Premise
Observations on coping skills and task completion efficiency in teachers and learners in a family medicine teaching clinic.
Main points
Work in progress. Draft outline only. Not for full distribution.
Most docs hate them but are able to keep up with routine Tasks in the EMR
We don’t teach or assess them at all, apart from casual on-the-fly modeling, and even this is minimal because residents are not included in the regular stream of tasks.
Importance of Conscientiousness Index as a marker for low performers
Mention DIRT concept and proposal. Delegated Instructional Resident Tasks. See obsidian://open?vault=PKM%20and%20Obsidian&file=Writing%20stuff%2FOLab%20book%2FDelegated%20Instructional%20Resident%20Tasks%20(DIRT)%20project and email discourse at Email thread on Resident Tasks Project.docx — could you create another OLab mini-case series to test whether residents do their 4D on tasks appropriately? Test their thinking of how to handle a delegated Task, including defer (by changing the due date) and by how much. Created stub map 1993 for this.
Low-low on the Uncertainty-Consequence quadrant
Results from Lab Rat Doddle (aka Rushing Roulette) —
Note that Rushing Roulette was published in PeerJ but not peer-reviewed. so find what we said then. Diverted by the Arduino and stress testing. Good intro phrases re non-exam context. Does not mention ConscIndex stuff at all. Mentions Covey quadrant III, urgent but not important. Compare this quadrant with our quadrants. Used 30 second clock, not the reducing clock that we used with Arduino. 20 cases, 6 with minimally relevant abnormalities, 3 more significant, and 1 which needed a patient recall. Touched upon Receiver-Operator curves – did not do enough of this to be useful. Omit this. Higher degree of intervention with earlier learners.
Covey Quadrants vs Eisenhower matrix. Covey Matrix Vs Eisenhower: What Is The Difference? — similar grid. Did Covey plagiarize? Not the same as ours.
Also published the map to Borealis: Rushing Roulette – OLab book materials — but no report details to go with it.
Discuss the purpose of the other maps in this series:
- Make it stop
- It’s probably nothing
- Lab Rat Doddle
- Also …
- Can you hold please? (although this was a bit different)
- The Secretary Problem
- War of Fog
Pasted segments
There are many other job functions in daily practice: Why don’t we teach and assess these?
- lab investigations: ordering and review of results (this part of HEIDR)
- radiology ordering (where DI recommendations magnify the costs)
- referrals processing
- summarizing reports e.g. distilling essential elements from the consult letter, imaging reports, discharge summaries, patient transfer summaries
Also preparing learners for changing practice dynamics – for instance …
Note that summarizing is the sort of task that can now be delegated to a large language model (LLM) or generative pre-trained transformer (GPT)? We suspect that this will increasingly occur, because it is easier. It might not be entirely accurate but may be good enough for most purposes in our quest to be “efficient” (ref Thermodynamics of Medical Education). For example, many physicians have lost the ability to do ECG interpretation and simply rely on the machine’s interpretation, even though we know these to be rather error prone.
Task management not taught modeled or assessed
We assume that the boring daily tasks will get done and that learners will figure it out. But, as with our earlier examples in Chapter 20 on decision-making and dysrationalia, some do not. And this is NOT picked up.
In the bottom left quadrant (low-low), this type of decision has little impact and this is probably why it is ignored. But dysfunctional decision-making in this quadrant is common. How common? We don’t really know because nobody is paying attention to it. No measurement = problem does not exist, to misquote (as seems to be commonplace) Deming, Kelvin and Drucker.
But we see the damaging effects in daily practice. This is much more common than realized once your awareness is raised. Worse, it will be completely hidden in isolated practice because nobody sees another doctor’s task list. The serious exceptions, at the extreme end, eventually come to light during regulator reviews of bad practice when a doctor has already been flagged.
What we have observed ourselves is the massive accumulation of Task lists in the EMR. Most doctors are a wee bit behind in this thankless chore – it might have been interesting to calculate the average lag. But when doing an in-system locum, a typical amount observed would be 1-2 weeks latency for most tasks. Some doctors would have a marked selection of more complex tasks that were less urgent.
However, there were a notable 1-2 docs out of 24 who were excessively behind on their Tasks… up to 18 months in one instance. Buried in amongst these were important and significant problems that needed action. DT was uncomfortable with how these were left ignored and tried to tackle some of these. The Task List Owner (TLO1) was quite angry about this and did not want her Task List touched. This raised a concern about liability because some of the clinical issues were significantly overdue for attention. The compromise was that DT was not expected to touch the delinquent Task List at all… which does not really address the underlying problem.
When asked about the overdue Tasks, TLO1 was not unconcerned. However, TLO1 felt that these Tasks were worthy of detailed consideration, not hurried task handling. This seemed to represent a displaced conscientiousness: TLO1 was fully intending to tackle these Tasks and was not ignoring them but also felt that they needed sufficient time and attention (which TLO1 did not have).
To really judge the significance of this, there would need to be an independent audit of the Tasks, comparing the issue, context and eventual plan. This would be a lot of work. TLO1 left the department /clinic? before this was resolved. It is not clear what happened to the unresolved Tasks.
For another user, whom we shall call TLO2, who was behind on Tasks, this created a lot of stress for TLO2 who felt overwhelmed and unable to keep up. Some simple coaching helped to recontextualize the priority and importance that TLO2 assigned to these Tasks.
TLO2 had become bogged down in trivial decisions with an over-emphasis on consequences where there was minimal difference in likely outcome. This might be considered as over-conscientiousness. TLO2 just about quit medicine over the stress arising from this task management and was very thankful for saving her career via the coaching applied.
We also observed people like TLO3 who are real experts at task management but don’t realize how much better they are at it or, more importantly, why they are better. Is what they do learnable? At least it can be modeled, even in a mentoring role, because it may not show up until well into practice. In modeling it, there is the risk of “I taught my dog to sing; he just ain’t learned yet”. And if we don’t measure it, how will we know if it is learnable?
Can everyone learn it? Our sense is that the majority can, especially with good modeling. But there will be outliers, with many reasons for this e.g. OCD-trait (“I must not miss anything”); or a previously experienced tragedy which causes subsequent over-sensitivity and guilt; excess scolding/belittling/punishment for previous errors or missed information clues. And there is likely to be a spectrum of skill here with our equivalents of dysrationalia or task-blindness, or the previously used term “aphantasia for tasks”.
Do all doctors need this? Maybe not but some jobs will need it more than others. Consider the chaos of a busy primary care office or emergency room where doctors will need it more. A nice, stable, regulated super-specialist’s office, with an hour per patient, and few interruptions, will need it much less. We tend to assume that learners will figure it out eventually. Like heuristics, it can be modeled and learned but there are some who are blind to this and cannot even see the problem: aphantasia for Tasks.
Task blindness, executive dysfunction, aphantasia for tasks
We likened this to an ‘aphantasia for tasks’: there are cognitive phenomena that serve as functional equivalents to aphantasia for tasks, often referred to as a “task blindness”. While aphantasia is specifically the inability to create mental visual images, this “task blindness” or “executive aphantasia” refers to an inability to mentally simulate, plan, or imagine a sequence of actions.
How Individuals Compensate
Just as people with visual aphantasia use verbal thought or factual recall to compensate, those with “task blindness” often rely on explicit checklists, external aids, such as calendars, timers, and diagrams. But in the above examples, the EMR provides all these aids and we still find some that struggle with “keeping on task”.
While aphantasia is considered a variation of human experience rather than a disability, these task-related limitations can make certain types of planning and memory-dependent tasks more challenging. But do all individuals compensate? Our experience is that some do not or do so poorly.
Some modeling and mentoring can help. For example, for many, the decision options for lab tasks could easily fit into the 4 action categories that we used in the Lab Rat Doddle case: (see below)
Is this just a variant of procrastination? Or is it a variant of the inability for some people to be timely?
This seems to be another spectrum of behaviours, which many people will oscillate within, just like others have noted a spectrum from aphantasia to hyperphantasia. See E Quill in Nature. These are rich research areas: how to construct a supportive context for efficient task resolution?; what are the internal reward structures, dopamine/serotonin pathways for each activity type? And some people have deficits where they don’t have access to some parts of the behavioral spectrum i.e. always procrastinating, frittering.
Quill does note, as have others, that aphantasia does not necessarily affect behaviours. When it comes down to it, for our purposes, we are more interested in the behaviours that result from these processes. We mostly measure how people behave, rather than how they think.
Will this become worse in years to come with the prevalence of dopamine-inducing fritter activities such as TikTok? Some would say that this is obviously “Yes”. The perils of Doomscrolling etc. But is there evidence to support this?
Or is this another example of the naysaying that is often cast towards many new activities (computer gaming etc), as amusingly portrayed in “One Million Ways to Die in the West”, with that great clip about the stick hoops (aka ‘hoop rolling’). A Million Ways to Die In The West – Stick Hoop
Excessive conscientiousness
We mentioned that there may be a spectrum of behaviours. Consider another of our colleagues, whom we shall call TLO4. TLO4 was already super conscientious and tended towards a higher degree of completion of tasks. He found that the reminders and added functions provided by the EMR to ensure that all tasks were completed and checked off simply reinforced the ever-extending possible improvements or tasks to be completed. Instead of working a 12-hour day (and only being paid for 8 hours), he was now working a 16-hour day and was exhausted.
In a way, he was an individual representation of the 27-hour problem. He quit the department because of this, which was an enormous waste of an award-winning full-service physician and a talented teacher. This was as much of a travesty and loss, as we see when teachers quit because of disastrous learners. (Did we refer to this in the earlier chapter about Lackadaisical Larry? If not, it is important to include this.)
Lab Rat Doddle case series
(Originally named Rushing Roulette but this name and approach have diverted attention away from the boring, plodding nature of routine lab tasks.) In this case series, we wanted to examine how participants would perform when not in the high stress context of an examination. We constructed a set of 20 mini-scenarios where the participant is presented with the typical amount of minimal background information that experienced clinicians consider when going through routine lab results. There was a small amount of time pressure: 10 minutes for 20 easy tasks, the pretext being that clinic will start soon.
There are 4 action categories that we used in the Lab Rat Doddle case:
- File it
- Let patient know
- To come in
- Act now
This was intentionally similar to the 4D method for dealing with email: Delete, Delegate, Defer, Do.
There was a mild hierarchy of effort involved from #1 to #4. There were only a very few actions that fit into category #4, and even with these, this would be something like “you need to adjust your Synthroid dose” as a phone call by the doctor.
Obviously, in actual practice, one might break these down into more nuanced actions. But the eventual outcome was unlikely to be much different between these, for the vast majority of the Tasks. The important point to get across was not to agonize over the fine details but to go ahead and act.
We found a range of behaviours but generally, less experienced medical students tended to overcall the situation, gradually trending through residents towards experienced teachers who were more efficient and limited in their energy expenditure. Not surprising in itself, but we also found quite a range within each group and considerable overlap between groups.
At the time of the study, apart from the reaction from one of the face validity testers (“that was so boring. It was just like doing lab task chores”, which was exactly what we were looking for), we only noted the expected range of actions from less to more experienced clinicians. Now that we are faced with the consequences of the career-destroying range of task completion problems described above, we can use this same series to assess for and mentor such weaknesses.
For example, Make It Stop explores the problems illustrated by our colleague TLO4, whose tendency was to follow up on every last possibility; or It’s probably nothing is useful for assessing those residents who tend to be minimizers, measuring their tendency to be overly dismissive and not act when they should.
Delegated Instructional Resident Tasks
Also include case reference to map 1993 for the DIRT project.
See case in OLab4: Delegated Instructional Resident Tasks (DIRT)