- ‘Wannabe’, Spice Girls
ED Clinical User Interface v Clinical Information System From Terry Lopes, Chief Architect NSW Health IT – modified by iCIMS and reproduced with permission
Clutter = cognitive load Clicks = time
On the opening page, the screen headed “All patients medical officer”, there are 63 spaces to click. That is without counting the actual patients that can be clicked on. There are a row of 4s and a row of 8s that do not seem to have a function. It appears that they represent an advertisement for NEAT and EAP. They can be confused with triage categories and time in ED.
On progress notes and other documents, there are excessive headers and footers. They are not in a background colour or with some signal to say “this is not the part you would usually read”. There is almost no use of colour, or background colour, or font to provide clues like an individual’s handwriting did in the old written records. So it is not possible to scan for where that doctor has made another entry. Or to see by documentation style where a doctor has written versus where a nurse has recorded a routine check.
Function is overall poor. Rather than create a narrative medical record, that can be read as a whole, it is more akin to creating a filing cabinet, with myriad individual pages slipped into it. Every nurse visit is a progress note and so is the one visit from the surgical registrar. The closest functional record is the ED assessment, but then there is no scope for later reviews and records of developments by other treating doctors.
To print a letter it is not possible to just hit the print button and get the letterhead which identifies the hospital and provides Gps with phone numbers and so on. To do this you would have to log out of the letter, and go to a select a chart option and re-open it, and do more steps… such that almost all letters go without letterhead.
The clicks to sign an Xray request also make you first highlight the question you are answering, then put in the answer. The answer is often the same as is has been every time, such as my contact phone number, ut it has no capacity to remember this so its asks always. But it asks with multiple clicks rather than just opening up the space for the answer.
It is not possible to sign at the end of documenting. Instead, at the start of documenting you must log off from the previous user, then lo on as yourself, then document, then sign. Staff are still caught out by this at least a dozen times a day. It leads to dysfunction. Sometimes they copy and paste their work into a word document, then log off, log on and paste it into the record, then sign it. But signing other people’s work is rife, and there is even password sharing, and frequently they seem to give up and abandon documenting at all.
There is no separation of visits. Just a row of dates, and sometimes even visits to different hospitals on one day are inseparable. To copy and paste, you can only Ctrl+C and Ctrl+V and even then only form certain parts of the screen. Often the work has to be re- done. Some staff only ever write in a discharge summary to avoid this problem, but that creates other confusion, and lots of strange emails to GPs.
When it jams, it does not seem to be possible to ctrl+alt+delete. That computer gets left alone until a staff member who can fix it comes along. Failure to have a narrative record created leads to a work around where later staff go into earlier records and amend them continuously. This leads to a better narrative record, but tracking of who wrote what is lost.
There is no short term project function that is customisable. That is, for example, I would like to have the CAM ICU delirium screen available electronically and then have a report that did an audit of how often it was used on, say, pateints over 85. I would have to do this paper and forms and cardboard boxes. There does not seem to be a search function, so I cant find things that I suspect are there, like the stroke guideline. Every patient is recorded as being treated by the director of the ED.
There are some works that seem to be related to American practice, like “attending” Signing the discharge leads to an email to the GP, but not signing it leads to other people having trouble accessing the work. So GPs get puzzling half done letters whenever patients are handed over. To open, say, an ED assessment, you have to click in two places to say that’s what it is. For the dictation software, a single voice command can do lots of clicks to take you to where you need to be. But the same is not available on the keyboard.
There is no language screening. That is, it could pick up words like confusion, disorientation, fluctuating level of consciousness, and then offer an option to record a diagnosis of delirium and offer advice and collect data. ECGs are imaged later, so they are not available in the first few days after discharge. No diagrams.
Some other electronic devices don’t communicate so nurses must transcribe, eg from urinalysis. You have to click through annoying pop-up adverts for spinal wards every time you log on. Results don’t follow the patient to wards. So results checking is a failure. Radiology results are done on paper. The pause after log off leads everyone to open the eMR manually before the automatic screen opens, and then you get a scolding about having it up twice, until you learn to meditate.
eMRs have not delivered clinical improvement and have a problem with inflexibility NEDIMS modelling saved time and reduced cognitive load eMRs have tangible benefits but the evidence for them helping overall in EDs is limited. The Effect of Computerized Provider Order Entry Systems…Review. Georgiou, Prgomet, Paoloni. Ann Emerg • Med 2013;61:644-653 Efficiency Achievements from a user-Developed Real-Time Modifiable CIS. Bishop, Patrick, Besiso. Ann Emerg • Med 2015;65:133-142 Health Information Technology and Victory – Editorial. Wears. Ann Emerg Med 2015;65:143-144 •
NEDIMS type system can sustain QI and ad hoc audits Rapidly solves problems like sign work at the end
There seems to be enough evidence such that it would be unethical not to trial a new interface. A trial needs a whole ED, but more manageable in a small one.
Needs to look at all KPIs (wait by triage, NEAT, EAP, TOC, DNW), time to teach system, ease of obtaining reports for QI, ease of use.
If we say it is research on something that affects patient care it needs ethics approval Would need IT co-operation.
The trial would need to be funded.
Recommend
More recommend
