Thursday, January 23, 2014

Healthcare and learning technology, i.e., HIPAA crit

     With all the new technology everywhere you look, why is it that intelligence in using it is so rare?
                                        --- me

I visited Carle's Convenient Care a week ago Tuesday.  The visit itself went reasonably well.  Some of the details will be included below, but just to explicate the learning technology issues.  Yesterday, I received an email with a link to a survey about the visit.  I suppose that receiving the survey is an adjunct of my using the MyCarle portal for communication about my health information.  I wanted to be cooperative so I started to complete the thing.  But I found it so poorly constructed that after it rubbed me the wrong way initially I got more and more dismayed until I became completely disgusted, at which time I closed the browser without submitting it.  This post considers what purpose a reasonable alternative survey might address and how that purpose might be designed into the survey.

There are some visits to the doctor that really are one and done.  One evening when the kids were young, my wife was doing something away from home so I brought the boys over to Burwash Park to play.  There is a wooden structure and one of the boys got a splinter by dragging his hand over it.  My wife not being around and she being the splinter remover in the family, I needed an alternative solution to the problem.  So I took the kids to Convenient Care.  I was hoping a nurse would help us.  But it wasn't very crowded and we actually saw a doctor.  She had the hand soak for ten minutes and then removed the splinter without causing additional pain for my son.  Wunderbar!  I don't even think they charged a co-pay for that.  I'd give a big thumbs up all around.

There are other visits to the doctor where a return visit is unlikely, but where a treatment is required to provide a remedy.  For example, the patient goes to the dermatologist because of a rash on his wrist that itches and annoys him.  The doctor prescribes a cream to be applied topically twice a day for three weeks.  After two weeks or so the patient should have a reasonably good idea whether the treatment is effective.  This is a good time to survey the patient, who can then reconsider the office visit in light of the treatment.

Most of the adult medicine I've experienced has been for chronic conditions, such as my arthritis.  The most recent visit happened because the arthritis became activated in a new site - my neck and left arm.  And just like the first time arthritis was identified in December 2009, when what I had thought was sciatica proved to be arthritis and bone spurs in both hips and the lower back too, this time around I thought it simply was a bad stiff neck that resulted from having my head at an awkward angle while sleeping.  I might have been less clueless when it didn't go away after a few days, as stiff necks usually do, but instead I was confused and anxious.  Chronic conditions require diagnosis, prognosis, treatment, and coping strategies.  Further there must be periodic reassessment because the situation is not static.  Part and parcel with all of this is learning - for the patient and the healthcare provider.  That learning is facilitated by dialog between the two.

Do's on the survey

1.  Ask questions about what the patient cares about - his medical situation.
2.  Situate questions about the care received in light of the patient's medical situation.
3.  As a practical matter, (1) and (2) above imply few if any Likert-style questions and at least one and perhaps a few short paragraph questions.
4.  Conceive of the survey as part of the dialog, with the aim of improving care for this patient and other patients as well who are in similar situations.
5.  Share findings with patients, both participants and non.

Don'ts on the survey

a.  Ask questions about the consultation and care the patient received in a manner disembodied from the patient's medical situation.  The patient has no reason to provide this information.
b.  In an undifferentiated way ask about what  the nurse did and what the doctor did.  Their roles are different.  The questions need to be written in a way that reflects this.
c.  Use the the survey responses as thumbs up or thumbs down indicators on the providers' performance.  Take a clue from Higher Ed where Course Evaluation Questionnaires are used for this purpose, but are ineffective.  The students don't want to complete them and they lead to grade inflation.
d.  Collect information that nobody will look at.
e.  Collect information that is already available by other means.

Let me review my recent visit and follow up in light of the above.  The only part of the visit itself that could have been done better happened between when I first saw the doctor and when they took me to do x-rays.  Apparently there was a queue for that.  I stayed in the examination room for about a half hour and then decided to find out what was going on for myself.  By coincidence, they were coming to take me to x-ray just then.  It was annoying not knowing about the queue, but it didn't have any impact on the diagnosis whatsoever.  In other words, it was small potatoes.  If the information is collected with no reference to the patient's medical situation, and some of this information is patient attitude about the visit, then it may not be possible to distinguish small potato causes of patient attitude from more serious concerns.  That's just bad social science.

The doctor told me at the time that while she would do a cursory read of the x-rays then and there, to come up with an immediate treatment for the pain, their standard process was for an orthopedist to do a more in depth review, write a report, and then they'd contact me with the results.  Those results were more detailed but essentially confirmed what the doctor first told me.  They also were written in a way that I couldn't decipher them, which speaks to the next issue.

Because I had gone through the ringer after my rotator cuff repair last year (for the other arm) I was and remain extremely reluctant to find a surgical solution to the current issue.  So when the doctor suggested to schedule a follow up visit on my behalf, this time with my primary care physician, I rejected that offer as I feared it would accelerate matters toward surgery.  In retrospect, my decision was a mistake.  The doctor prescribed an anti-inflammatory drug - helpful for dealing with the pain but not useful for addressing the underlying cause.  That does need to be addressed and having realized that I've since scheduled an appointment with my primary care doctor.  But what if I hadn't changed my mind?  Should the process be neutral on that score?  Or should it prod me to be more pro-active?  That's the sort of question which is important to consider when looking at the survey as a piece of the overall information flow.

Finally, let me turn to the HIPAA crit part of my post title and tie that to part (e) of the don'ts list.   At the beginning of the form, they asked for a variety of pieces of information that together amounted to pinpointing when my appointment occurred.  One of those was the date.  They asked me to type that in even though it was already included elsewhere on the form!

It's harder for me to tell whether they could readily get the rest of the information they asked for some other way.  All the information is collected when you sign in for Convenient Care and make your co-pay.  Is any of that information actually HIPAA protected?  I really don't know.  The survey itself was not administrated by Carle but rather by a third party and even if that information flow is not HIPAA protected it may be sensible to not share it.  (Credit card information is often collected as part of making the co-pay.  If sharing appointment date information would put the credit card info more at risk, then not sharing the appointment date info makes sense.)

But it's also true that Carle itself tends to blithely ignore the impact on the patient from asking for the same bits of information again and again.  Immediately after checking in at Convenient Care, the patient is given a form on a clipboard to write up what the problem is for the nurse.  There is much duplication here with what is provided in check in.  In my head I've rationalized this because the second tome around all the information provided is hand written and there is no ready way to combine the novel bits with the stuff previously provided that is in digital form at this point, though I'm not sure I believe my own rationale here.

After my vitals were checked, I was surprised that the nurse asked me for my preferred pharmacy for picking up prescriptions.  Carle has had that information on me for years.  For whatever reason, that information wasn't shared with Convenient Care.  I have no idea why that's true, though it is evident that on a transaction by transaction basis it is far less costly to ask the patient for the same information over and over again.

But overall doing so conveys the impression that the data flows within the organization haven't been well managed.  Likewise, returning to the design of the survey, asking the patient for what appears to be redundant information creates the appearance that survey design wasn't well considered.  So if the data actually can't be had in other ways and if it really is necessary to collect, then the patient is owed a brief explanation as to why and that explanation should be included in the survey right before those questions are asked.

Treat the survey participants as adults and show respect for their time and effort in completing the survey.   It's the critical aspect of survey design.

No comments: