A really good study/article pointing out the obvious that more isn’t always better…
Jon
2 Likes
Thanks for sharing this paper Jon. It’s an interesting piece of research that analyses a huge number of clinical notes to reveal the amount of duplicated text. Apparently the English version of Wikipedia contains some 3.9 billion words; the University of Pennsylvania Health System has apparently amassed some 8x this number of words in clinical notes, around half of which are duplicated text. Not surprisingly, ‘cut and paste’ is considered a likely culprit!
The article describes a “note paradigm” that has re-created paper-based functionality in electronic systems and “contributed to two major documentation hazards”, “information overload” and “information scatter”.
To me, this paper highlights the importance of moving from a clinical notes approach towards a data-driven design that supports multi-disciplinary clinical workflow.
@NathanK - not sure that ‘Emerging Tech’ is the best place for this post. I think it would be of interest to a wider audience.
1 Like
It is such a real and tangible problem. And deeply rooted in the paper origins (paradigm) of clinical notes. In addition, the findings of this study reflect the common tendency to cut & paste in modern EPR usage.
Personally, I believe that we should have a single wiki page for all patients in NZ with several sections. This can be updated / changed as their situation changes, and be freely accessible to the patient and whoever they choose to share it with. There is a medications section (of course) which is expected to reflect ‘what the patient actually takes’. When acutely unwell, you simply have an “Actute Illness” section up the top.
Let me know if you’d like to have a crack implementing that! The technology already exists, but the cultural change required is very much a big mountain range to climb. Thinking Himalayas rather than Southern Alps!!! But it is just the sort of clever innovation that little old NZ could wow the big boys with and change the clinical world forever…
Thanks for bringing that to my attention Chris.
Actually, Emerging Tech isn’t a bad spot; it is viewable by the entire Forum. You could make a case for it to be in Academic, but I think for now we should just tag it scientific-papers and leave it here.
1 Like
Hi Nathan,
So……interestingly - this is the approach Christchurch clinicians have started to take using the toolkit available to them.
Departments can utilise a living document at the top of the patient’s documentation record which can be updated by anyone in the care team and snapshotted out as deemed appropriate. This has been adopted across the nursing clinical team with their care plans and is gaining uptake as a new mode of documentation for medical (doctor) teams.
There is also bi-directional (optional) re-use of clinical data between workflows so that important and pertinent clinical data can be electively carried over between static documentation (i.e. ward round notes) and/or into and out of the more dynamic clinical summary/care plan “living documents”
2 Likes
The Anaesthetic Preoperative Assessment System at Southern (which I co-designed / developed with @lance.elder) takes this approach - partially.
An assessment is started by a triage nurse, subsequently updated by a clinic nurse, and then taken over and completed by either a doctor or a clinic nurse.
For ongoing care and representations for a new operation, the old note is used as the base for the new one.
Unfortunately it is limited to the preoperative context, but the idea is there!
That looks like a really positive step in the right direction! I’d love to see it in action. Would you (or someone else such as @sax) be willing to demonstrate this to me or anyone else who is interested?
I agree amazing study
-
American notes may not translate here. They are significantly longer than notes in other countries
-
In trying to build streamline notes we have observed behaviours that hard to break and i think for following reasons
-
People put notes together to help formulate their reasoning and thinking. Therefore they repeat alot of stuff in record. It helps them come to a conclusion.
-
think they also do this so when u read a note retrospectively you can piece together why they made their decision. It gives context and is impt for times when retrospective review useful
-
We tried to streamline by recording actions. E.g ability to note an action on ward round like reviewed obs and they were normal. Simple check box. . However people still liked to write them out. Same for bloods. When transcripting a ward round people often dont have time to curate they simply transcribe. Again also gives context
-
To solve this issue we need to truly understand what the underlying purposes are of the record as it has so many functions for reader and writer. My bet is a complex system and so we cant jump to soln but need to probe sense and amplify what works.
-
At cdhb we on the journey but the really hard ones are short stays as curating a master summary takes time and people dont see value unless it cascades forward.
-
Maybe the problem is our methods of recording is based on the written language. Still paper behind glass. Perhaps if was verbal or video that could be accessed easily and accurately. Then we have paradigm shift. Also understanding what is likely to be needed ongoing and extracting that with another layer behind for the detail that hardly ever required.
-
Looking forward to increasing use FHIR and interoperable data sets to see how we can change behaviour.
-
Big opportunity for same research to be done at canterbury given the notes generated by cortex. We make 45000 notes a week. How does problem look in nz. Especially on a multi-d platform that offers asynchronous chats to solve problems and answer questions quickly
2 Likes
Hi @derek.b
It’s really up to the organisation/department to decide how they want to use the living document functionality. There are some use cases where the living document could be for a single encounter and snapshots taken at appropriate times.
Others may want the living document to persist across admissions (which also possible)
Most departments are still creating daily notes depending on context and speciality but the ability to electively reuse clinical data from previous or other workflows provides the toolkit to do things more elegantly than just a straight cut and paste.
@sax has given more context above, but from our POV the key is that this transition is, as stated, “a journey” with lots of twists and turns still to come. FHIR, Hira and new regional integration are going to be huge drivers of ongoing change in how the digital clinical documentation and care coordination space continues to evolve.
2 Likes
Our team is working hard to understand what ‘high quality’ clinical documentation actually looks like, create an improved documentation culture and drive education through our University education. Whilst we await an EHR we have a great opportunity to hardwire good practice, as opposed to taking our bad habits with us #copyandpaste #ClinicalDocumnetationImprovement cdi.
Years ago our documentation was for limited purpose and readers, now it’s read my thousands and suits hundreds of purposes. No wonder it’s got complicated!
3 Likes
From my perspective the systems we are using are still in the dark ages of assembling information. Whilst there are some fields in an EMR which are very easy to replicate information, (as opposed to duplicate) into different areas such as name, DOB etc there are many other fields of information which are context sensitive and that’s before we even get to the challenge of free text. Taking the time to carefully work out which fields of information to pull together in a system, depending on the context is extremely expensive. This is even more so in the clinical environment where we need to be 100 percent sure we are not altering the meaning of the information which could result in patient harm.
Perhaps advances in AI, machine learning and NLP will mean EMRs can address this challenge at the macro level, instead of humans trying to manually work out which fields of information can be stitched together to reduce duplication.
1 Like
There are some tremendous observations and insights in this chain. As Saxon points out, “notes help to formulate reasoning and thinking” - I am doing this now and my guess is that if I had to use tick boxes, or even a SNOMED-like terminology, then the result would not accurately reflect even these simple thoughts. “Paper behind glass”, carving on stone tablet, or writing on the cave wall is, of course, how our written communication has evolved (although there is sure to be someone out there who will tell me that ancient Egyptians used check-boxes!). Free-text will be around for a long time I suspect and structure terminology, as one of the main drivers of interoperability, is not going anywhere even though its adoption over the years has generally been less than enthusiastic. A paradigm shift will surely be required to overcome these ingrained behaviours.
1 Like
Just re-reading this post. I’d love to see how this works from my/CDI perspective too. I wonder if CDHBs clinical coding team believe the documentation is better too. Then the other interesting piece would be whether there has been a reduction in incidents of patient harm, medication errors, and improvements in quality indicators. Is there an improvement in DRG complexity splits since it’s been implemented?
Hi, I haven’t noticed any comments here from a GP perspective. Duplication and contextualisation notes is arguably even more of a problem in general practice. GP software is generally far behind the hospital software. Is anyone working in this space?