Crazy Ideas


A patient calls or emails me with a problem.  I talk with them over the course of a few days, using whatever form of communication works best.  Eventually, they need to come to the office to be seen – either for something needing to be done in-person (examination, procedure, or lab test), or because of the advantages of face-to-face communication.  At the visit, I not only deal with one problem, but there are other issues needing to be addressed.  Finally, after the visit, follow-up on the problem continues until it is either resolved, or at least is not causing much trouble.

So how do I document that?

In the past I would\’ve had a clear structure for the \”office visit\” and separate \”encounters\” for the documentation of the communication done outside of the office.  The latter would be done largely with narrative of the conversation, and some direct quotes from the patient.  The former, the \”office visit\” would include:

  • A re-telling of the story of the \”chief complaint\” and what\’s been happening that caused this encounter to be necessary.
  • A sifting through other symptoms and past-problems to see if there is any information hidden there that may be useful.
  • A documentation of past problems (already in the record) to support the thought process documented later in the visit.
  • An overview of the physical exam, again to support the  decisions made as a result of the visit.
  • A discussion of my thoughts on what I think is going on.
  • A telling of my plan on how to deal with this.
  • A list of any advice given, tests ordered, medications changed, prescriptions written, and follow-up as the details of that plan.
  • A signature at the end, attesting to the validity of what is contained in the note.

But here\’s the problem: it\’s not real.  I don\’t make all of my decisions based on the visit, and the patient\’s story is not limited to what they tell me.  Details may be left out because they are forgotten, questions aren\’t asked, or things just haven\’t happened yet.  This signed and sealed unit of care, represented as a full story, actually represents only fragments of the story, of many stories actually, and only as a moment on the continuum of the patient\’s care.

But there it stands, the office visit, the center of the patient\’s medical record.  It is what my past life defines as \”health care.\”  But for me now, it is an anachronism: an old-fashioned idea that has nothing to do with my present reality.  My care is no longer episodic, so why should my records be?  I no longer need \”visits\” as units of commerce, and no longer need \”problems\” as the goods for which I am paid.  This took me quite a while to figure out, and has me making some radical (crazy?) changes to how I think about care.

1.  Stop Artificially Defining Units of Care from My Perspective

The \”office visit\” is a unit defined from the physician\’s perspective.  This may actually undermine a clear thought process, though, fooling me into believing I\’ve finished something that isn\’t complete.  Instead of breaking care into a temporal unit, why not define it another way:

  1. The narrative – what is going on with the person.  The patient narrative goes over their lifetime, some of which happens in my office.  Narrative would best be a combination of patient input and my observations as a third-party.  My job is not necessarily to do the narration, but to guide it by asking questions and to prioritize that which I think is most salient.  Each \”problem\” may have its own narrative, but those lines often blur, as the chronic disease causes depression, and both the chronic disease and depression lead to fatigue and increased pain.  How do you divide those?  Do you need to?
  2. The investigation – both through my questions, examination, and through tests I order, I work to solve puzzles.  The patient wants me to make sense of confusing parts of their narrative, so I step in and investigate.  Tests are ordered for a reason, and exams are generally problem-driven.  One of the most important \”tests\” to do is to wait and see what happens.  Sometimes stories only become clear as they unfold, and it is my job to know when it\’s best to do nothing other than wait and observe.
  3. Interventions – medications, lifestyle changes, education, and procedures are what most people think of when they define \”health care.\”  Yet these are entirely driven by the narrative and investigation.  They add to the narrative and often become part of the investigation, but they never stand alone.  Well, they shouldn\’t stand alone, despite the fact that they are what doctors are largely paid for in this country.
  4. Destinations – what are our goals?  What is the goal of the patient in the bigger sense, and what are our goals for individual narratives?  The destination is the entire purpose of the doctor-patient interaction.  The patient doesn\’t like their destination and I am called on to help them go somewhere else, or they just want to be sure they are not inadvertently going to a bad locale.

Can I do a record based on these categories?  If I do, would it end up as a crazy jumbled mess?

2.  Find Strings to Tie the Pieces Together

Each part of the record of care is tied together in different ways, with different things that they share with the other pieces of information or narrative.

  1. Diseases – these were referred to as \”problems\” in my past life, a euphemism if there ever was one, but part of each record is usually devoted toward the prevention, treatment, or management of disease.  Narrative, investigation, intervention, and destination could address these aspects of disease.
  2. Symptoms – these also (perhaps more appropriately) were listed as \”problems\” in my past life.  They are how health directly impacts the patient.  Symptoms may or may be related to a one or many diseases, or may be mysterious.
  3. Risks – These are sometimes determined by choices of the patient (smoking, drinking) or not (family history).  The reason we at all care about smoking or family history is the risk they confer.
  4. Events – Big events, such as heart attacks or motor vehicle accidents can result from or cause disease, but they are tied together in a temporal way.  In the same way, an office visit is a temporal way to tie together interactions with a patient on a given day.  (I didn\’t say it\’s entirely bad to consider visits as units; it\’s just bad to only define them that way).
  5. Situation/Environment – What\’s going on in the other areas?  Is the person getting a divorce?  Is it allergy season?  Is there a flu epidemic?
  6. Population factors – some narratives, investigations, interventions, and destinations can transcend a single patient.  My ability to see the population as a whole will help me to give better individual care.

Where does this lead?

What does a record in such a world look like?  I can\’t say I\’ve come to a clear idea, but there are some things I\’ve started doing:

  1. Treating narrative as a separate entity – Whether communication happens via secure message, on the phone, or via email, narrative is narrative.  It is the fuel that propels good communication (my listening to it) and by which proper care is given.  Lumping narrative with the other areas tends to artificially categorize it.  The result is a separate documentation of narrative from exam, intervention, etc. in the record.
  2. Tying narrative and the other areas together using things they have in common.  I have started using diseases, symptoms, risk factors,  and events as tags, tagging narrative, consults, lab results, medications, and anything else with these fathers (so I can see all conversations, examinations, lab tests, consults, hospitalizations, etc. related to a person\’s diabetes, for example, or seeing all other parts related to a specific medication).
  3. Allowing free-flowing use of tags to give me different views of the same parts of the record.  In general, the more perspectives one has when looking at something, the more accurate the picture.  This means tagging needs to be simple (perhaps automatic) and robust.  It also means that I need to be able to quickly get those views when I (or the patient) need them.
  4. Working with the patient to come up with a destination – an overall plan of what they expect to get out of my care.
  5. Going over the progress toward those destinations on a regular basis.
  6. Making all of this fully available to my patients.  This assures the accuracy of the information and keeps it focused where it should be: on the person for whom this whole system is built.

Sounds crazy?  Sometimes I wonder.  But then I look at what my past life as a doctor involved, and what \”care\” was defined as, and I begin to grin.  That was crazy.  This is simply being myself, which, for once, doesn\’t seem so crazy.

5 thoughts on “Crazy Ideas”

  1. Mark Wheatley

    Try this – what you need maybe is a help desk ticketing system. Each time a patient interacts with you a ticket is opened. Into each ticket you can pour all the info you want about that interaction. So a patient contacts you about a matter. Poof a ticket opens. They can keep the ticket number, refer back to it, call back with more info, what have you. You can always refer back to it. Cross-reference it from other tickets if you like.

  2. I was also thinking help desk ticketing system, but rather that a ticket being an interaction, the ticket is the symptom/condition, and the times it is addressed (by whatever interaction) is a comment in the ticket. Also, the tickets can be cross referenced, eg. one symptom is related to another. Lab reports are attached to the ticket, and so on. Help desk ticket systems also can have defined workflows “investigating” “treating” “resolved” “chronic” etc. it also manages access control (eg a parent can be a “watcher” on a child’s “ticket” for their asthma and so on. A few days with software like JIRA and you’d be there (though of course they are not HIPA compliant, if that is a requirement).

  3. Not crazy and in fact it is very much what I do in case management. We “manage” the information which comes from all different sources; narrative, interview, patient, family, significant others, medical records, etc. We then formulate our list of problems (nursing diagnosis) and make a plan. Then we implement that plan (some interventions the patient will do, others that I will coordinate, and some with input from MD or other) then I obtain and evaluate the outcomes and revisit for potential changes in care. It is an ongoing process and one that is never limited to a face to face encounter. I typically get more information from my patients IN BETWEEN visits than they can remember AT the actual visit. I do have to give you kudos for the tagging though. That is ingenious!

  4. As an IT professional, what I see in your note are glimmers of systems analysis thinking. Information you collected in your past life doesn’t suit your current purposes. I would suggest that you develop and maintain a vision of your information needs and use that as a touchstone for your ongoing system and process design work. My simplistic view is that you need to collect, structure and retrieve information for two main reasons: to support the decision-making you need to do now, and to support decision-making you’ll need to do in the future. The needs for the short term may be much simpler, especially since you don’t have the coding for billing problem any more. The problem of selecting information to record now for decisions you’ll have to make in the future seems like it would be harder.
    The requirements tests in the short term might include ‘do I have enough of the right information to make the treatment decision I ought to make now?’. If you don’t think you have enough information, you may want more tests or need to ask more questions. At some point you conclude that you do have enough information to make the next treatment decision. Your system needs to hold the information you had in front of your for that point in time. You need that for ongoing process improvement (e.g. ‘how did I miss that?’, and ‘how did we get to this point with your condition’) and for any sort of legal defense.

    The requirements test for the longer term might include things like ‘what kinds of treatment decisions am I or my partner(s) likely to have to make for this patient?’ I imagine you’d collect baseline information on vitals and on emerging chronic conditions. I think you’d want to make sure you’re recording enough about current treatment decisions so that the current treatment can serve as useful input to future decisions. You’d make baseline assessments about the patient’s ability and willingness to follow instructions for medication, lifestyle changes, and other aspects of self-care and following your treatment plans. You’d make some minimal assessment of their mental state, both as a factor in their health and as a factor in what you ask them to do.

    If this kind of perspective resonates with you, you might want to recruit a local IT pro with systems and requirements analysis skills to serve as a sounding board for you. Some people with those skills might call themselves business analysts. Most programmers would have some skills in this area too.

    Best of luck


  5. Patients or Care-Givers should be able to open and update tickets in addition to the doctor. Anytime a ticket is updated (by patient, doctor, or care-giver) all parties should receive a notification.

Leave a Reply