like, “…she’s a real pain.”
I know we have lots of readers who are IT gurus, and many are working on various kinds of collaborative and social networking sites in healthcare (a field that has assumed the moniker, “Health 2.0”). But the vast majority of this activity has focused on activities like patient-to-patient support groups (such as patientslikeme), patient information sites (e.g., WebMD), and personal health records such as Google Health and Microsoft HealthVault. I’ve seen far less discussion about creating a Facebook for the medical record.
I’d love to hear from some of our IT-oriented readers – perhaps there is more activity in this area than I know about. I hope so – this seems like a vital tool to support the kind of collaborative care that sick patients really need.
Bob,
As always, a provocative presentation. This would be a great experiment. Think of all the unintended consequences, on both sides. You should perform this work at UCSF as there is nary a company or company employed software engineer who will come orders of magnitude close to gettting it right. Just as your teenagers preferentially work on facebook instead of homework, do you think that there is just a slight chance that health care professionals will prefer “patient book” to seeing the patient? What you suggest might work if the patient has equal access!
Best regards,
Menoalittle
Bob,
You are right on target here. Though CPOE and the EMR have undoubtedly made many significant contributions to patient safety, one of the many unintended consequences has been to take the “inter” out of interdisciplinary. All too often now we have segments of the healthcare team looking at the patient in isolation and not seeing the whole picture.
For example, we did a recent column demonstrating how most CPOE systems have resulted in pharmacists just seeing a short “order string” that conveys an order for one specific medication. They lose the pattern recognition capability they previously had. That pattern recognition allowed them to eyeball a complete order set and get a “snapshot” of the patient.
Note that the same loss of pattern recognition may affect nurses and physicians to a lesser extent. Nurses used to take off all the physician orders so they had the opportunity to see that “patient snapshot”. Now the medication orders go directly to the pharmacy, PT orders to the therapy department, etc. and nursing may only see the nursing orders. Nurses often tell us that the admission orders plus the “face sheet” tell them much more than the admission H&P does!
So how do we deal with this lost opportunity to use pattern recognition? We need some way to restore that patient “snapshot”. One way is by customizing the computer screens for pharmacists and nurses, much the same way we allow physicians to customize the screen layouts to their liking. Most CPOE screens have the patient name, date of birth, other identifiers, room number, physician name or name of service, and allergies listed at the top of each screen. But the view can be customized to show select laboratory data, a problem list (though you’d be surprised at the difficulties generating and maintaining good working problem lists!), and a full medication list. You can also show select data relevant to a specific order. For example, when a physician orders digoxin, we show him/her a popup screen with the most recent K+ level, creatinine or GFR, and any recent digoxin levels. No reason we can’t popup that same screen for the pharmacist when he/she opens that order for digoxin.
But neither of those solutions provides all the things we saw when the full admission order set was visible. Therefore, ensuring that anyone can get a printout (yes, even a “paperless” facility can use paper sometimes!) or full screenshot of the full admission orders is advisable. The key point is that you need to work with your end-users when you are planning your CPOE rollout to find out what is important to them.
Your concept of doing a “Facebook” or “My Space” type of format that would allow much easier access to parts of the EMR that are relevant at various times and would promote communication between the multiple disciplines caring for a patient is most intriguing! Maybe we need to cast HIPAA aside and let our kid play with the EMR for a few hours. They’d show us how to get the most out of it!
Bob,
The comments of Dr. Truax ring both pertinent and paradoxical as he depicts yet another unintended consequence of HIT, CPOE, and EMR that have been deemed (by who?) to be safe. Might it just be that the “undoubtedly…many significant contributions to patient safety” by CPOE contrivances are counterbalanced by the “unintended consequences”? For accurate description, the touchy feely industry sponsored phrase “unintended consequences” should be simply renamed as defects. The administration of medical care has been disrupted and it now needs a facebook communication program (among other things) to salvage it.
From the abundant reports of CPOE device defects that caused the mistakes and unintended consequences, its safety and efficacy, that has yet to be determined, is highly suspect.
This experiment on the healthcare professionals and patients is fascinating as it unfolds.
Best regards.
Menoalittle
It’s hard to engage the IT leadership and administrative leadership in this sort of risk taking at a large, community teaching hospital. Resources are always strained and the risk of this kind of “open” record is not likely one they would seize; these organizations may not want to lead, they want to use technology to maintain their secure positions where they are. Even in the RHIO models there’s great concern about control. Imagine the response to other discussions about engaging patient-consumers, such as the cautions about Health 2.0 published at http://is.gd/2FNR or http://bit.ly/2jefxA.
From another perspective, as an ED workgroup member at CCHIT, there’s a continuing dance of push and lead among professionals and vendors with a focus on getting standards out the door. It’s very hard to persuade these ‘leaders’ to learn about social media–they don’t know what they don’t know.
Certainly social networking has great potential to improve communication, with even greater potential for chronic illness, in which there are even more transitions than in a single hospitalization.
For more on the development and functionality of UCSF’s Synopsis application, please see the related article in the Journal of Hospital Medicine (full disclosure: I am one of the authors). Here’s the link.
Interesting read. I found your website through AHRQ when I was looking for patient safety information. I’ve worked on 14M a few years ago in 2005 and I was frustrated with the delay in order entries such as faxing down orders to pharmacy and waiting. Sometimes the secretary is bombarded already and the RN needs to stand there to “wait” for the fax to go through in fear of the fax being jammed.
The re-writing of the MAR was also very cumbersome and very prone to illegibility due to penmanship errors. I am sure you know of all these details but as a nurse, it really pulls time away from bedside care.
Thanks for reading.
Christine
I am a new reader to your blog, and first want to say bravo. Very well done, very professional, well laid out. Possibly the best I’ve seen.
Onto my comment. You are right on the money. These clumsy things we use in hospitals, masquerading as EMRs, are simply mediocre paper records re-hashed into computer form. Despite the fact they do indeed provide great value (when was the last time you had to actually go to the micro lab and physically get the information that you needed. Or maybe you don’t even need to go to the radiology “file room” any more), they certainly have a long way to go to meet the actual needs of facilitating improved communication and collaboration amongst providers of care.
But Bob, do you really believe these companies are developing what we want? They barely can give us what we need. I have to say that I have personal experience with two of the largest in the acute care industry, and they have no plans on the drawing board that even comes close to meeting what you are talking about. I (along with others) have envisioned the EMR in the hospital with the ability to update providers with immediate and automated notifications of when others have seen the patient, what their assessments and plans are, when tests have been completed, etc. Unfortunately, I have never been able to successfully engage anyone in such “fancy” notions. I have seen the progress in the industry continue to show improvement. Its unfortunate that it is painfully slow. Time money and rescources.. there’s never enough.
Thank you,
Joe Sucher
Anytime you get spontaneous spread of HIT applications, you know you have something.
Not sure how Synopsis works in regard to mobile devices. One driver for Facebook applications is mobile posting and integration. It is critical to get updates to travel with and alert the provider that something has changed. I would rather review well presented updates (and acknowledge) vs. intrusive paging.
I have observed many physicians and nurses already have a “poor mans” version of this by using text messaging on pagers. There is questionable IT security and HIPAA compliance with this approach, but providers vote with their feet.
The caveat to all of this… this frees us for vital discussion on the phone and face to face. The benefit of the electronic tools is to take out the mundane, asynchronous communication. Unfortunately, I doubt most medical schools are spending much time emphasizing the value of effective communication for patient safety.
My partner and I head two hospitalists groups in the Boston area, one acute care, the other a rehab hospital. For years our handoff communications went through paper mail or fax. We were very diligent about communication. Even so, specialist from acute care settings and primary care physicians in the community complained that our group was like a black box – that they were not getting good communication about the care we were providing. The hospital even setup a physician portal so that any on-staff doctor could log in remotely and access their patient’s information. But this “pull” model never caught on, as most doctors expect data to be “pushed” out to them.
One of our new physicians suggested we look at Concentrica, which is an online network for secure clinical communication. This is free to physicians to communicate with each other. The national directory of physicians meant that we could quickly send to any physician, without having to know their fax or email. Like an online email system, recipients can reply and forward messages, so now we could get immediate feedback from colleagues in other locations, and in important cases, have a real dialog about patient care. The “Group Discussions” feature allows the specialist in town, the hospitalist, and the PCP to all join in an online dialog about one patient. The application works well on our smartphones.
When our group wanted to send documents on our behalf, we upgraded to the subscription version, which cost less than paying someone in our office to fax the documents. There is an audit trail so we can see who received their messages. One feature we really liked was that if the message was not accessed online it was faxed, so we knew our clinical work was getting there.
For our group it made it easy to communicate with other physicians, to get our documents out, gave a way for others to respond, and was cost effective.
Arthur Williams, MD
Wonderful article! (Still relevant 6 years after it was originally posted.)
Of course I agree wholeheartedly that nurses are the eyes, ears, and often even surrogate hands for physicians. As a surgical resident, online collaboration and real time communication with nursing sounds like a dream (despite being completely within the realm of existing technology.)
What if, for example, rather than page the night float covering tens of patients at 3 am to switch the patient’s medications to PO, the nurse could post a quick note for the day team? Or what if, when I lay my head to the pillow and can’t sleep, worrying about a patient, I am able to quick ask for an update from the respective nurse and s/he can answer me at their convenience when they return from the patient’s room? Improved peace of mind and improved patient care.
Thanks for posting. I look forward to the practice of medicine catching up to existing technology.
Heather Logghe, MD
@LoggheMD