Human Factors in Healthcare Blog

I just started hearing that stupid brake squeal from the front of my 10-year old truck…again! The new brake components installed a few months ago were to replace the new brake components installed a few months before that.  I suppose I will have to go BACK IN in order to sell this pretty nice looking, but pig-sounding, truck.

Returning your human body back to the hospital is probably even more irritating.  Did “they” do something wrong? Did you?  Did both? We might decide the “bounce-back” is no one’s problem.  This might be the nicey-nice route, but now there is no incentive on your part or their part to do anything different (i.e., it will happen again).

Reducing readmissions is a key part of health reform and many quality efforts.  Bob Wachter (of Wachter’s World), has a pretty clear discussion of this at his Blog.

Question:  Can we reduce discharge “bounce-backs” with huge computer systems, 85 million pseudo-checklists, and a pile of brussel sprouts (think sulphury tasting carrots)?  I am not sure…

My wife has been discharged from same day and overnight hospital stays many times in the past few years.  What did the healthcare staff do to keep the complications from medical procedures and strong medications from pushing her back into the hospital?  I will save that for her to describe in specific detail another day.

Overall, however, the main tools to tell us how to avoid “bounce-back” were verbal instructions (doc), verbal instructions (nurse), written instructions (standard-printed), written instructions (hand-written), and whatever was printed on the medical items (bandages) or meds we were given.  Oh, and whatever was left over in my head from medical school-residency 25 years ago; and whatever she might have read on consumer or medical web sites or medical journals.  Lastly, in some cases we would get informal consult or input from my medical pals.

Depending on how you want to count, that is 8 or 9 sources!!!  I can hear some of you readers saying, “go with what the doctor says, and downplay the others”.  Some might advise, “go with the handwritten, it is the most customized to your situation”.   Others will recommend, “consumer web sites have the gold nuggets from real patient experiences and tips.  Those doctors and nurses have never REALLY experienced XYZ disease, what do they know.”

Tools (computer or otherwise) need to help streamline or sort this out in a more understandable, easy to use way.  What do I mean by easy to use?  You need usability testing by patients and their caregivers — most realistically when they are sleep deprived, scared, in pain, etc.  Not a simple task to do this usability testing, but CRITICAL to success.

You know that area of your house outside where the phone, electric, gas, etc come into the house (basement)?  Well, do NOT try to use your weed whacker to clear out weeds.  The weed whacker might have problems cutting tall or wet grass, but it chews right through the outer coating of your phone line.  It’s actually a delicate “chewing”, since it just exposes the copper wire to the rain and snow, so that the phone failure takes 4.75 months to happen.  By that time, you have forgotten that you meant to call to have the gnarled up wiring fixed in a routine, non-emergent fashion.

Do any of you know what happens when you call to get phone service fixed?  Yep, it is very much like trying to get an urgent (but not emergent) medical condition addressed by the primary care doctor you have not seen in 6 years.  You go through voice prompt systems, type in you phone or medical number 50 times, and finally talk to someone who tells you that someone else will call you back about the plan…  All that seems understandable, since these customer service people need to be efficient with their time and that of the professionals that they protect, er, help manage finite resources.  You realize the problem, though, when the first person can’t tell you who the second person is, or how to call that person if they don’t call.  Worse yet, you miss their call and you go the “back of the line”, even though you have given them every cell and land line number you own.

Like delayed flights in the last blog post, human factors engineering analysis can give us some clues about design features to improve timeliness of care (or return to phone service).

1) Offer the same ability to track your process as Fed Ex gives you to track your $29 vase of roses.

2) At the risk of being a broken record:  offer clear alternatives to give choice to fit a person’s need.  Offer a phone message every 4 hours, or text, or email…  Or, give a direct number to call if they have not called in X amount of time.

3) Offer up a web page form so you can enter the key details about your medical condition (or phone disaster).  I see why this is avoided, since clinics worry they would be liable for reacting slowly to ominous symptoms.  Okay, but there needs to be a better method than verbally, and less systematically, having someone go over the same story.  Think of this as trying to design stuff to create the best shared model about what is going on, and what should we do next.

You look up over the top of the newspaper you planned to read later.  You see the first sign of trouble:  two agents backing away from the counter whispering to each other.  Yep, it is 10 minutes past “official time”, and that darn door is not propped open.  You could ask the agents, but you know they will not tell you anything unless the delay becomes tragi-comic.  Perhaps not until they are sure you are 30 minutes late.  Sometimes the folks working in back will bypass the frontline agents, so you decide to check other sources of “truth” about how late you might get home.  Hopefully, you have “planned” to be late and do not have critical child care or work tasks.  Suddenly, you feel that hot, Elaine Benes-like, irritation welling up.  PLEASE MOVE!  Or, just tell us when we will all get moving.

Of course I am tricking you!  I am talking about almost any encounter with healthcare settings.  Although my level of irritation at delay last night in Providence did inspire this posting (60 minutes delay, not as bad as 12 hours in Baltimore last year).

There are many human factors engineering lessons about how to address delays when seeing a physician, getting a scan, or other services. 

First, vague concepts like “transparency” or “customer first” are hard to translate into design.  You end up with pictures of people rowing on the wall and signs that say “speak up if no one has talked to you in 30 minutes”.  Instead, think of delays with your computer.  Which helps more:  a) the little rotating symbol that indirectly indicates a web page might be loading; or b) a progress bar and estimate of time for download.

Second, people want clear options with costs and benefits that make sense to their condition.  NOT a good example of options: 1) stick around in earshot of a person you have never met who will yell your name over the top of other noises; 2) leave the area and lose your spot in the queue.  More useful options: 1) stay in earshot to see the surgeon and visit loved one in post-op in seconds; 2) stay within 5 minute walk and carry pager; and 3) give us your email for us to send you the outcome, and see your not-as-much-loved one upon discharge.

Third, situational awareness of those serving and those being served is tricky to do, but everyone benefits.  Do you think that gate agent last night wanted someone screaming at her when we were only 30 minutes late?  Why don’t they have a status light that is

• Green when they have actionable information and are about to tell you,
• Yellow when they have only partial information, and not actionable.  However, yellow is a promise a “green” light within 5 minutes
• Red is when everyone is the dark, and communication and trouble-shooting are broken.  It is always accompanied by a best guess number.  You are free to rebook at a $100 fee; or get a pager and go eat (they will pay if the pager goes off before you eat); or get a $50 certificate and stay in the area.

I know that last set of options would not work, but it tries to get at the idea of all of the key people having the best situational awareness possible.  Your thoughts?

- John

My son decided to stress test his parents with 60% body rash last Saturday.  We keep a pretty “clean” and food consistent house due to our daughters allergies to 9 food ingredients (from peanuts to poppy seeds!).  So, we could not figure out the rash and were worried it could “spread” to his respiratory system (this has happened with our daughter).  On Sunday, 5AM, his rash worsened and moved into his eyelid area and mouth.

Every time I visit the emergency room with a family member, I cannot dismiss the fact I know that colleagues of mine study the relative design chaos that “lives here”.  Plus, sifting out the bad illnesses from horrible illnesses from run-of-the-mill illnesses is not that easy.  Missing diagnosis is a common reliability issue with super experts, much less someone half asleep and over-worked.

In short, after six physicians in three settings across 2 days saw my son with 40-60% body and facial rash, we took prednisilone and Benadryl until he was bouncing off the walls!  Even though the medications did not seem to work at first, in a few days he was back to his normal bouncing-off-the-walls self, with no rash.

The hardest thing for my wife and I was keeping track of the medication and dosing.  6.25 ml of Prednisilone…really?   10 mL Benadryl as every 6 hours, 5mL Ranitidine as every 8 hours, and 6.25 (then 5mL) Prednisilone as every 12 hours (or twice a day).  Oh, and sometimes we forgot one and then gave it, but 2 hours off cycle.  Needless to say, we used paper medication records to keep track and still messed up (double doses and missed doses).  Anyone who can figure out how to design a way out of this design mess, in a robust, usable, and well-accepted by physicians, will be a billionaire!

Oh, yeah, the markings on the picture of that medication measuring cup are kind of hard to see.  REALLY exciting at 2AM!