Hi,

I am working on a project looking at usability issues with medical devices and human error that is influenced by device design. For example, sending an email without an attachment, recording the wrong TV program, and assembling flat packed furniture wrongly are examples of human error which could potentially be improved through better design.

I am a researcher new to the medical domain and so would love to hear more about:

1) Device design interaction issues you are aware of.
2) Stories of human error and near misses that you think were influenced by device design.
3) If you consider usability, Human Factors or HCI in purchasing? And if so how?

Apologies if some of these questions appear a bit naive... I look forward to learning more about these issues from practitioners over the course of this work, and if the project generates more discussion about these issues then all the better .

The project is called CHI+MED and a link to a short film I've made for it can be found here: Man-Machine Nightmares

Thanks

Welcome to the forum, Dominic.

All that sounds very interesting. And yes, I reckon you've come to as good a place as any to secure feedback on the subject (ie, this forum).

It must be nice to have funding for a six-year project!

I've just enjoyed the video. It's nice to see at least that infusion pumps are (generally) better designed than microwave ovens (what does the Chaos button actually do, I wonder)!

I'm pleased to see that you've homed in on infusion pumps.

There was some work done by the US Air Force a few years ago which looked into the possibility of standardizing pump interfaces. Alarm tones in all medical equipment, but especially pumps and patient monitors, is another obvious area. Not to mention alarm fatigue!

Defibrillators are another area worth looking at. I would say that most modern ones have the design problems more or less sorted out. And so they should, by this time!

On a slightly different tack, I see a lot of older equipment myself, wherein it is often easy to see (once you open them up) where better design decisions could have been taken. Broken parts that could (should) have been more robust, etc. But (as my long suffering colleague has been known to remark) ... isn't hindsight a wonderful thing?

Also, there have been a number of items (of medical equipment) which have truly evolved over the years. That is, with each new version building upon the solid design of the original, and adding sensible features as the technology developed. The Ivac (Alaris, whatever) P-series syringe pumps are an example which spring readily to mind. And, conversely, many others for which time has seemingly stood still (the MS-16A ambulatory syringe driver, for example ... a subject for "research" if ever there was one)!

Dominic,
Apart from general equipment management one of my roles is medical device training, so people using equipment correctly is high on my agenda. I also have a personal interest in ergonomics in general so I am glad to see some serious work being done on this.
Have you seen the recent publication aimed towards pump manufacturers telling them what we, the end users, want and what we think is a good idea? I thnk it was the NPSA who published it.
Happy to talk more about this subject.
Contact details in my profile.
Robert

Dominic, bear in mind also (and as I have said a few time before - but you probably haven't seen it) that most medical equipment is already well designed (it has to pass through the FDA, and all the rest), and that clinical staff are not inherently stupid (far from it, in fact). But it is the conditions* under which the stuff is used that is the real "problem"!

Think "hospital design", "the hospital as a system", "design in the round" etc. Whilst, of course, avoiding consulting architects at all costs!

However, if your focus must be kept within the narrow framework of equipment per se ... then surely the keyword, as far as operator->equipment interfacing is concerned, must be:- "intuitive"!

And, lastly, you may not realise how rugged medical equipment needs to be. In short, stuff gets dropped (sometimes a lot)!

* Stressed out, chaotic, dramatic, rushed, over-worked etc., etc.

Dominic, You may want to take a look at the NPSA document 'Design for patient safety - A guide to the design of Electronic Infusion devices' Ed1 2010. It can be accessed from their website. I think this may the document that Rojo was writing about. Lots of info' in there that is relevant to what you need.

Link

See also the further links offered at the References section towards the end of this nice document. Including (for instance) this one and this.

Thanks for the tips and pointers - really useful .

We have a project meeting at the beginning of October and one of the items on the agenda is to discuss the NPSA's new Design for Patient Safety booklet. We have been looking at this already and hope to be in a position to contribute to a version two sometime in the future (maybe?) (we have a representative from the NPSA on our steering committee ). It seems like they have done a good job, both in content and clarity. Thoughts on this would be useful to feed in to the meeting if you have them. I'll have a look at the other refs.

Glad you liked the film Geoff. Here is an explanation of the 'chaos button', which was found by another friend who was intrigued when they found one on their hotel room's air conditioner! Apparently the latter uses choas swing technology . All sounds very clever but perhaps not immediately 'intuitive' to a user .

Pointers to other equipment to look at e.g. MS-16A ambulatory syringe driver and defribulators are really useful... I'll have a look at the other threads and the one on alarms.

I am definitely interested in the evolution of devices. I've heard of patients sharing knowledge of how to silence their own infusion pumps after an occlusion alarm sounds (typically from bending their arm) even though they're not allowed to use the pumps themselves. They watch the nurses and copy them which is bad. Also annoying for nurses having to manage these often trivial alarms. Apparently the Asena pumps sound, then retest the line, then stop the alarm and continue infusing if it has cleared. This uses less nurse time and means patients don't break the rules. Clever.

I'm glad the role of context was raised as investigating clinical situations is more my part of the project. Other areas include device design, cognition and stakeholders. There is always a focus on device interaction but I also look at understand the work around them. I have recently finished a small stint at an Oncology Day Care Unit and still doing the analysis. One nurse said it's annoying when a pump runs out of battery part way through a long treatment. This rarely happens as nurses are aware of this, and the interface shows the battery level. The device alarms and counts down when the battery is low so it can be changed. However, I wondered why the pump doesn't tell you it won't last if you program a 2hr infusion and it only has an hour of power left?

Our perspective definitely doesn't consider people as stupid. My film tries to convey that we all make mistakes, everyday: To Err is Human. A nurse I spoke to acknowledged this and said her mantra was "Check, check, check." People are often very clever at monitoring their own activities and errors, and finding workarounds and strategies to make less errors e.g. leaving the umbrella by the front door. I'd like to find examples like this in the medical context.

I like this example of a design glitch - not so clever : Opticlick design vulnerability

All good stuff, Dominic. Be sure to keep us updated.

I'm not convinced that we're quite ready yet for a Chaos button on infusion pumps, though.

"Check and Test" used to be one of our (unofficial) mottos in the army too. When we had nothing else to do (and sometimes even when we had) we would check our kit* - again and, er, again (to the point where it would probably appear a bit anal to an outsider).

Which Asena pump do you have in mind there, I wonder? We need to be clear about the distinction between volumetric infusion pumps and syringe pumps. In general, infusion pumps deliver fluid volumes, whilst syringe pumps deliver drugs.

Certainly some of the newer pumps (of both type) "pull back" a bit on occlusion, but whether they cancel the alarm (upon sensing a return to normal line pressure) ... well, I'm in doubt about that. Hopefully someone on here can clarify that one (otherwise I'll need to run some tests on the next Asenas I come across)!

I like the idea of "predictive battery capacity technology". Now that we have so-called "intelligent" battery packs, the sort of thing you describe should not be outside the realms of possibility, I should have thought.

And, as for copying the nurse. Many pumps have a "lock out" function, so all should be well as long as the nurse shields the key-press sequence involved. If a higher level of security is required in our New Spec ... then we shall need to think about a swipe card, chipped key fob, or some such thing. But then ... what happens when the nurse looses the key?

Lastly:- "umbrella by the front door". What about the endless argument in New Hospital Design about the so-called Doctors' Hand Basin? I have been in hospitals that have this (small) basin positioned (at great expense) near the door in every patient room. Yet some - many - of the "blighters" (yes, I'm being polite today for some reason) still didn't use them to wash their hands between visiting patients! What's the answer there - cameras? (no, due to patient privacy). Please discuss!

* I'm referring here to the "green kit", of course ... rather than medical equipment (although we certainly checked and tested that enough, as well)!

That's a brilliant document!

Anyone know where we can purchase Infusace pumps?

OK, let's get this one back into play.

How come that calculators, and numerical keypads on computer keyboards have 7-8-9 along the top row ... whilst telephones, mobile phones etc. have 1-2-3?

Which is better (more intuitive, or logical), and which should medical equipment use? Assuming, that is, that we want to standardise.

Meanwhile ... I have two "well known" infusion pumps right here in the workshop. One has the 1-2-3 layout whilst the other has (yes, you've guessed it) the 7-8-9.

Interesting question... I've just found that pages 24-27 of the NPSA Design for Patient Safety booklet recommends 1-2-3 layout to follow the wide use of mobile phones. And warn against mixing both layouts in the same context.

It also recommends considering chevrons and other analogue controls so the user monitors the screen more rather than punching in the keys as they will be more sure of their actions and less likely to check.

I've asked my colleagues but is there any more detail or science behind this?

I have always considered one of the "great leap forward"'s of medical machine interface to be the introduction of the Marquette (now GE) "trim knob" control (I may be a bit biased here as I worked for them at the time).

This simple interface control device allows the user to concentrate on the screen/menu/values/etc without the need to look at what their hand is actually doing. No various button pressing just a simple left/right movement and press to select. I think the evidence of just how good this interactive device was is just how many machines now use something similar.

Not nurse proof I'll grant you but simple and effective.

I also like the (so-called) "trim knob"* approach. But I first saw these on Dräger monitors (back in the mid-1990's). Not a lot of good on infusion pumps, though.

I also like the up and down chevrons on infusion pumps. A pair for each of the hundreds, ten, units (ones?) ... and (if need be) decimal positions.

Meanwhile, I have yet another "well known" pump in front of me now that simply has a 100, 10 and 1 button. Each press on each button increments by 100, 10 or 1 respectively, and then rolls-over to 0 again when you press again at 900, 90 or 9 (if you see what I mean). As always (and as Bruce sings in one of his songs), "you can get used to anything ... sooner or later it just becomes your life"!

* Twiddle and click? Very quick ... just as long as you can see what changes you are scrolling through, and then commiting to. Once again ... feedback!

Yes, it's called (visual, tactile?) feedback. Therefore "a good thing"!

Why?

Actually, you could be right there, Mate. Take a look at the old Terumo TE-311, for an example.

But my own take would probably have to be that the "trim knob" would be too quick and easy for insfusion pumps. Better, I reckon, to have something that enforces them (the users) to actually stop and think about what it is they are trying to do.

An example of that? Why, the "good old" MS-16A et al. Ha, ha.

Yes, quite, the MS16/26A, an example of bad user interface if I ever saw one.

However, what about the Omnifuse or that one that had the wheel thingy on the right hand end, um, err, the ....... (? sorry, it's late and my brain has slowed to a crawl).

Examples of "rotary switch interface" in infusion devices me thinks.

Ok, enough, bed time for me.............

... Terumo?

Meanwhile, here's the Omnifuse. I have never come across one of those myself. But, yes, that looks like an interesting interface.

How's that? It needs a tool in order to change the mm/hour (mm/24 hours) settings. Not a bad "design solution" considering its intended use, I would have thought. What's more, it has "stood the test of time", has it not? As usual, "less is more" (simple is best, whatever).

Well, it's in terms not of the physical user interface but rather that they are the only infusion devices I have ever seen that work in mm/hr rather than ml/hr. Confusing? I think so. A trap for the unwary? certainly.

As to the use of a "tool" to change the settings, yes, a nail file or the cap from a Bic Biro will do just fine

Bearing in mind that I didn't actually design the thing myself, I would say that it is "proven", if only by it's sheer length of service (what is is now ... forty years) and almost universal acceptance (or should that be, uptake in the absence of any serious competition*)!

It works in "mm" (rather than, I presume "ml") because that was about all the technology of the late 1960's allowed. It was designed by Pye - remember them? It also happened to the first (ambulatory syringe driver)! Not to mention cheap ... due to the use of components readily available at the time. Like micro-switches (no LED's back then)!

Confusing? To today's harried nurses, perhaps. Maybe nurses of yesteryear had a bit more time to actually think. Who knows? They didn't have degrees back then, after all.

They are not complicated at all. In fact the word "simple" comes to mind. Also reliable. And (I would assert) appropriate to the setting(s) in which they are used. And ... the adjusting screws could easily be covered if need be (by a bit of surgical tape - the nurses' favourite - for example) - not to mention the syringe cover itself.

Meanwhile, I believe I have mentioned elsewhere the nonsense of testing ancient technology like this to standards way beyond the original design spec (as I have seen happening to MS16A's in certain NHS biomed shops). And that goes for the design itself as well. That's a bit like condemning the original Morris Mini after taking a look at a modern "BMW" Mini (a bad example, perhaps - because the 1959 design was actually superior)! Oh yes, the "retrospectoscope" can come in very handy at times like this!

Sorry Dave. Like it or lump it, the MS16A/MS28 has already gone down in history as a classic design (of medical equipment). Hey, I've already got them in my museum!

* Until recently, that is.

Seems to me here Geoff, that you are making "points of argument" against points I never made in the first place.



Can't remember ever saying it was not "proven" only that the "device usability" was less than first class.



Well, there you are, we are talking about "usability and human error" after all.




I never said they were.



Never said it wasn't a classic just an example of poor usability and prone to human error (? the worlds most dangerous infusion device)

If we can agree that the word "quality" can be taken to mean "meeting the users requirements in all respects" ... then I would submit that the MS16A/MS28 is (or, rather, was) a Quality Design! High marks for usability, then.

Whether or not nurses cannot be arsed to use them *correctly, get themselves trained, ask a "senior" ... or whatever, is, I grant you, an "issue". But it is one that goes across the board (that is, may very well apply to any piece of medical kit), I would have thought.

OK ... let's move on to the next exhibit!

* Especially after all this time, the amount of controversy surrounding the device, etc., etc.

No where near, you should have seen (or investigated) the Vickers VP55 syringe driver.

The MS16/26 syringe drivers are brilliant they are simple, can be used regardless of syringe make or size, do not have pressure alarms (that only really let you know that you have a problem with the pipe work between the pump and the patient), or the myriad of additional features that are rarely if ever used.

Staff not trained on this or any other medical device should not be using that device.

Most if not all of the errors I’ve seen with these devices are human errors, you can not totally stop human error no matter what design you come up with.

I am afraid that in my experience as the design of equipment has improved (safety wise) it appears that the users become lazier or less vigilant.

Lee

What we need are more MS16-type infusion pumps ... but big 'uns! Ha, ha.

But somehow I guess that's not the answer Dominic is looking for. He has a six-year project, after all!

Have you looked at the "Design..." .pdf yet, Lee? Quite interesting, I reckon. Clearly laid out, too. A good "discussion document", in fact. So ... please discuss!

@ Lee, have never seen the Vickers VP55 (from the same people who brought you the Vimmey bomber presumably !!) so can't comment. Perhaps I should have said the the worlds most dangerous contemporary infusion device.

Do people become lazy/slack/dependent when technology is available to take out the thinking part, no argument from me. All the "safety nets" provided can remove the inherent checking that will happen with more basic systems but, frankly, that is a path that we are so far down we cannot, nor I think, should we turn back from. We are not yet at the end of that path, it's just that the bumps are still a bit "lumpy"

I am reminded by this discussion of a comment I heard from a "vehicle safety expert" some years ago when asked about air bag technology. His comment was that, even though air bags were a wonderful technology, in some ways he would much rather see a large steel spike attached to the steering wheel, pointed towards the drivers chest, that way, he said, people would be a lot more careful instead of feeling that they could/would not get hurt if things went wrong. I think he had a point (pardon the pun).

Indeed. Most old cars (British ones at least) generally had the steering column configured as a lance for that very purpose!

By the way, have the Aussies started making their own aeroplanes yet? Interested? Then take a look at the .pdf.**

Actually, there was nothing much wrong with Vickers Medical kit, in my opinion. Clean designs, rugged (that is, made of steel rather than plastic), reliable, and all the rest. Ideal, in fact for extended service throughout the Empire, and beyond.

And, I might add, still sought after for equipment* donations in the "tougher" parts of the world.

*Certain other British manufacturers deserve an honourable mention there as well. Eschmann, for instance. Even good old (but sadly long gone) Cardiac Recorders! And the famous SAM pumps by M.G. et al. Plus, of course, Blease, Penlon and the older Ohmeda stuff (BOC-Medishield, whatever). Plenty of examples of Good Design to be found amongst that lot!

**With the state the world is in now, one wonders if they could have followed that same route today.

Footnote: the Vickers Vimy was built in the Bedfordshire town of Leighton Buzzard ... not far from where I am right now!

Are the MS16A's the ones that people were getting confused - I recall someone saying that there were some incidents of mixing 1hr syringe drivers with 24hr syringe drivers... as per the warning on page 86 of the guide

We can say this confusion should never occur as people should be aware of the differences, be fully trained, have time to think and check everything they do and focus on what they're doing without interruption. However, it seems this utopia rarely exists, and in rare cases error gets through the safety checks and alarms and things go wrong. It's probably true that human error cannot be eliminated but there are things that can increase and decrease its likelihood. Human error is a term that is used broadly and sometimes people use it too readily to blame 'lazy' users when really the system and devices need to be thought about as well.
(From the perspective of our project we probably have a bias toward not blaming the user but trying to understand why the error happened (in terms of device design, cognition, and the context), often because other 'normal' people in the same situation would have done the same)

I must admit that converting between 'ml' and 'mm' sounds like a nightmare to me. I'm not sure what's involved at this stage - but my intuition is the less conversions and calculations the better. The volumetric devices that do most of the calculating for you seem friendlier. It'd be nice if the prescriptions included the exact information that is needed for the device to reduce the translations and calculations nurses have to do. I think this would reduce the likelihood of error.

There was an important point made about dumbing down work as the devices do things for you e.g. the calculations. Here we become more and more reliant on the technology... we're pretty stuck if the battery runs out on our mobile phones as we don't know the actual numbers any more, and I am sure more people get more lost when their SatNavs fail. However, this can also allow us to do more. A nurse raised the point with me that they do a lot more nowadays... when she started the max amount of infusions per patient was more like 2 or 3, in the most extreme case she had 15 pumps hooked up to a patient and was managing them on her own. Real consideration needs to be given to doing more, doing simpler, and deskilling people... you might make the system more efficient but it might be less resilient and more prone to error.

The issue of feeling safer and taking less care reminds me of people reducing visibility at roundabouts, i.e. they would plant hedges so as you approached a roundabout you were unable to see oncoming traffic, therefore you were more likely to actually stop. With better visibility people take more risk as they look ahead, anticipate and squeeze into the traffic. Ironically then the safer system is less usable, less efficient.... again careful consideration needs to be given here. It does emphasise thinking about the system holistically to make it safer.

James Reason has come up with a ' three bucket ' model which I think nurses are meant to use to be more aware of 'risk'. Basically the more proverbial in the buckets the riskier the situation and the greater need to back off, think and get help. (I'm not sure if there are plans to incorporate a fan into this model )

Yes, that's the one.

The MS16A and its famous friend the MS28 come in for a lot of flak. But, personally, I don't buy it.

These things are used in special settings (such as Palliative Care), and not - generally speaking - by every Tom, Dick and Harriet who happen to have nothing else on!

As the quote says, the real issue with those little pumps may not be the "difficulty" in recognising that they are designed to deliver a linear distance rather than volume, but the fact that they may simply be confused (that is, mm/hr versus mm/24hr versions). Which, of course, is a serious mistake to make. However, they are colour coded. How "dumbed down" does kit have to be?

What would prevent similar mistakes being made if the things were calibrated in ml?

Reason and his buckets? Interesting for some, perhaps (that is, those who like to spend a lot of time at training seminars and the like), but it sure looks like (yet another) box-ticking game to me (and designed for whingers, at that)! "Culture and power distance"? Eh?

Roundabouts? Yes, I consider them to be the most dangerous feature of the British road network. Take a look at the A1 ... much improved since many of the roundabouts have been eliminated. Surely I have mentioned before my campaign to introduce some kind of standardisation of roundabouts on British roads? Getting it all down to something like three or four (well-designed) types. Not to mention the complete removal of all so-called mini roundabouts! In that way, we could probably save many more lives than fiddling about with medical equipment (which has, on the whole, already been well designed).

Let's get this one back into play!

After all, we could come up with the specs for the "perfect designs" right here, could we not?

How about we start with:-

1) Syringe Pump
2) Volumetric Infusion Pump
3) Defibrillator

For all I know (and I believe that I may), the "perfect" examples of each could already exist!

Would anyone like to weigh in with their own candidates (or, indeed, their own design specs)?

It would be easier to redesign the human, especially the 'brain' part

Whilst there will always be equipment users from the shallow end of the gene pool, don't forget, Neil, that our nurses all have degrees now.

Quote "Neil, that our nurses all have degrees now"

Thats usually a few degrees above room temepature.....

Simple mistakes are made by people right across the board from cleaners to surgeons... We all make mistakes and so I think it's a bit of an easy option to blame the users. There are different sorts of errors that people make and 'slip errors' are even made by people regardless of their expertise.

We need to move away from simple models of blaming users because systems/tools/devices/procedures can be better designed. If there is an error/incident I think people follow this train of thought:
- was the device working as the instructions suggest?
- if not, then manufacturing fault.
- if yes, then user error: blame, discipline, train, train, train.

This is a great video, inspired by an accident, which has an alternative view. It proposes bringing a better learning and safety culture into healthcare and uses the aviation industry as a good role model:
http://video.google.com/videoplay?docid=-6738698910009425483#

As a rule, we don't apportion blame. We just pick up the pieces.

The trouble is, though ... that the users like to blame the equipment when someone has "slipped up", and then (if possible, or even by association) us.

Physical design does play a part in user errors in my opinion but I find that with increasingly sophisticated medical equipment being introduced, software design plays a bigger role in user related errors.

Syringe pump = Protected arm, mains cable locked in place, large clear display (with drug being used and mL/H or equiv), On/Off button, occlusion change button, rate change button and start/stop infusion.
It forces the nurse to do a purge (therefore eliminating air in line and backlash).
It is quick to start.
It is water/fluid proof.
It has no stupid fiddley bits that will break off or not get cleaned.
It is light.
It has an alarm that makes them do something with it.

On our side= Software that works when you have to calibrate.
Four screws max.
Cheap parts.
Modular mechanism for quick repair. (but with all components available for repair)
Test equipment at a reasonable price.


Will get on to the others after I finish this pump in front of me.

Billy

Had a play with a Niki T34 syringe driver today (the one that seems to be taking over from the MS16/26A in this part of the world). Very nice I have to say. Simple user interface, I could drive it without reference to manuals or any such nonsense . Preprogrammed to a drug protocol (hospital/clinical area specific) and works in a similar way to all other pumps (you know, mls/hr unlike the MS series).

Have not seen the programming tool but others of my team have and the reports are good.

A step forward in patient safety by the looks of things.

bcarlise
good list on the needs of a pump but I would not agree with the user being able to change the occlusion this should be set at an upper limit that is safe .If the pump alarms occlusion there is a reason and the user should not be able to alter the limit so as to just silence the alarm

The problem is the viscosity of different drugs used. With the gloppy stuff(drugs knowhow) they use on CCU and ITU there is a need for them to up the occlusion. Also when a bolus is administered.

If a way could be found that this would all be taken into account with the drug database versus size versus rate calculation then by all means the occlusion could be removed. One less button for them to break.

We call them mckinley T34 here. If you are looking at these make sure you have a look at the price of the parts. You can not just change the LCD here it is full front. If you have the older type as we do you have to get the back aswell as the new front doesnt fit the old case. Circuits are different.

bcarlise
Im based in a 16 bed ICU we have set the max occlusion at 500mmhg for our syringe pumps and 750mmhg for infusion pumps the pump has a programed bolus that allows for higher occlusion for the duration of the bolus this iset on all pumps in the hospital (360 syringe 450 infusion approx ) and have had no problems

How do these databases get updated as new drugs become available? Will we still be able to update in (say) ten or twenty years time?

As an aside:- for the first time in my sheltered existence, I have just opened up a Baxter Colleague. Interesting pump mechanism ("shuttle"?), but not so easy to open up the unit. That is, a bit "fiddly" with the various connectors. No doubt I'll get the knack!

Both our infusions and syringes are unlocked. Staff have hopefully been trained to use the lowest occlusion setting to achieve their aim.

We set them when they leave here at midrange (approx 500mmHg).

Billy

For those interested in this thread it's World Usability Day (WUD). Yes, yes there seems to be a day for everything nowadays but more seriously this gives people interested in making usable technology a chance to promote their field and discuss important HCI issues. The theme for this year is communication, and at UCLIC we are lucky enough to be able to showcase a great piece of communication by our MSc students on UCL TV.

2min film: "Why buttons go bad" shows the importance of user-centred design. The film and its relation to patient safety can be found in this blog post http://chi-med.posterous.com/an-amusing-short-film-illustrating-the-import

Please consider sharing this to inspire people to think more about the tools, technologies and interfaces around them; those that are user-friendly and those that are less so.

Two usability issues from recent discussions:

1) PCA - as I understand it patients have a button that they can press to administer pain relief. However, there are controls as to how much drug the pump will deliver no matter how much the patient presses the button, e.g. it ignores the extra button presses beyond a limit. Does anyone know what sort of feedback, if any, these devices give the patient and clinical staff?? Does it just ignore the patient with no feedback? Does it make a dull sound to show it cannot comply? Does the nurse have access to how many times the patient has requested a dose?
These different configurations could have important implications for the patient's own management of dosing/pain and the awareness of patient comfort by clinical staff.

2) ECG monitors. I heard ECG machines are commonly reported broken when actually the wrong paper has been used or it has been put in upside down. Apparently this is a persistent and annoying problem. Do other's agree this is persistent? Any good ideas on alleviating this issue?

Thanks

How about this for item 2):- always make a note of which way round the old paper fitted, before trying to replace it.

Hint:- could this be why so many healthcare professionals have university degrees these days?

As far as PCA goes, I only have experience of the ubiquitous Graseby 3300s (never got my hands on an Omnifuse PCS sadly) - other pumps may do things slightly differently.

With the 3300, you get the same beep from every button press, I guess just so the patient knows it was acknowleged and doesn't lie there pressing the button harder and harder. And yes it does record how many requests, and how many were actually given.

As for ECG, the main problem I get is phantom lead faults. Incorrect / upside down paper is annoying, but not what I would call persistent (of course, larger hospitals may be more prone to it simply due to there being more equipment, therefore more opportunity for it to happen).

However, since you're asking specifically about that;
The incorrect paper seems to stem from the fact that while most units take (or can accommodate) the same size of paper, the page markers vary - it might be a hole, it might be a small black patch, positions may vary. Now, if manufacturers would agree a standard, that would be ideal, but that won't happen. They want you to buy their paper, after all.
Besides, it would take years for that to trickle down and replace all the older machines with their many different ways of determining where the next page starts.

Standardising on one model within a hospital would help, although obviously it can't prevent the wrong paper being ordered. But equipment replacement is piecemeal at the best of times, and is often dictated by who will give you the best deal, so you end up with a smorgasbord of stuff.

As for upside down paper, to me that's a training thing. If the machine doesn't immediately give the expected result, the operator should maybe check one or two things before picking up the phone. Maybe that's a rather optimistic viewpoint though...

Hello

Yes the Graseby beeps, from memory it displays how long to the next available bolous and it definitively records how many successful and unsuccessful doses it gives, i have seen cases where the patient has pressed the button unsuccessfully hundreds of times. Whilst the patient is still under the after effects of sedation and in pain they often keep pressing the button until the pain stops, obviously from pressing the button to actually receiving the pain relief takes time.

On the ECGs you can add the problem of staff ordering the wrong size American/English paper to Gordovan's list especially when the same model can be set up for different paper. Most of the problems I’ve seen stem from lack of or poor training regarding the taking ECGs, but I’ve also seen way to many paper loading problems over the years.

Lee

More movies!

This is a great video of a widely used device which shows interaction problems - thanks for sharing Geoff! (I had seen the video before but didn't realise that Harold had put it on YouTube). I do like a good video

Firstly, is there really a problem? The anaethetist gets his job done after all.

Is this common? Intuitively one could imagine lots of interruptions where this timeout becomes an issue - perversely this may be at its worse at the most critical times! In more normal situations, how many restarts happen needlessly around the country/world everyday - in the video it happened twice in just one operation!?

Does the anaethetist just need more training, surely he should know how to use the device properly and press cancel rather than switching it on and off to restart.

Why is there this timeout function anyway? Couldn't the number entry be better designed so we don't have this problem and the need for training to cope with it?

How about a return to the thumbwheel rate switches, as used on Imed's first volumetric pump (the 922)? No room (excuse) for dumb mistakes there!

Meanwhile, anyone interested in (older) infusion devices in general might like to have a browse through this one.

This thread seems to have gone a bit quiet ...

... so here's another movie.

And here is the Q-Core website.

Been quiet on the forum for a while but the project is still moving along. To mark World Usability Day last week one of our professors gave a public lecture on usability, human error and patient safety so I thought I'd share it for those that are interested:

Very interesting! And thanks for sharing that with us.

Here are a few of the points I picked up there:-

1) Checklists are still given the thumbs-up (all engineers make lists, do they not?)
2) The realisation that nurses are too busy caring for patients to be focusing on technology
3) In general, hardware does not get Beta tested in the way that (most) software does
4) Some NHS Trusts are apparently negligent in the way they have un-checked (insufficiently trained, or just incompetent) people working in home-care settings

And, right at the end, the speaker could not give a definition for one of the words used throughout the presentation!

Meanwhile, here are a few questions of my own:-

1) How (and when) do manufacturers decide to introduce new equipment?
2) How is new (to market) checked out (in the UK) before being offered for sale?
3) Can anyone cite an example in recent times when it was the equipment (or its design) that was clearly at fault in any of the tragic cases reported in the press?
4) How about the ones that the Public were never told about?
5) In cases where equipment was clearly at fault (if any) ... would it have been likely that such faults would have been picked up during regular maintenance?
6) In cases where user error had tragic consequences ... would it have been likely that such errors would have been avoided by user training on the equipment involved?

Hi Geoff,

Thanks for taking the time to watch the video. Following some of your points I thought this might be interesting: I stand to be corrected but I think there is a gap between the current understanding of 'user error' and 'device error' that people don't readily look at. The default position appears to be that if there is a mistake and the device has done what it says in the manual then the nurse needs more training.

Psychological studies show that training is effective in reducing mistakes and knowledge based errors, but not effective in reducing slip errors. Knowledge based errors are things related to not knowing how to work something, e.g. not knowing how to operate a new blood glucose monitor. Slip based errors are errors associated with pressing the wrong button because they look the same, mixing things up like putting your orange juice in your cornflakes, forgetting where you have left the patient notes or controlled drugs cupboard keys etc.

The example (which is now dated so it wouldn't happen now but it shows the gap I am referring to): A nurse was in a very busy emergency situation where there are multiple audio alerts going off and she has to do too much. One of the infusion pump alerts is easily silenced by pressing the alarm silence button. The reason the pump alert was going off was that it wanted more information (I can't recall the reason). Whilst dealing with all the other things the nurse silenced this alarm three times as it repeatedly went off. However, she was actually raising the rate of the infusion each time she silenced the pump because she was hitting the up button rather than the silence button - this also meant the alarm was silenced.
The view of the expert reviewing this case was that it was clearly a user error i.e. the nurse should have been more careful about the buttons they pressed and should always check the device after interacting with it. Therefore there is no issue with the device and the nurse is blamed and needs more training (if she still has a job).
I thought that the design was at least partly at fault. The expert considered my view and conceded that the buttons could have been placed further apart. However, I wanted him to go further than this, why did the up button silence the alarm in this way which led to the mistake? A good design would have prevented this error. To me this interaction set the nurse up to fail.

I can't recall where I read it but we can use what is called a substitution rule in these instances. We don't know enough about the incident, but from what we do know, if we substitute that nurse with another nurse could we reasonably expect them to act in the same way. I would say 'yes' in this busy situation. This to me points to the design being at fault so it shouldn't only be classed as a user error.

So I would argue that this is not just a 'user error,' and it is not a 'device error' as it follows the instructions. I think there is a gap - a 'usability error' or 'design error' or 'interaction error' whatever you want to call it. I think these sorts of errors are confidently and wrongly categorised similar to the expert above, and that people don't readily look for usability issues. I'd be interested to hear people's views on this.

Hi Dominic

Have a look at this video ( in fact the site is good ) :

http://www.risky-business.com/talk-18-story-of-bethany-bowen-2.html

Note the comments about the investigation and the expert witness into this death by the mother ( very tragic )

Geoff these could answer some of your questions if you can get copies :

• Wang B, Fedele J, Pridgen B, Rui T, Barnett L, Granade C, Helfrich R, Stephenson B, Lesueur D, Huffman T, Wakefield JR, Hertzler LW & Poplin B. Evidence-Based Maintenance: Part III, Enhancing patient safety using failure code analysis, J Clin Eng, 2011, 36:72-84.

• Wang B, Fedele J, Pridgen B, Rui T, Barnett L, Granade C, Helfrich R, Stephenson B, Lesueur D, Huffman T, Wakefield JR, Hertzler LW & Poplin B. Evidence-Based Maintenance: Part II - Comparing maintenance strategies using failure codes, J Clin Eng, 2010, 35:223-230.

• Wang B, Fedele J, Pridgen B, Rui T, Barnett L, Granade C, Helfrich R, Stephenson B, Lesueur D, Huffman T, Wakefield JR, Hertzler LW & Poplin B. Evidence-Based Maintenance: Part I - Measuring maintenance effectiveness with failure codes, J Clin Eng, 2010, 35:132-144

@Dominic: good post!

Yes, we all make mistakes. But it is my sense that the "kit" gets blamed more often than not just to ease the embarrassment that other options would involve!

In the kind of examples you cite (and in the Real World), people often work in chaos. Sometimes they are under tremendous pressure, but sometimes (based on my own observations, at least) that crazy situation appears to be "self inflicted".

"Do one thing at a time (and do it properly)" wouldn't be a bad idea, in my view.

Easy to say sitting here, perhaps. But there needs to be a discipline (oh yes, there's that word again) about the workplace (any workplace), especially when working as part of a team effort, and/or in critical, stressful etc. situations.

So, what's the answer:- staffing levels! Back to the Numbers Game, then.

@George: thanks for that. Yes, I know a bit about "codes" (job codes, and the like). Believe me, when it comes to listing, numbering, classifying, categorising and coding ... then I'm your man! Sometimes to the n'th degree. How sad is that?

But take my tip:- sometimes analysing stuff to death can end up as mere procrastination. And, again, I write from great (chronic) experience of that, believe me.

If you can provide links to anything like those papers you mention, I would be very interested, Mate. Although I suspect that what those august academically inclined blokes are saying amounts to the same old stuff.

That's the trouble with research papers and the like:- when are we going to see some practical benefit from it all at the Sharp End?

1) What's the point of measuring maintenance if you're not doing it in the first place?
2) What's the point of measuring maintenance if you already know you're unlikely to ever have enough resources to do it properly?
3) What's the point of measuring maintenance when the Bean Counters (suits) can't understand what you're talking about anyway?
4) Etc. etc. ...

There's some interesting stuff at that link you gave, George. I'll take a look later on (when we're into the "quiet phase" of the day). But ... surgeons feeling infallible? Well, they need to be confident in their approach, do they not? After all, it's no good them dithering about all day, is it?

Surely competence and confidence go hand in hand. But when does confidence become over-confidence? And at what point does over-confidence become arrogance? And ... how often should they be let get away with it by the rest of the team? Now, there's a whole new topic for debate!

Thanks for sharing that video George - very moving and powerful.

There really is a need for people to admit their fallibility and try to create a better culture of openness and learning. There are big difficulties in changing the culture but after having just watched the video I wouldn't want to detract from its message. I hope others watch it too.

I posted this previously but this video by Martin Bromley makes a nice companion to Clare Bowen's tragic story: http://video.google.com/videoplay?docid=-6738698910009425483#
Martin Bromley is also trying to spread the lessons of the accident his wife was involved in through the Clinical Human Factors Group http://www.chfg.org/ . These people should really be applauded for their efforts.

Hi Dominic ,

Yes I have seen that one – notice how the nursing staff are ignored ( just like in this case http://www.courts.qld.gov.au/__data/assets/pdf_file/0004/86755/cif-mcvey-os-20081128.pdf - page 10 & 27 ) - and in the case of Clare Bowen – the expert witness had never used that particular device ??? ( for those who haven’t watched it )

As a hospital based biomeds, competent nursing staff are our partners in safe outcomes with respect to equipment 99% of the time – we all know this and many of us treat Nurses with the same arrogance when it comes to User Errors – and at the same time having never to deal with equipment issues under the type of pressure you exposed to in the Bromley video – its easy criticise users when you got a device in your workshop , a cuppa and some Mozart in the background .

A competent nurse can set the tone and manage all the background chaos in theatre when things are not going well and the tend to be first responders to most incidents in hospitals – I setup and demonstrated Phaco machines for Allergan (back then) for 8 years (for those who don’t know – most patients are awake so there are lots of hand signals, glares and short abrupt statements when things go wrong )

So rather than just talk about it - I have done this : http://bioclinicalservices.com.au/

( sorry to harp on this Geoff ) – making sure nurses have access to product specific info 24/7 is , in my opinion , the first step to changing the culture of all staff to medical equipment

Geoff ;

You quite right – if we not doing it , what’s the point – these guys looked a 8 hospitals and + 40 000 devices - User Error ( they spoke about NFF and User induced errors ) was the most common issue with respect all device categories – PM detected very few device errors

Its interesting – the Alaris Asena has a recommended 3 yr pm schedule and most other syringe drivers have a 12 month one – and I don’t really see much different inside each – and that’s the issue when resources are limited .

I think the point is that someone needs to go first with respect to putting their neck out and saying “ we don’t have do stuff for the sake of doing it “ ( I’m not saying they are first )

Maybe we need to do something else – which adds value – eg No Fault Found = User Error before someone gets hurt ( most of the time ) - lets count NFF rather than the number of items tagged or not tagged

I think most people who make a effort to post stuff on this list are on the same page – as biomeds we just seriously disorganised ( in fact biomeds [ and other healthcare people ] tend to be slow adopters and if we , as a group , were involved in the internet you would be pen pal [ remember those days ] )

http://en.wikipedia.org/wiki/Ignaz_Semmelweis

How did you know that Dr.Semmelweis is one of my heroes?

You right about the nurses. We are there to support them (not the other way around).

But ... regarding PM, the answer is:- Risk-Based (sorry to harp on about that, but I have looked at it all long and hard over the years, and have been unable to come up with anything better).