Just wondering if anyone has developed a solution to provide doctors handover functionality?
I’ve had a request from our clinical colleagues to provide the functionality to deliver capability for our doctors to be able to create handover notes and the receiving doctors view and confirm these actions.
Happy to develop something with the team and share but was just checking has anyone been down this path before? Could save me a bit of time and effort.
Cheers,. Ben
Hi Ben
I’d be happy to have have a chat to discuss what we could do or point you towards someone who maybe able to help
Best Wishes
Chris
Handover solutions and ‘patient list’ type requirements were a regular feature of NHS Hack Days - usually at least one of the pitches at NHSHD would be asking for something to solve this problem, which - as per your request - still isn’t solved!
I’m not aware of any solution that is good enough to actually recommend as a clinician. The problems as always are that solutions when implemented:
I would just implore you to open source whatever you produce @ben.sewell82 and try to make the solution generalisable to other trusts (this can be done with how the software is designed - don’t hard-code anything specific to your trust, use adapters which plug into your integration points allowing other trusts to write their own adapters for their integrations… etc) Happy to help with the open-sourcing bit, winning hearts and minds etc as well as the technical.
What is involved in the handover?
I’ve only done army watch keeper handovers where we would describe what we’ve done, what needs to be done, situation, etc…
This would be centred around a log or journal.
On health I’ve heard about ward rounds and tasks that need doing.
More interested in knowing what we are trying solve as I suspect I already know ways of solving without reinventing the wheel.
Medical handover takes various forms depending on speciality and location, acuteness of patients/illness severity, etc. It follows a similar pattern to the army watch handover though - what you’ve done, what needs to be done. I’ll try to bash out a few bullet points here, and I’m hoping to be able to obtain some real paper handover sheets to help illustrate.
Usually handover would be 12-hourly as one shift hand over to the next shift. In some areas where the shifts are shorter (eg ED) then handover would be 8-hourly. Handover can last anything from just a few minutes (eg ED when it’s inexplicably super quiet) to a full hour (eg on an ICU), depending on the amount that needs to be handed over.
When I started medicine this would be a handwritten list, each doctor would make and maintain their own copy. If you were on all weekend for example, you could just update yesterday’s list in some places since the patients may well be the same.
Later on (say 2005 onward) the default started to be a Word document (a table) or an Excel spreadsheet, where each row would be a patient, and the columns would be their details. Each team member would generally still print out a hard copy, which they would annotate and add to as they worked clinically (much of our work is away from the computer) then they would update the Word version
For most hospital specialities we would go through a list of the patients that service is caring for. This ‘patient list’ feature is the thing that a lot of EPRs just don’t support well. We need to be able to add and remove from this list easily.
It’s also important that we can easily make notes and changes across the whole of the list without having to go into each patient’s actual EPR as this is time consuming.
In some places the handover documents are an organisational safety/clinical governance record that a handover meeting took place and the sheets may actually be signed, dated, and kept. In other places the handover sheets are shredded once the shift is over. (And of course a small proportion accidentally taken home, left on buses, found in the bottom of work bags after moving on to the next rotation, turned to mulch in the washing machine…)
The ‘digital master’ can be a blank document which is simply printed and populated with handwriting. However, in some care settings it can be a digital document which is kept updated with actual patient information and a new version of the file is saved to the PC. This practice is discouraged because of the IG risks posed.
(ie a paper handover sheet, with or without some form of digital ‘master’ that is updated later)
(ie a paper handover sheet, with or without some form of digital ‘master’ that is updated later)
DON’T BREAK USERLAND As you can tell from all of the above, it is a real mess at the moment, from many perspectives. But whatever solution replaces these ad-hoc arrangements MUST NOT BREAK CLINICAL WORKFLOWS. Because, bad as these handover forms may be, they are working, and we clinicians will have to continue to deliver clinical care with whatever replaces them. ‘Don’t Break Userland’ quote source
Writing all this I’m actually just quite amazed that despite all the money that’s been poured into NHS tech in the past 20 years, the current state of handovers - a fundamental part of safe medicine - is so poorly handled as to show no evidence that anyone has ever tried to solve it.
Psychiatry
Critical Care (ICU)
If anyone’s interested, I could go on to try to delineate some of the general principles that might form part of a digital solution to handovers…
Thanks for the replies. Really good responses. I’ll have a catch-up with the team and clinical view and post back with an update, I’ve got a small dev team propping up a lot of systems so another decision is priority and phasing of work. Agree, the solution should be light as possible and loosely coupled to support adapting and porting where required.
I think this is focused around Task. I’m looking at this for a hack event next and some other initiatives.
For the hack I’m going to be focusing this on the patients list of Tasks (e.g. order a repeat medication, diary entries from EMIS, Mediication Reviews etc).
This sounds similar with the focus on a ward or something similar.
At a basic level it sounds like we have a patient id, optional code for the Task (maybe SNOMED) or just a simple text string, maybe some notes/annotation.
These Tasks have status like the ones in FHIR Task so ready, completed, in-progress.
I’ll if I can mock up a simple example, I only have more complex ones at the moment designed to move tasks between doctors.
I’ve been finding this ‘administration of care’ a bit of a weak area. I think it’s a big gap in what we do.
We have tons of knowledge on document patients records or patient administration but workflow such as this and the ones I’m seeing all seem pretty weak.
Which in a way is a good thing as you don’t have to battle with all the existing processes, viewpoints or standards. It’s greenfield
It’s interesting that “Tasks” seems to be as much part of the care before handover as at handover e.g. “obs every 30 minutes”.
I guess that is not just recorded at handover.
But maybe if that is the short term situation, which might not be relevant at the end of shift, then the clinician doesn’t record it.
It only becomes an issue when they have to hand over.
But if it was recorded instantly - as hard is that is - then handover would not be so specific a job.
So maybe this isn’t all about handover.
‘Handover’ means different things to different disciplines and as Marcus has demonstrated is an evolved process, often specific to both medical specialty and hospital site.
To clinicians, it’s running through the Git commit history of the patients to ensure an incoming team is capable of making the mission-critical decisions over the coming shift.
To clinical governance teams, it’s a crucial safety feature of good quality care (and as such the process should leave a patient-specific auditable footprint).
To information governance teams, it’s one of those pinch points where PID coalesces.
I’m an intensive care consultant, and my paper record of the handover becomes my scratchpad for ideas, prioritisation tool, aide memoire for emergency decisions (“this patient is crashing, shall we put him on the X/Y/Z?”), to do list, jotter for phone numbers and memos and a lot more besides… until it is dutifully shredded at the end of a shift when any useful information from it is either in the EPR or the brain of the incoming team.
How much of this can be distilled down into something reproducible across disparate teams and sites I would be fascinated to see, though if this were an easy challenge you’d have expected it to have been done and dusted by now.
DON’T BREAK USERLAND As you can tell from all of the above, it is a real mess at the moment, from many perspectives. But whatever solution replaces these ad-hoc arrangements MUST NOT BREAK CLINICAL WORKFLOWS. Because, bad as these handover forms may be, they are working , and we clinicians will have to continue to deliver clinical care with whatever replaces them.
Wholeheartedly agree with this. Having spent a reasonable amount of time making bad forms better (audit trail, incremental backups etc) there’s nothing more than I’d like to see than something that negates the (modest) duplication of effort, the IG concerns… but not at the expense of reducing the quality of clinical care.
Good luck, Ben! 
One of my thoughts at the moment is we should be designing around workflow and NOT building forms, datasets, etc.
When we do that, it all becomes about the forms or datasets. It should be people+process+technology and not all about a friggin spreadsheet (being sent around via JSON and REST).
