Next in our series on my experience with OpenNotes, a project sponsored by the Robert Wood Johnson Foundation’s Pioneer Portfolio.
This item has nothing to do with OpenNotes itself – it’s what I’m seeing now that I’ve started accessing my doctor’s notes. In short, I see the clinical impact of not viewing my record as a shared working document.
Here’s the story.
In OpenNotes, patient participants can see the visit notes their primary physicians entered. Note – primary, not specialists. I imagine they needed to keep the study design simple.
So, here I am in the study, going through life. Five weeks ago I wrote my first realization: after the visit I’d forgotten something, so I logged in. There it was:
IMPRESSION AND PLAN: …
3. Dermatologic. I think this is actinic keratosis and needs to be removed. Refer to dermatology.
I’d reported a crusty lesion on my forehead at the hairline. I’d had the same kind of lesion last year, and the dermatologist had frozen it off. (I have a history of skin cancer.) This time, the need to handle it had slipped my (ever slippery) mind.
Several weeks had already gone by so I emailed Dr. Sands (in the PatientSite secure portal) asking if I should insist on a fast appointment. He said no.
Notice that all this happened without any phone calls, because I could do it online. And that means it all happened sooner (act in the moment, no phone tag). And it captured the action in the moment when I thought of it.
[These are all well known advantages of being online – in other industries, but still debated in healthcare, sigh. How can an industry founded in science ignore evidence from other industries?]
Last Wednesday I saw the specialist. On my way out I found myself expecting that I’d go home and update the notes in PatientSite. Then I realized I can’t – because at present the visit notes are read-only.
More to the point, today the visit notes are an unstructured blob of text. I realized:
- In my record we should have an open issue, under dermatology (or “skin,” for users who want simple vocabulary): “Have dermatologist check out actinic keratosis.”
- This would essentially be just like a customer service issue-tracking system, or a software developer’s bug-tracking system, with data like this:
- Reported by: Sands, D.Z.
- Severity: unknown
- Follow-up plan: see derm
- Status (open)
- On that item, the specialist (or I) would add
- Seen by Dr. Digby, Chestnut Hill, 8/4. Lesion is gone.
- Status: resolved.
- As with customer service systems, I could filter my view various ways: Open Issues, Dermatology History, date ranges, etc.
- All the software to do this has existed for many years, folks.
- If this doesn’t seem important for you yourself, think about it as empowering you to take care of your child, or an aging parent. All we’re asking is that we let our doctors help us help ourselves. With fewer phone calls.
Note: there are two big deals here –
- My ability to add notes, even if another doc isn’t in the system.
- I’m still constantly annoyed that doctors at my hospital aren’t required to be on PatientSite, because they’re not employees, they’re independent businesses, and that business relationship trumps better care. In essence although it looks like an office of my hospital, it’s really just a “doctor mall” with shared office services.
- So I can’t secure-email them and I can’t make appointments online – that still requires phone calls when the office is open. Ugh. And sometimes it requires dealing with an impatient clerk. Double-ugh.
- Issue tracking. At present the visit notes might just as well be a Word file: one big blob of text.
- No, actually, a Word file can include indexing, headlines, pictures, boldface for emphasis, table of contents… today’s system is more primitive. (How do doctors function with tools like this??)
- So today it’s just like a plain text file (Notepad or TextEdit).
In short, the medical record should become a joint working document among my providers (all of them) and me (a collective noun, including my circle of supporters).
I welcome comments from people (patients, providers, vendors) whose systems do have these features. Vendors, just keep it factual, non-commercial, please – link to fact sheets or demos on your website.
(If this sounds like patients climbing into the driver’s seat, well yeah.)
Dave, I’m pretty sure participatory medicine does sound precisely like patients wanting to climb into the driver’s seat. Which is rather ironic when you consider that metaphor in any detail.
We’re the cars (no, not the band). We can’t crawl into the driver’s seat.
I can see why the RWJF didn’t foresee a need for the patients to interact, though. I can’t interact with my 20th Century, paper records, either. I’m not saying I agree with that, but I do understand overlooking it.
I agree with both of your suggestions, but I’m another patient. Highlighting is so easy; there’s really no excuse for dumping the whole thing into an ASCII file. They could even, if they wanted, code the notes readers to do some of the highlighting (rather the way code is handled in gedit or in Notepad++). That way, terms brought up by the doctor could be marked in one color for mild (AK might come up in green) and another for more severe problems (and melanoma could be red). Italics could be used to mark terms as chronic (rosacea) and highly contagious problems could be bold-printed (Ebola virus).
Even without highlighting, categorization and organization should be possible. You could, for example, convince your doctor to list open issues with leading hashmarks
and list closed issues with leading asterices
Who says RWJF foresaw (or didn’t) a need to interact??
Maybe there’s something unclear – RWJF didn’t provide any software – this is being done entirely within each system’s existing medical records system. The only software change is that I can now view things that were in the past only visible to doctors.
Methinks if they’d installed new software, it would GROSSLY confound any conclusions: would any effects be due to the visibility, or to the system providing it?
I’m intrigued with the idea of an evolving convention similar to hashtags: ## = open, ** = closed.
Otoh, I shudder at the idea of explaining that to people who don’t comprehend Twitter at all. And my impression is that anything we (all) develop must be usable by all, not just those of us who understand what a convention is.
Nonetheless, I like the thinking…
Twitter? Oh, hashtags. Well, I’m a relative newcomer to twitter. I was actually thinking like a tech writer and code geek. Oddball ASCII characters make excellent bullets. Different kinds of bullets make a good replacement for different code formats.
I agree that it would be nice if all notes were online. Perhaps this will come to pass in our lifetimes.
Your other point, providing issue tracking through the electronic medical record, is a great idea but it’s far from new. This feature is built into some electronic medical records and in others it is a function of the problem list. The issue in either case is that physicians often don’t update these lists.
When one looks at the data in online (or even paper-based) medical records–something you’ve written quite a bit about–one finds that the information is sometimes inaccurate or outdated. One major reason for this is physicians’ unwillingness to update information in the record. This applies to most parts of the record that aren’t automatically updated (medication lists, problem lists, flow sheet entries, etc.). In the case of medication lists, medications often are added (during the process of care), but old ones often don’t get deleted. Problem lists are even more problematic, because problems are often not added or removed from the patient’s problem list (an exception may be when the addition of problems helps the physician submit a bill). Likewise, flow sheet entries that are imported directly from a laboratory or the practice assistant’s vital sign entries get updated, but manual updates don’t occur.
The bottoms line is that it’s easy for a developer to add fields to a database, but it is a wasted effort unless users enter data into them (and prune old data). Unfortunately, it’s nobody’s job to do that today.
Ideally, this would be a responsibility shared between physician and patient, but this may not be practical.
Perhaps issues lists will be utilized by physicians who are financially held accountable for providing top quality care for their patients, rather than being encouraged to provide high quantity care.
1. Is there no hope?
2. I wonder if a new population will arise, perhaps as the fuddies age out, in which the clinicians and patients do indeed collaborate on the visit notes.
I recently heard “culture trumps strategy every time,” and bet we could say “culture trumps mandates” just as accurately.
Hey, you integrated systems – Geisinger, Mayo, KP, GroupHealth – what is reality? Is anyone doing (or eager to do) records as an active collaboration?
It’s not a matter of hope. Behaviorism applies here–people do what they are rewarded to do. When the Joint Commission started requiring hospitals and hospital-based practices to perform medication list reconciliation at every visit, it motivated many people to change their behaviors, and it is likely that many patients have more accurate medication lists.
We could get to a collaborative part of the record with proper incentives (perhaps notes), but tickler systems for tracking issues are less likely to work this way, for two reasons:
1. Since these systems remind physicians about specific activities, they really need to manage these themselves to maintain accuracy; and
2. Most patients will not want to maintain these lists either.
Ironic, isn’t it?
I’m clear that most patients won’t want to maintain these notes too. I didn’t, until I had my “NOW I care” moment.
But every time I meet someone involved in an intense care, like the “mama lion” cases we’ve heard recently, or my mom & sister caring for Dad, I see cases where someone DOES want to, and is not yet supported by it. I want us to design to enable great outcomes, not be constrained by the slacker median.
And that’s especially important when you consider how often someone like Regina Holliday would have LOVED to be on top of to-do’s, if she could. Today’s unstructured, unhelpful records emasculate any such effort – either by the clinicians or the patient/family team.
I am sure you have heard this before but at Group Health when we went live with Epic our members (it is a co-op of 580,000 members) were able to write to their medical records even before the docs could (they were read only) via an email that went into the chart. http://www.ghc.org/cddemo/member/MyGHTour.jhtml
There is also a much more interesting model called the “shared care plan” that is being used up in Bellingham Washington. It was co-developed with patients (funded by RWJF grant) a few years ago. https://www.sharedcareplan.org/OtherPages/Phms.aspx It combined an EHR with case management and actually prevented so many hospitalizations that there was push back from local specialists who “lost income”
I am fairly sure that the UW is in the middle of rolling Epic out to their ambulatory clinics so it should support problem lists although Dr Sands identified some common challenges. It is all about the workflows versus the technology.
So, that’s why I asked you (plural) to show up: there’s an existence proof that shared records exist.
But is it structured, by issue, or just dueling journal entries?
btw, re “it’s all about the workflow” – are you talking about configuring the workflow so the doc MUST take the required steps? I’d think you’re instead talking about cultural (or other) resistance to the effort of updating the record for the benefit of the NEXT user (patient or doc).
I heard a term for that recently – wish I could remember it. Something about an altruistic motivation because the payback was deferred to a future user. Sounds like a vital insight.
BUT, if medicine transforms from being “all about the star du jour” to being a team sport, then altruism shifts to focus on a team victory.
“Yes” to all, or put another way, what doctor doesn’t want to be on the same page with their patient? I agree with Danny’s points above, and the reality is that RWJ through the OpenNotes project is doing now what was asked of the medical profession 37 years ago. Imagine all the innovation we would have seen had we taken that step then!
So glass half-full, glass half-empty (I prefer 75% full), is that we are now seeing the impact of sharing what the doctor writes. Now, who will innovate to make this about what the doctor and patient write together?
Agree – who will innovate to enable doctors & patients writing together??
EPIC, Cerner, PracticeFusion, anyone – who’s working on this? We’ll sure jump in the pool with you and help make it better!
(You WILL involve patients from the ground up, right? Would not be smart to *guess* what patients want, when they/we are happy to work it out with you, on the fly!)
Great idea! It would be wonderful to set up a simple tracking system, that the patient and his or her health professionals can view and update (and presumably correct when necessary).
In fairness to Tom Delbanco and his OpenNotes project, though, this suggestion probably relates more to the architecture of electronic health records than to the narrow purpose and design of this limited, but still interesting, project.
Terrific discussion. Thanks Dave, for kicking it off. One of the appealing qualities of the OpenNotes approach is that it works pretty well within existing systems and practices — i.e. notes are documented electronically and the systems have provisions for selectively sharing portions of the record. The ideas being floated here do run up against the limitations of contemporary EHRs. So what are the solutions? I keep going back to designing EHRs as platforms with separable substitutable apps (ref Mandl & Kohane). But what do others think?
Thanks for the time machine: way back in 3/09 (early in MY health IT life) Mandl & Kohane (NEJM said what the rest of the software world seems to have discovered twenty years ago: a modular, plug-in oriented architecture lets innovation run rampant, and eliminates the need for (pointless) centralized planning.
A key reason Adobe Photoshop totally dominates the world market for professional photo editing is that they didn’t try to figure out everything everyone would WANT to do with it. And they made it trivial to install a plug-in: just drop it in the Plugins folder. Result: anyone who thought they saw a need was free to invest their own time and money trying it, and if others agreed, it would float.
Newspapers could invent their own custom plugin if they wanted, even if nobody else ever used it. Commercial developers could invent stuff if they believed there was a market. And people could add stuff just ’cause they felt like it.
And here’s a tip for business managers who fear that they’ll lose a captive market if they open up: Adobe’s kinda doing okay, last I checked. Even after 20 years of this.
I haven’t scrutinized that NEJM piece, much less anything more about Indivo, but I can’t imagine why any developer who wants to cash in on the crashing need for better EHR wouldn’t use an approach that enables open innovation. Why on earth would you take on the burden of figuring it all out (years into the future), when you can say “Hey geeks, come make me better!”?
It’s like offering free harnesses at the front of your wagon.
Practice Fusion is taking this approach now — see their challenge as part of the Health 2.0 Developer Challenge: http://health2challenge.org/blog/category/challenges/. And they seem to be drawing interest from developers — 16 teams have signed up so far.
there is a web site, my first obgyn used, I log on to my accout to get lab results & then they stay there. they even had him dictate the message so I can hear someone tell me I’m ok. (I forget the site on that on)
ihealthrecord.com by medem you can contact your doctor & they can contact you, remind you of appts.
reliefinsite.com is good for me keeping my symptoms &pain diary.